Poster Session 2

 

Abstract

 

 Ground deformation occurs even in geologically stabile areas for reasons such as the swelling and shrinking of clay minerals or frost activity. Such changes can be considered potentially important geo-hazards when they pose threat to the stability of important infrastructures like roads, bridges, buildings and apartments. An interesting spatial setting can be found in counties like Finland where stable bedrock areas occur in close vicinity with such instable sedimentary areas where both frost and thaw dynamics as well as the shrinking and swelling the clay sediments may induce local-scale ground deformation.

Modern remote sensing techniques have made it possible to detect and monitor ground deformation within even short time periods in millimeter scale precision. In particular the synthetic aperture radar (SAR) systems that have been installed in Earth orbiting satellites provide an interesting opportunity. SAR Interferometric techniques (InSAR) which compare two or more SAR images acquired in different times have become an important, precise, appropriate, and low cost tool for the monitoring and mapping ground deformation.

The combined use of different InSAR techniques revealed various kinds of spatially and temporally accurate ground changes in this study area, Particularly the joint use of Earth Resource Satellite (ERS-1/2) and Advanced Environment Satellite ENVISAT SAR, Sentinel-1  (Synthetic Aperture Radar) C-band data gave the potential to monitor, detect and estimate the ground deformation during both long and short-term with high spatial and temporal resolutions in Turku.

The present research is focused at the detailed scale detecting and analyzing of spatial ground deformation processes using a novel approach of long-term InSAR.

The research hypothesis is that many parts of the city of Turku suffer from observable ground deformation dynamics and that damages in urban constructions will be found in those areas. The research highlights the resulting economic and safety hazards and offers methodological guidelines and results to contribute to spatial planning. It also helps to find the best managing procedure to reclaim and protect infrastructure degradation and deterioration.

Due to the lateral erosion of the river, a 40-50 meters high high bluff was formed along
the right bank of the Danube in the Late Pleistocene and Holocene. The high bluff is
composed up from Late Miocene (Pannonian) sand, silt and clay layers, and
Pleistocene loess and paleosoil sequences. The high bluff was dissected by valleys
and further eroded by the river during historical times. Due to the lateral erosion of the
Danube, the geological and hydrological settings of the area, this region has became
one of the most landslide affected region of Hungary. Settlements and infrastructure
built on the fossil landslide masses or close to the edge of the high bluff are
endangered by different kind of mass movements (e.g. deep-seated rotational
landslides, lateral spreading) till nowadays. However, intense research of the
endangered high bluffs has been started just after the catastrophic landslide events of
the 1960s and 1970s.
Despite of the intense research activity and efforts designed to stabilize the built-up
areas during the last 50 years, several landslide event and building damage were
reported from the area. New prevention campaigns were started by local municipalities
in the last years, nevertheless, integrated monitoring of the affected areas are still lack
here. Researches and recent monitoring surveys are based on field and DGPS
surveys and deal with only very specified local areas, moreover, only mass movements
of the Pentele Loess Plateau were investigated, using DInSAR techniques.
Integrated landslide monitoring of the built-up areas along the high bluffs is now
feasible, using DInSAR techniques. Based on freely available satellite acquisitions of
archives can cover the landslide history of the last 20 years in built-up areas.
Descending and ascending ERS-1 , ERS-2, Envisat and Sentinel-1 acquisitions have
been analyzed from 1992-2010, 2002-2010 and 2014-2016, using the SarScape 5.3.
module of Envi 5.3.2. PS and SBAS analyzes on satellite images (stacking
interferometry) have been done to calculate vertical and line of sight (LOS) horizontal
displacements of the area. Results were validated using the two different stacking
techniques on the same dataset.
Mass movement detection in the above-mentioned time-spans were done with success
on the inhabited parts of the high bluffs environment, where backscatters were more
stable than on vegetated areas. Methods were useful to obtain slow- or extremely-slow
moving landslides and to control the effect of stabilization efforts. It was demonstrated
that Sentinel-1 acquisitions are powerful basis of landslide monitoring and mapping.

This study is about the use of Enhanced Small Baseline Subset (ESBAS) technique for monitoring carbonate karst induced ground subsidence around the LAR valley in the northeast of Tehran, Iran. The recently developed SBAS technique relies   on   an   appropriate combination of SAR images with small temporal and spatial baseline. The SBAS technique was extensively modified to be able to deal with various aspects of de-correlation and atmospheric problems affecting InSAR observations.  In the last decade, the technique has been becoming increasingly an operational tool for time series deformation monitoring. During the ESBAS processing, several aspects within the chain of the SBAS processing have been modified. These modifications include filtering prior to   phase   unwrapping,  topographic   correction   within   3-dimensional  phase unwrapping,   reducing the atmospheric noise either by GPS data or based on the band-pass decomposition of both topography and interferometric phase, and removing the ramp caused by ionosphere turbulence and/or  orbit errors to better estimate crustal deformation, especially in the mountainous regions. We evaluated the effectiveness of our ESBAS method in monitoring of a significant sinkhole-related subsidence around LAR dam regions. The dam has been constructed to supply drinkable water for Tehran, capital city of Iran. In this regard, ten sinkholes were observed in the area and within few kilometers away from the dam location on the upstream side. The Lar catchment, 33 percent is occupied by exposed karstified carbonate rocks in which sinkholes. As the main conclusion, here, we showed that the use of enhanced processing makes the long-term monitoring of subsidence more efficent as compared to the SBAS processing.

 

In order to determine the condition of a high-speed railway bridge, various monitoring methods and a complex set of analyzes are necessary. Among them, the performance of movable bearings is one of the key aspects to be analysed. For this, it is necessary to carry out a long-term monitoring and evaluation of the bearing state, assess the bearing performance degradation, repair or replace the damaged bearings timely. As for the bridge displacement monitoring, two methods are commonly used. One is in-situ measurement done by workers periodically, which is strong subjective, and has no real-time capability. The other is the installation of displacement sensors to monitoring the longitudinal displacement at each pier, which can get real-time displacement. Both the methods are based on limited sets of sensors mounted near the movable bearings, collect point-like deformation information and have the disadvantage of providing a non complete displacement information.

In this paper, an Persistent Scatterer Interferometry (PSI) approach is used to monitor the displacement of the Nanjing Dashengguan Yangtze River High-speed Railway Bridge. 29 European Space Agency Sentinel-1A images, acquired from 25 April 2015 to 5 August 2016, were used in the PSI analysis. 1828 measurement points were selected on the bridge. The results show that there is a maximum longitudinal displacement of 150 mm, on each side of the bridge. The measured displacements show a strong correlation with the enviromental temperature at the times of acquisition of the used images, indicating that they are due to thermal expansion of the bridge. At each piers, longitudinal displacement models calculated with PSI and in-situ measurement were compared. Their good agreements demonstrate the capability of PSI technique for monitoring the performance of movable bearings. Morevover, the high density PSI measurement points has a potential to perform the health monitoring for the whole bridge.

Global Navigation Satellite System (GNSS) and Satellite Synthetic Aperture Radar (SAR) interferometry are two of the most important systems used to monitor the movements of the Earth’s ground surface from space. Therefore, the creation of a unique surface motion map that merges both GNSS and SAR techniques is desirable.

GNSS provides results referenced to the geocentric International Terrestrial Reference System 1989 (ITRS89), whereas satellite SAR differential interferometry techniques provide results referenced to the component identified by the Line of Sight (LOS) between a satellite and ground points. Since interferometric SAR data have no absolute reference datum, geodesy plays an important role in the datum alignment of SAR products, and the determination of precise velocity fields from GNSS permanent stations is essential to accurately correct data.

Methods of validating and correcting datasets are presented and applied to create a fine-scale surface velocity map of the central-eastern Po Plain.

A C-band radar placed into a geostationary orbit will combine continuous view capabilities, super-continental access and all-day-all-weather interferometric imaging, leading to a novel and unique observation system with quasi-continuous revisit.

The major scientific drivers of such mission will be presented starting from user gaps, analyzing requirements, and discussing potentials and scientific readiness.

 

In particular, applications will focus on:

Water vapor maps generation at high temporal and spatial resolution to support meteorological applications, particularly short term Numerical Weather Prediction and Nowcasting

Mapping of snow mass on land surfaces and snow accumulation on glaciers, monitoring the diurnal variation of snow melt /refreeze and 3D surface motions of glaciers

Monitoring slope instability, ground deformations, and underground fluid storage, and sinkholes, where short latency and fast delivery time are paramount

Observing Infrastructure instability due to excavations, mining, oil extraction, etc.

Monitoring fissuring and dyke injection, co-eruptive deformation and lava flow motion on volcanoes

Monitoring rapid sequences of deformation events caused by earthquakes and seismic sequences

For these applications, examples taken from LEO and Ground Based SAR will be reviewed and a quantitative assessment of performance will be provided based on a mission whose cost fits in an Earth Explorer budget.

In interferometry technique, phases have been modulated between 0-2π. Finding the number of integer phases missed when they were wrapped is the main goal of unwrapping algorithms. Although the density of points in conventional interferometry is high, this is not effective in some cases such as large temporal baselines or noisy interferograms. Due to existing noisy pixels, not only it does not improve results, but also it leads to some unwrapping errors during interferogram unwrapping.  In PS technique, because of the sparse PS pixels, scientists are confronted with a problem to unwrap phases and due to the irregular data separation conventional methods are sterile. As unwrapping is a key step to estimate deformation from an interferogram either in conventional or PS techniques, several methods have been proposed to unwrap PS data. Unwrapping techniques can be divided in to path-independent and path-dependent in the case of unwrapping paths. Path-dependent technique can be divided into sequential path and region-growing. A region-growing method which is a path-dependent technique has been used to unwrap PS data. In this paper an idea of extended Kalman filter has been generalized on PS data. This algorithm is applied to consider the nonlinearity of PS unwrapping problem as well as conventional unwrapping problem. A pulse-pair method enhanced with singular value decomposition (SVD) has been used to estimate spectral shift from interferometric power spectral density in 7*7 local windows. Furthermore, using Delaunay triangulation to reduce sparse data on a regular grid makes strict polygon boundary, hence moving from one polygon to other neighbor polygons is too rough; however, edges can be passed gradually and smoothly by means of Kalman filter. Moreover, a cost-map is defined to prioritize PSs which should be unwrapped. The cost-map consists of phase derivative variance (PDV) which is enhanced with geometric properties and neighbor polygon distributions. This algorithm has been implemented on simulated PS data. To form a sparse dataset, 1.06% of points from regular grid are selected randomly. The results of this algorithm and real unwrapped phases were completely identical. Additionally, this algorithm is implemented on real PS dataset. Real PS data located on the southwestern part of Tehran have been extracted from 22 ENVISAT ASAR images. The proposed algorithm does not require any prior knowledge of deformation model which makes it distinguished among other PS techniques. In order to evaluate the results obtained from the application of the proposed model to the real data, precise leveling measurements are applied. The comparison results show the significant performance of the proposed method.

3D SAR Tomography (TomoSAR) [1-4] and 4D SAR Differential Tomography (Diff-TomoSAR) [8-12] exploit multi-baseline SAR data stacks which represent an important innovation of SAR Interferometry, to sense complex scenarios with multiple scatterers mapped into the same SAR cell. In addition to 3-D shape reconstruction and deformation retrievals in complex urban/infrastructure areas [2,4], and recent cryosphere ice investigations [5], there have been emerging tomographic remote sensing applications regarding forest scenarios [3,6,7], e.g. tree height and biomass estimation and sub-canopy topographic mapping. However, these scenes are characterized by temporal decorrelation of scatterers, orbital, tropospheric and ionospheric phase distortion and an open issue regarding possible height blurring and accuracy losses for TomoSAR techniques particularly in densely vegetated mountainous rural areas. Thus, it is important to extend characterisations of temporal decorrelation, orbital, tropospheric and ionospheric phase distortion. We will study on 3D point cloud reconstruction (especially in vertical layers) over densely vegetated mountainous rural areas using 3-D SAR imaging (SAR tomography) and data stacks of X-band COSMO-SkyMed Spotlight and L band ALOS-1 PALSAR data stacks over Dujiangyan Dam, Sichuan, China. Atmospheric and ionospheric correction method are discussed first to remove tropospheric and ionospheric delay. Then the TomoSAR method is described to obtain the number of scatterers inside each pixel, the scattering amplitude and phase of each scatterer and finally extract their 3D positions and motion parameters.

This work is partially supported by the CSC and UCL MAPS Dean prize through a PhD studentship at UCL-MSSL.

Keywords: 3D; point cloud; tomography; densely vegetated mountainous rural areas

[1] A. Reigber, A. Moreira, “First Demonstration of Airborne SAR Tomography using Multibaseline L-band Data,” IEEE TGARS, 38(5), pp.2142-2152, 2000.

[2] G. Fornaro, F. Serafino, F. Soldovieri, “Three Dimensional Focusing With Multipass SAR Data,” IEEE TGARS, 41(3), pp. 507-517, 2003.

[3] M. Nannini, R. Scheiber, R. Horn, “Imaging of Targets Beneath Foliage with SAR Tomography,” EUSAR’2008.

[4] F. Lombardini, F. Cai, D. Pasculli, “Spaceborne 3-D SAR Tomography for Analyzing Garbled Urban Scenarios: Single-look Superresolution Advances and Experiments," IEEE JSTARS, 6(2), pp.960-968, 2013.

[5] L. Ferro-Famil, C. Leconte, F. Boutet, X. Phan, M. Gay, Y. Durand, “PoSAR: A VHR Tomographic GB-SAR System Application to Snow Cover 3-D Imaging at X and Ku Bands,” EuRAD’12.

[6] F. Lombardini, F. Cai, “3D Tomographic and Differential Tomographic Response to Partially Coherent Scenes,” IGARSS’08.

[7] M. Pardini, K. Papathanassiou, “Robust Estimation of the Vertical Structure of Forest with Coherence Tomography,” ESA PolInSAR ’11 Workshop.

[8] F. Lombardini, F. Cai, “Evolutions of Diff-Tomo for Sensing Subcanopy Deformations and Height-varying Temporal Coherence,” ESA Fringe’11 Workshop.

[9] F. Lombardini, “Differential Tomography: A New Framework for SAR Interferometry”, IEEE TGARS, 43(1), pp.37-44, 2005.

[10] Xiang, Zhu Xiao, and Richard Bamler. "Compressive sensing for high resolution differential SAR tomography-the SL1MMER algorithm." In Geoscience and Remote Sensing Symposium (IGARSS), 2010 IEEE International, pp. 17-20. IEEE, 2010.

[11] F. Lombardini, M. Pardini, “Superresolution Differential Tomography: Experiments on Identification of Multiple Scatterers in Spaceborne SAR Data,” IEEE TGARS, 50(4), pp.1117-1129, 2012.

[12] F. Lombardini, F. Viviani, F. Cai, F. Dini, “Forest Temporal Decorrelation: 3D Analyses and Processing in the Diff-Tomo Framework,” IGARSS’13.

A new fleet of satellites developed by ESA for the European Commission, the Sentinels, has begun in 2014 with Sentinel-1a systematic acquisition of Earth Observation data over the globe. Acquisitions will continue for the next decades, with follow-on missions of existing satellites and new satellites with different observation capabilities.

Technological and knowledge issues are partially preventing user's uptake of such large volume of data.

We intend to present a service aiming to overcome such issues.

The Indian Ocean InSAR Observatory (OI²) is a component of the National Services of Volcanological Observations, one of the 20 National Services of the French National Institut of Earth and Space Sciences of CNRS. One of the main goals of OI² is the regular production and diffusion of ground displacement measurements, related to volcanic activity at Piton de la Fournaise, La Réunion Island. The displacement measurements are obtained from radar interferometric remote sensing data. They are exploited both in a near-real time operational framework, as a component of the geophysical dataset used by the Piton de la Fournaise Volcano Observatory scientists to monitor the volcano activity, and for more fundamental researchs interested either in methodological developments or in improving our understanding of the way the volcano works. On this poster, we first give a short description of the Piton de la Fournaise geological context, then we describe the OI² missions, operations and database. Finally we present some examples of scientific exploitation of the OI² data, with results related to the recent activity at Piton de la Fournaise (between april 2007 and september 2016).

As is well known, a major earthquake of Mw 7.8 occurred in Nepal on April 25, 2015. In three cities (Kathmandu, Laritpur and Bhaktapur) in the Kathmandu Valley, catastrophic damage was reported such as several temples registered as world heritage collapsed. The authors participate in the project of “Hydro-microbiological Approach for Water Security in Kathmandu Valley, Nepal (principal investigator: Prof. Kazama Futaba, adoption fiscal year: FY2013)” adopted into SATREPS (*1), and are working on satellite image analysis in order to contribute to solving the problem of securing quantitative water safety of Kathmandu. This project aims to evaluate the safety of water supply and to construct a low-cost and sustainable water treatment system by a combined research approach of water volume, water quality and microorganisms. The stable supply of water and its impact assessment to be addressed in this project are indispensable for all recovery activities related to reconstruction of afflicted areas, urban rebuilding, restoration of world heritage and others. Currently, collaborations in each field with the cooperation of counterparts Tribhuvan University and other Nepalese organizations have been promoted.

In this research, InSAR time series analysis using ALOS/PALSAR data and conventional DInSAR analysis using Sentinel-1A data were carried out. The subsidence information obtained by this study can be effectively used for securing quantitative water resources and planning for urban reconstruction.

For the analysis by using PALSAR data, we processed 16 images acquired from January 30, 2007 to December 26, 2010. 59 interferometric combinations which a vertical component of the spatial baseline was less than 1 km were selected, and time series analysis by Deguchi et al. (2009) was applied. The processing software was "dpro" developed by the first author. As a result of the time series analysis, 47 cm, 31 cm and 8 cm displacement away from the satellite for 4 years were measured in Kathmandu, Laritpur and Bhaktapur respectively (Figure 1). Since these displacement amount ​​were in line of sight direction, 60 cm (15 cm per year), 40 cm (10 cm per year) and 10 cm (2.5 cm per year) of land subsidence occurred for in Kathmandu, Laritpur and Bhaktapur under the condition of assuming that the surface deformation was dominated only by the vertical component.

The population of the Kathmandu Valley, which was 680,000 at the time of 1991, increased to 1.08 million in 2001 and 1.74 million in 2011. The population growth rate is actually 2.5 times in 20 years. By superposition of the population distribution map of the Kathmandu basin area by LANDSCAN (as of 2010) and the land subsidence map produced by InSAR time series analysis, it was recognized that land subsidence anomalies corresponded to the densely populated area (more than 10,000 people per km2). When the land use map (as of 1995) by the International Center for Integrated Mountain Development (ICIMOD) was overlapped with the land subsidence map, the large subsidence had clearly occurred not only in urban areas but also in suburban development areas. In comparison with high resolution images on 22 March 2007 and 6 February 2012 posted on Google Earth, at least five tower buildings or commercial facilities were newly established in the area where the maximum subsidence amount was detected. It was inferred that modernization and urban development in recent years had influenced a large amount of subsidence. As described above, a large amount of groundwater may be collected in order to cope with the rapid increase of the population as well as the urban development occurring in the central part of the Kathmandu Valley.

Unfortunately, ALOS/PALSAR already stopped operation in April 2011. Accordingly, DInSAR analysis using data of Sentinel-1A launched by the European Space Agency (ESA) in April 2014 was conducted experimentally. As a result of the analysis, a phase anomalies indicating the tendency of ground subsidence was also detected after the earthquake. In the future, we will consider the application of InSAR time series analysis by using Sentinel-1A data and the implementation of leveling survey for verification in the field.

 (*1) SATREPS (Science and Technology Research Partnership for Sustainable Development) is a Japanese government program that promotes international joint research. The program is structured as collaboration between the Japan Science and Technology Agency (JST), which provides competitive research funds for science and technology projects, and the Japan International Cooperation Agency (JICA), which provides development assistance (ODA). Based on the needs of developing countries, the program aims to address global issues like as energy/environment issues, disaster risk reduction, infectious disease control, and food security, and lead to research outcomes of practical benefit to both local and global society.

As part of the Ring of Fire, Shiveluch volcano is one of the largest and most active volcanoes on Kamchatka Peninsula. During the Holocene, only the southern part of the Shiveluch massif was active. Since the last Plinian eruption in 1964, the activity of Shiveluch has been characterized by periods of dome growth and explosive eruptions. The recent active phase began in 1999 and continues even today. Due to the special conditions at active volcanoes, such as smoke development, danger of explosions or lava flows, as well as poor weather conditions and inaccessible areas, it is difficult to observe the interaction between dome growth, dome destruction, and explosive eruptions in regular intervals. Consequently, a reconstruction of the eruption processes is barely possible, though important for a better understanding of the eruption mechanism as well as for hazard forecast and risk assessment.

A new approach is provided by the bistatic radar data acquired by the TanDEM-X satellite mission. This mission is based on two nearly identical satellites, TerraSAR-X and TanDEM-X, flying in a close helix formation. On one hand, the radar signals penetrate clouds and partially vegetation and snow considering the average wavelength of about 3.1 cm. On the other hand, in comparison with conventional InSAR methods, the bistatic radar mode has the advantage that there are no difficulties due to temporal decorrelation. By interferometric evaluation of the simultaneously recorded SAR images, it is possible to calculate high-resolution digital elevation models (DEMs) of Shiveluch volcano and its surroundings. Furthermore, the short recurrence interval of 11 days allows to generate time series of DEMs, from which finally volumetric changes of the dome and of deposits can be derived.

Here, this method is used at Shiveluch volcano based on data acquired between June 2011 and September 2014. Although Shiveluch has a fissured topography with steep slopes, DEMs with a resolution of about 6 m can be calculated and the changes caused by volcanic activity can successfully be derived and quantified. The presentation focuses on the required DEM corrections as well as the error analysis of the DEMs and the volumetric changes.

Monitoring ground surface displacement of volcanoes over a long period of time helps improve understanding of the volcano structure, dynamics and mechanisms. On the Piton de la Fournaise volcano, we analyze the spatio-temporal behavior of the long-term displacement field affecting the volcanic edifice between 2007 and 2014 using continuous GNSS times series and a large X and C band InSAR data set ( ~ 210 images acquired by ENVISAT, TerraSAR-X and COSMO-Skymed satellites). We use a classical SBAS processing approach to process this large SAR data set but using an original approach based on principal component analysis to mitigate the long-wavelength tropospheric artifact in the interferograms. The resulting InSAR time displacement maps show that several ground displacement patterns overlapping both in space and time affected the major part of  the volcanic edifice. Recent lava flows are mainly affected by subsidence decreasing in time with nearly no long term eastward motion. We show that the amplitude of the post-lava flow emplacement subsidence affecting a give lava flow is generally proportional to the the thickness and is also in relation with the age of lava flow. From these relations we determine an empirical law in order to estimate the contribution of the lava flow subsidence in the surface displacement field of Piton de la Fournaise. During the time interval of this study, a slow-rate summit deflation is observed only affected by strong and sudden motions during different eruptive events. More importantly, a widespread sector (~ 20 km²) affected by time-dependent downward and eastward motion is observed on the Eastern Flank of Piton de la Fournaise. The combined analysis of InSAR and GNSS time series allow us to confirm that this large mobile sector on the Eastern Flank underwent deformation for at least 7 years following the end of the historical March-April 2007 eruption, including a stage of transient strain from the end of eruption to February 2011, followed by a stage of steady-state strain until late 2014.

Maritime structures such as breakwaters have a large economic importance as they protect the harbours, the ships and their contents from sea waves. Rubble-mound breakwaters are one of the most used types of these structures and they are composed by rocks forming a trapezoidal cross-section covered with rock or concrete blocks whose goal is to dissipate sea wave energy. The breakwaters are often exposed to severe meteorological conditions, especially during storms, which may move the blocks from their original position, reducing the protection capability of the structure. An early detection of those changes is of the utmost importance as it may minimize costs both for the breakwater reconstruction and for the consequences of not having it working properly. Structural health monitoring of breakwaters is often performed through visual inspections. Due to the shape of the structure, the size and dimensions of the protective blocks, and the fact that it is surrounded by the sea, in situ measurements are difficult to perform. Photographic surveys using unmanned aerial vehicles are sometimes hard to accomplish due to the strong winds that are observed at the coast. Multitemporal InSAR techniques can overcome the hindrances of the mentioned methods and provide a useful way to identify the damages.

In this study, damage assessment is performed for a breakwater located at Ericeira town, Portugal, using multitemporal InSAR techniques. This breakwater was built in the 1970s and it is exposed to northwestern Atlantic Ocean waves which are very energetic and led to severe damages several times. From 2008 to 2010 the structure was reconstructed and expanded in order to increase its protection capability. A dataset of 70 SAR images acquired between 2011 and 2015 by the sensor on board COSMO-SkyMed 4 was used to evaluate changes occurred at the breakwater during this time interval. Displacement and amplitude time-series as well as coherence matrixes are analysed for damage assessment using the SARPROZ© software. Time-series analysis shows a good agreement between changes at displacement and amplitude data and the occurrence of storms at the Portuguese coast. Spatial patterns are found that enable the identification of the most fragile areas of the breakwater which may need maintenance interventions in the near future.

Tangible cultural heritage includes various categories of monuments and sites, from cultural landscapes and sacred sites to archaeological complexes, individual architectural or artistic monuments and historic urban centres. Such places are continuously impacted and weathered by several internal and external factors, both natural and human-induced, with rapid and/or slow onset, including natural hazards, such as earthquakes or extreme meteorological events, cumulative processes as well as the effects of humans. The list of List of World Heritage in Danger encompasses sites threaten by armed conflicts, however, a comprehensive picture of endangered sites is not available as yet. New space technology based on radar interferometry (InSAR) is now capable to monitor, since 1992 and with mm precision, surface deformation for reflective targets named persistent scatterers, which consistently return stable signals to the radar sensor. The project PROTHEGO: PROTection of European Cultural HEritage from GeO-hazards (www.prothego.eu) will make an innovative contribution towards the analysis of geohazards in areas of cultural heritage in Europe. The project is led by the Italian Institute for Environmental Protection and Research (ISPRA) in collaboration with the Natural Environment Research Council (NERC), Cyprus University of Technology (CUT), University of Milano-Bicocca (UNIMIB) and the Geological Survey of Spain (IGME), and funded in the framework of the Joint Programming Initiative on Cultural Heritage and Global Change (JPICH) Heritage Plus (2015-2018). PROTHEGO will apply novel InSAR techniques to monitor monuments and sites that are potentially unstable due to landslides, sinkholes, settlement, subsidence, active tectonics as well as structural deformation, all of which could be exacerbated by climate change and human interaction. To magnify the impact of the project, the approach will be implemented in more than 400 sites on the UNESCO World Heritage List in geographical Europe. After the remote sensing investigation, detailed geological interpretation, hazard analysis, local-scale monitoring, advanced modelling and field surveying for the most critical sites will be carried out to discover cause and extent of the observed motions. The project will enhance cultural heritage management at national level, reinforcing institutional support and governance through knowledge and innovation, identifying, assessing and monitoring risks, strengthening disaster preparedness at heritage properties in the future.

The crisis of Land subsidence is one of the most widespread hazards in Great cities. Subsiding in urban and suburban areas is mostly associated with imbalance of ground-water levels and dehydration of alluvial aquifer systems. This phenomenon threats to the integrity of nearby infrastructures.

  Subsidence in Tehran plain during time span of 2003-2013 is mainly covered by cultivated lands between shayriyar-Eslamshahr (center latitude: 35.63°N, center longitude: 51.13° E) and some residential areas in South-West of Tehran with maximum rate > 200 mm. The drop in water level reaches to 11.65 meters during 28 years ago (1984-2012). In 2003, the number of wells permitted in the Tehran plain was 26,070; by 2012, this number had risen to 32,518.

The repeated X band Cosmo_SkyMed (CSK) data over Great Tehran have been acquired by the  Italian Space Agency and makes it possible to apply multi image processing with high resolution for monitoring Land subsidence. The CSK data-set has been processed with the GAMMA software.24 scenes between 2014-2016 along As and Des orbit were used .The CSK data analysis indicates that the 17, 18, 19 districts of Tehran have remarkable subsidence especially 19 district as well as part of No7 subway line under construction near Railway station. Maximum accumulated subsidence for first and second region over a period of 3 years (2014–2016) was about 10.5-3 cm and 3-1.5 cm.  Satellite data analysis of 20150324–20150511 indicates that some areas in city have potential of ~1 cm subsidence in a 48 days period. In Iran, the summer drawdown season typically occurs from May to October and the rest of the year is the recovery season. The study confirmed that land subsidence caused by groundwater pumping is a serious threat to Tehran and surrounding.

Key words: Subsidence, Aquifer systems, Tehran, Cosmo_SkyMed

In 2011, a large number of building damage reports showed up in the city of Böblingen (near Stuttgart, Germany). A few years earlier, from 2006 to 2008, geothermal drillings were carried out in some of the affected areas. The suspicion that these drillings were connected to the building damages soon rose. Incorrect drilling procedures may have induced water leakage into anhydrite layers beneath the city and provoked its transformation to gypsum. Later on, the swelling of the gypsum may appear as an uplifting area on the earth’s surface.

Soon after the first damage reports, local authorities ordered investigations to get a more detailed insight into the incidents. With Airborne Laserscanning, precise Levelling campaigns and a first Persistent Scattterer InSAR (PS-InSAR) analysis two major uplift areas could be identified. The measured displacement rates were of about 6 mm/month and an accumulated displacement of 37 – 45 cm since the beginning of the uplift was reported. These investigations concentrated on recent displacements but gave little information on the temporal and spatial beginning of the uplift. As this is seen as valuable information, a corresponding project was launched and funded by the Ministry of Environment, Climate Protection and Energy Sector in Baden-Württemberg. In this ongoing project, we carry out a PS-InSAR analysis with Envisat data from 2003 to 2010 (one ascending and two descending tracks) in order to pinpoint temporally and spatially the beginning of the uplift phenomenon in Böblingen.

After processing the data with StaMPS (Stanford Method for Persistent Scatterers) and getting the PS-InSAR results, the main focus lies on one of the descending tracks and a detailed time series analysis for selected points in one of the main uplift areas. To gain information of the temporal beginning of the uplift, the detection of change points, i.e. trend changes in the time series, is necessary. The matlab package “Shape Language Modeling (SLM)” is applied to PS points with high uplift rates in order to find such change points. This tool allows fitting linear or cubic subfunctions to a data set. A smoothness criterion guarantees that these subfunctions are neatly joined together at the break of slopes as to get a continuous overall function. Additionally, a least-squares-approach is introduced to find the best fit to the data.  The position of the change points is determined by analysing the third derivation of the overall function and its changes. The user has to specify the expected number of change points as well as the expected displacement behaviour of the PS. In our case, we assumed one change point in the time series and anticipated that the PS displacements follow a linear behaviour. First results show that the uplift started in the beginning of 2007, shortly after the first geothermal drillings had been carried out. The determination of the significance of the resulting change points is done with bootstrapping methods and still has to be investigated in detail.

Multi-temporal InSAR (MTI) applications pose challenges related to the availability of coherent scattering from the ground surface, the complexity of the ground deformations, presence of atmospheric artifacts, and visibility problems related to the ground elevation. Nowadays, several satellite missions are available, providing interferometric SAR data at different wavelengths, spatial resolutions, and revisit times.

High-resolution X-Band SAR sensors, such as the COSMO-SkyMed constellation, acquire data with spatial resolution reaching metric values, and revisit time up to a few days, leading to an increase in the density of usable targets, as well as to an improved detection of non linear movements. Medium resolution C-band SAR data have been thoroughly exploited in the last two decades, thanks to the ERS-1/2 and ENVISAT-ASAR missions, and Radarsat-1/2.  A new interesting opportunity is provided by the Sentinel-1 mission, which has a spatial resolution comparable to previous ESA C-band missions, and a revisit time reduced to 12 and 6 days, by considering, respectively, one or two satellites. It is envisioned that, by offering regular, global scale coverage, improved temporal resolution and freely available imagery, Sentinel-1 will guarantee an increasing use of MTI for ground displacement investigations.

The present work discusses opportunities of MTI applications to ground instability monitoring by assessing the performance of the different available satellite missions, according to acquisition parameters such as wavelength, spatial resolution, revisit time and orbital tube size. This performance analysis allows to foresee the quality of displacement maps estimated through MTI according to mission characteristics, and thus to support SAR data selection.  In particular, a comparative analysis is carried out, aimed at addressing specific advantages of different satellite missions in L-, C- and X-band. For instance, high resolution data increase the density of coherent targets, thus improving the monitoring of local scale events. Short (X-band) wavelengths improve the sensitivity to displacements.  Short revisit times allow collecting large data stacks in short times, and improve the temporal sampling, thus increasing the chances to catch pre-failure signals (high-rate, nonlinear signals). The precision of the displacement rate detection depends on the number of images and on the phase noise, while the precision of the residual height error estimation depends also on the orbital tube size. Sentinel-1 will provide data for the next years with short revisit time, and it is thus likely to provide reliable displacement estimations at large scale, and in quite limited observation time spans. However, due to its narrow orbital tube size, it has a limited height precision, which leads to poor geo-location quality.

An example of multi-sensor ground instability investigation is also presented concerning the site of Marina di Lesina, in Southern Italy, where several SAR datasets are available acquired from ERS, ENVISAT, Radarsat-2, COSMO-SkyMed and Sentinel-1, covering more than 20 years with varying ground resolutions, frequency bands and repeat times. The site is affected by sinkholes and uplifting caused by the interaction between the water coming from an artificial canal and the underground soil where gypsum with residual anhydride is present.  The data at C-band and medium resolution from ERS and ENVISAT are able to catch the large scale uplift pattern, since the available observation time span is suitable to provide the required velocity precision.  Radarsat-2 data improve the spatial density of detected targets, while, as foreseen by the model, Sentinel-1 improves the C-band performance, by providing, in a limited time span, precise estimation of the displacement rates. Finally, as expected, high resolution data from COSMO-SkyMed lead to a considerable increase of the PS spatial density, which allows to improve the delineation of the spatial deformation pattern. High resolution / short revisit time data are also very promising for detecting small precursory terrain movements related to the sinkholes.

ACKNOWLEDGMENTS

Work supported by the project “APULIA SPACE” (PON03PE_00067_6), PON Ricerca e competitività 2007-2013.  

The occurrence of geological events such as landslides, subsidence phenomena or structural failures is one of the main causes of damage in linear infrastructures such as roads, bridges, railways and retaining walls. The deformation of those man-made facilities provides important socioeconomic and human losses. The frequent and accurate monitoring of infrastructures and their interaction with existing urban infrastructures plays a key role in risk prevention and mitigation activities.

For this reason, a considerable interest towards innovative approaches useful to provide information on temporal and spatial trends of surface deformations has grown among the scientific community and land management institutions.

The present work suggests a methodology, which, using Differential Interferometry Synthetic Aperture Radar techniques, obtained from VHR images, allows to identify, the areas of linear infrastructures which could be affected by active movements. To this purpose, working in open source software, a semiautomatic GIS-environment procedure has been developed. A buffer zone has been generated along the linear progression of the infrastructures by mean of the implementation of regular grid: the Permanent Scatterers (PS) data within the grid have been then processed through statistical analysis in order to obtain a zoning map useful to identify sections characterized by higher damage susceptibility, where detailed conventional in situ measurements are required for further analysis.

The proposed approach has been successfully applied to monitor some roads in Campania Region (Southern Italy): the interferometric stack of the available COSMO-SkyMed SAR images, in time-span 2011-2014, thanks to its accuracy, high spatial resolution, non-invasiveness and long-term temporal coverage, integrated with topographic data and geomorphological field surveys, allows to identify the most critical areas and preliminary cause-effect correspondence, aimed to predict future conditions of damage and local failures of infrastructures.

InSAR analysis is effective tool to detect unknown displacements. And recent advances of time series analysis, there can be determined the accurate deformation rate with millimeter precision like GNSS continuous observation. Additional this, ESA's Sentinel-1 constellation allows us time series analysis shorter period than the previous SAR satellite missions at all over the world. For example, ALOS spent about 5 years to make 30 scenes InSAR pair in Japan. But Sentinel-1 could take only about 1.5 years to observe 30 scenes InSAR pair in Japan. Short period to InSAR time series analysis means more chance to warn landslides, sinkholes, and other disaster associated with subsidence. So it become more important to use Sentinel-1 constellation.

But there is two problem in Sentinel-1. One is the resolution. Sentinel-1 is acquired majorly by the TOPS mode, so the resolution is about 5 meter in range and 20 meter in azimuth. This cause the difficulty to detect like what building is being subsidence. And the second is wavelength. Sentinel-1 is C-band satellite, so it is difficult to detect deformation in rural area or vegetation area.

 In contrast ALOS-2 has more resolution and longer wavelength but less observation than Sentinel-1. Theoretically, there is no deference of urban result, but rural because of coherency of vegetation. But actually there is one deference selecting persistent scatterer of time series analysis among L-band and C-band even if urban area. The effect of this is less considered especially Sentinel-1 TOPS mode and ALOS-2 stripmap mode, even though these two operation are newest in the world.

In Tokyo, Japan, there is a tunnel under construction now. The underpass of this tunnel is constructing 40 meter under from surface or 10 meter under from upper supporting grounds of buildings because there is not required site acquisitions under Japanese law.

To monitor the subsidence by the construction under urban area, it is useful to use InSAR time series. Because it is difficult to measure over the all buildings like urban area. So the InSAR time series analysis is the only way to know accurate deformations in urban area similar to rural area, there are little measuring instruments.

 In this study, we analyzed the deformation rate using Sentinel-1 and ALOS-2 by GAMMA/IPTA module’s PS analysis. We chose the reference point as nearly point between two satellites. Because if the selections of reference between two are far from each other, there is possibility the deformation near the reference point to influence the results. And there are SAR looking angle effects, but we assumed the deformation is subsidence or uplift only. There is a way to divide uplift and NS-SW moving to use 2 satellite passes like ascending and descending pass. But we do not apply this, because the scatterer selected by processing may be defer from 2 passes, so it become difficult to divide what actually moving between building and ground.

And we use 28 scenes of Sentinel-1 from Nov. 2014 to Aug. 2016 and 7 scenes of ALOS-2 from Dec. 2014 to Sept. 2016. If the Sentinel-1 scene contains atmospheric effects, we omitted the scene from analysis dataset. Because we did not apply the atmospheric correction using numerical weather model like EMCWF here. Finally we used 23 scenes of Sentinel-1. ALOS-2 used all scene because of the limitation of the number.

In the result, we found 2 major subsidence. One is already reported using by ERS/Envisat result, and another subsidence is not reported before. This shows the first one is sinking continuously from 2001 to 2016, another one starts recently. And that this result can detect both ALOS-2 and Sentinel-1 means this is not analysis error. Actually this new subsidence area started the construction of tunnel of Tokyo-gaikan Expressway. So this means about only 1.5 years observation can detect the subsidence associated with tunneling using Sentinel-1 with millimeter precision.

Mainly prone landslide areas in Indonesia are located on densely vegetation cover environment that makes conventional InSAR method hardly detect the real exist moving of surface displacement. Consider to the drawback, multi-temporal stack of interferograms are needed to assess the amplitude and phase stability through a relevant configuration of temporal, perpendicular baselines, and different doppler centroid which ensuring there is no isolated clusters of this image-pairs network. A dataset of ERS1/2 SAR Images from April 1996 to Juni 1999 has been proceed in Ciloto, Indonesia where Indonesia National Agency for Disaster Management has classified this area as one of the highest landslide risk. The interferograms are formed on single-look images to filter out decorellated phase pixels at the highest possible resolution. We identify the candidate pixels with applying amplitude dispersion index (D_A) and slowly-decorrelating filtered phase (SDFP) algorithm developed by Hooper, 2008. These candidate pixels assumely include the phase change due to ground movement in the line-of-sight (LOS) and external factors that can not be avoided such as atmospheric delay, orbit inaccuracy, look angle error, and noise. On the western-north of our case study, mountainous area, we discover strong tropospheric artifact that could misslead to deformation signal. However if these external factors have highly spatially correlated terms then we could reduce these errors and able to better assess the phase change due to surface movement. The result shows slow-moving displacement has been detected with displacement rates up to ~ 30 mm/year on the top body of landslide risk area where mostly the land use is rural residential environment. The velocity consider non-linear because the geology condition and rainfall intensity strongly influences to the speed of movement which in general occured from november – february, wet season. We also figure out the accuracy estimation using the stable reference area. Moreover, the paper discussed possibility and the limitation of SBAS InSAR to monitor landslide hazard in which we find it is difficult generating pixel candidates on the main body of landslide risk due to very low coherence signal surrounded by agriculture land use. Regarding to this problem, we could not able to compare quantitavely the displacement generated by SAR interferometry to the previous terrestrial measurement.

Acknowledgements

ERS data were provided by the European Space Agency with the project scheme P31690. This research was supported by Deutscher Akademischer Austauschdienst (DAAD).

References

[1] Hooper AJ (2008) A multi-temporal InSAR method incorporating both persistent scatterer and small baseline approaches, Geophysical Research Letters, 35, . doi: 10.1029/2008GL034654

Phase imaging systems play a vital role in many present day technologies, namely in the field of surveillance, remote sensing, medical diagnostic, weather forecasting and photography. Often, in such systems, a physical quantity of interest is coded in an image of phase using a suitable coherent imaging techniques (e.g., InSAR, InSAS). Since the phase is closely linked with the wave propagation phenomenon, the measured signals depend only on the principal (wrapped) values of the original phase (absolute phase), which we term as interferometric phase (InPhase), usually defined in the interval [-∏, ∏). The interferometric phase is thus a sinusoidal and nonlinear function of the absolute phase, which renders absolute phase estimation a hard inverse problem. In addtion, the interferometric phase is usually corrupted by the noise introduced by the acquisition mechanism and electronic equipments, which further complicates the inverse problem which is the inference of the absolute phase from interferomertic measurements. This problem is often tackled in a two-step approach. In the first step, denoising of the noisy wrapped phase is taken care and in the second step, the denoised phase image is unwrapped. InPhase image denoising should be addressed with special care since the wrapping discontinuities should be preserved carefully for the second stage of unwrapping.

In this paper, we propose a novel approach to address the problem of interferometric phase denoising by modelling the complex domain phase using Mixture of Gaussian densities. The recent state-of-the-art techniques in image denoising are based on non-local self-similarity and sparsity [1], which may be exploited via nonlocal patch-based techniques. The fact that the phase images are natural images motivates the application of patch-based approaches in InPhase image denoising. Following the standard procedures in patch-based image restoration, the phase image is decomposed into small overlapping patches. Vectors corresponding to these patches are modelled using Mixture of Gaussian (MoG) densities in the complex domain. The parameters, i.e.,
the covariance matrix, mean and mixing coecients of the MoG are learned from complex domain patches of the training data. The learned MoG is used as a prior for estimating the interferometric phase images from the noisy images.

The main contribution of our work, which is inspired from the recent state-of-the-art image denoising techniques based on MoGs (see, e.g., [2]), can be summarized as follows: 1) an algorithm to learn the model; this is accomplished by formulating an Expectation Maximization (EM) algorithm for MoG densities in the complex domain; 2) computing the Minimum Mean Square Error (MMSE) estimates of the clean patches from the noisy ones using the learned model. The experiments conducted on simulated and real data of InSAR/InSAS shows results which are competitive with the state-of-the-art techniques [3], [4]. Also the entire process of InPhase estimation is illustrated by unwrapping the denoised interferogram using PUMA [5] algorithm.

One of the relevant contribution of our work is that it opens the door to the exploitation of  "learned priors"  from the specied classes of interferometric phase images, which can then be used in various phase inverse problems.

References

[1] J. Mairal, F. Bach, J. Ponce, G. Sapiro, and A. Zisserman, "Non-local sparse models for image restoration," in 2009 IEEE 12th International Conference on Computer Vision, Sept 2009, pp. 2272-2279.

[2] T. Afonso, M. Almeida, and M. Figueiredo, "Single-frame image denoising and inpainting using Gaussian mixtures," in Proceedings of the International Conference on Pattern Recognition Applications and Methods, 2015, vol. 2, pp. 283-288.

[3] H. Hongxing, J. Bioucas-Dias, and V. Katkovnik, "Interferometric phase image estimation via sparse coding in the complex domain," IEEE Transactions on Geoscience and Remote Sensing, vol. 53, no. 5, pp. 2587- 2602, 2015.

[4] Q.Kemao, "Two-dimensional windowed fourier transform for fringe pattern analysis: Principles, applications and implementations," Optics and Lasers in Engineering, vol. 45, no. 2, pp. 304 - 317, 2007.

[5] J. Bioucas-Dias and G. Valadao, "Phase unwrapping via graph cuts,"  IEEE Transactions on Image Processing, vol. 16, no. 3, pp. 698-709, 2007.

The greatly increased data availability associated with the successful operation of the first two Sentinel-1 satellites, combined with the maturity level of several SAR data processing techniques, have the potential of boosting research activities and commercial services based on satellite data, and represent a prerequisite for the development of cost-effective monitoring services. At the same time, new technological solutions are required to handle the increased data volumes and processing times.

 

An effective approach for some users and applications is to run computationally-intensive algorithms on supercomputers or distributed computing systems, which consist of a large number of physical machines, located in central processing facilities. On the other hand, several research institutions and SAR-data service providers currently rely on small to medium size in-house processing facilities (e.g. local clusters or just powerful workstations). This contribution addresses this latter scenario, and quantifies the performance of our cluster solution for two widely used multi-temporal SAR algorithms, namely the Small Baseline Subset (SBAS) and Persistent Scatterer (PS) methods for the measurement of slow surface deformations, and for a basic processing step, namely the generation of a geocoded interferometric coherence map, which is used for Coherent Change Detection (CCD) analysis.

 

Our cluster architecture consists of three main software components: one or more client applications (e.g. a program executing a SBAS, PS or CCD algorithm); a set of cluster nodes (one per cluster machine or per processing resource, e.g. CPU or GPU); one task orchestrator (i.e. a cluster entry-point managing available clients and nodes). Communications between components are TCP/IP-based. All clients and nodes connect to the orchestrator, which carries out high-level management tasks. Data-intensive transfers (e.g. to copy back to and from the cluster nodes), are carried out through dedicated communication channels between the clients and the nodes, to avoid orchestrator bottlenecks. From the user point of view, only some basic cluster configuration info must be provided at setup, e.g. the orchestrator and cluster node IP addresses, after which all software components run as an operating system service. The client and cluster machines can run indifferently on Linux or Windows operating systems (also mixed configurations are supported, e.g. a Linux or Windows client machine can work with a set of Linux and/or Windows cluster nodes).      

 

Concerning the SBAS algorithm, the processing time on a single machine is mainly consumed by the interferogram generation steps, which include image co-registration, interferogram generation and filtering, and phase unwrapping. These steps amount to over 90% of the total processing time, and can be easily parallelized, since they are carried out on single image pairs, with the exception of 3D phase unwrapping, which includes an initialization step involving the whole image stack.

Concerning our PS algorithm implementation, interferogram generation contributes to only about 50% of the overall processing time, also because spatial phase unwrapping is not carried out. On the other hand, compared to the SBAS method, a mean velocity inversion step, using the periodogram estimation technique is included, and significantly contributes to the processing time. Since the estimation is carried out independently on partially overlapping data patches covering the area of interest, it can be easily parallelized for cluster execution.

Concerning the generation of a geocoded coherence map, e.g. for CCD applications, in case of Stripmap acquisitions, the input image pair is split into patches, based on the number of available cluster nodes. In the ScanSAR or TOPSAR case, the data is naturally split into bursts, which are sent directly to the cluster nodes. Each node then carries out co-registration parameter estimation, interferogram generation and flattening, coherence estimation and geocoding of a data patch. Only a small overhead is required to split the input data, to derive image-wide co-registration parameters from those estimated for each patch, and to assemble the geocoded contributions returned by each cluster node. 

The performance of the three methods is assessed as a function of the available cluster nodes, in terms of wall-time (i.e. end-to-end execution time) and speedup factor with respect to the single-machine wall-time. Deviations from the ideal performance are discussed and quantified, including the impact of residual serial processing tasks (i.e. tasks which are executed on a single node even no matter how many are available), hardware resource contention (when more logical cluster nodes are instanced on the same physical machine) and data transfer overhead (to copy data to and from the cluster nodes). Results are presented for full-frame Sentinel-1a IW (TOPS) image stacks.

Due to the global warming and frequent human activities, the permafrost on the Qinghai-Tibet plateau is suffering serious degradation, which has caused severe ground settlement and hot-melt disasters, and then threatened the security of human engineering infrastructure including railway and highway. This study will use an advanced multi-temporal interferometric synthetic aperture radar (MTInSAR) technique to monitor the long-term surface deformation and thaw-melt hazards along the Qinghai-Tibet railway and highway. The SAR data include 45 C-band ENVISAT ASAR、19 L-band ALOS PALSAR and 30 X-band TerraSAR images, with their resolutions 30m, 7m and 3m, and time spans 7 year, 3 year and 1 year respectively. The study area locates at Beiluhe district between Tuotuo river and Wudaoliang, where the permafrost belongs to high-temperature and unstable type.

We develop an advanced MTInSAR methodology by taking advantage of existing SBAS and SqueeSAR technologies. To increase the point density per area, we synergistically analyze Coherent Scatterers (CSs) and Distributed Scatterers (DSs). The extraction of DSs is based on the identification of homogeneous pixels using a Kolmogorov-Smirnov (KS) test. Through analyzing the coherence matrix, constructing temporal attenuation model, and performing the filtering on both the spatial and temporal dimension, we deal with the problem of serious decorrelation which is common over the permafrost environment to amplify the extent and capability of MTInSAR technique.

Based on the deformation results we would map the potential and existing thaw-melt hazards along the railway and highway, analyze the properties of their spatial distribution and temporal evolution; At last, we study the deformation and damage features of railway and highway, and reveal the mechanism of the impacts of thaw-melt hazards. This research aims to promote the application of MTInSAR technique in permafrost environment, help monitor the thaw-melt hazards and protect the railway and highway.

Time series InSAR analysis gives a spatially dense set of geodetic observations of ground surface movement in the viewing geometry of the satellite platform, but with a temporal sampling limited to the orbital revisit of the satellite. Compare this to the GNSS technique, which can give a temporally dense set of geodetic coordinate observations in three dimensions but at a small number of discrete measurement points on the ground. Using both of these methods together can leverage the advantages of each in order to derive more accurate and validated surface displacement estimates with both high temporal and spatial resolution. In this contribution we will present results from a combined analysis of GNSS and InSAR data for the Sydney Basin, Australia, where unconventional gas extraction and subsurface coal mining is taking place.

A network of twenty geodetic monitoring sites has been established in the region of interest in June 2016 covering an area of about 20 x 20 km. Each geodetic monitoring site comprises a survey mark coupled with ascending and descending radar corner reflectors. Monthly GNSS campaign surveys are being performed at the monitoring sites since July 2016. Additionally, continuously operating GNSS instrumentation was established at four of the 20 sites enabling displacement tracking with high temporal resolution. At these continuous sites, daily coordinates are processed with the Bernese GNSS software V5.2 using a fixed network of surrounding continuously operating reference sites (IGS sites and high quality sites belonging to the national AuScope GNSS network). At the campaign sites, 24 hour GPS observations are used to derive a daily coordinate estimate on an approximately monthly basis using the same network of reference sites. The resulting displacement time series and linear velocities at the 20 sites are then further used for validation of InSAR. 

The coupled corner reflectors serve as tie points for the validation of surface displacements derived from Persistent Scatterer (PS) InSAR analysis with co-located GPS observations. RADARSAT-2 data has been acquired over the region of interest since July 2015 on ascending and descending tracks with 24 day acquisition frequency on each. In addition, archived data of seven Envisat tracks are analysed, mainly acquired between 2006 and 2010. The available SAR data acquired by different sensors and on different tracks are processed with the GAMMA software to produce interferograms from each stack of images. The interferograms are further analysed with the StaMPS software applying the multi-temporal PS and Small BAseline Subset (SBAS) approaches. The resulting line of sight (LOS) displacements are interpolated to a regular grid using Kriging for comparison and data combination. The data from different image geometries are then combined to robustly separate horizontal and vertical deformation components.

First results of InSAR analysis of Envisat image stacks indicate positive and negative LOS velocity anomalies of 10 mm/yr and more in the area of coal mining, whereas much smaller movements are detected close to the gas extraction area (less than 2 mm/yr). The inter-technique differences observed in these differing locations will serve as an important validation for the ability of each geodetic technique to resolve deformation of different magnitudes in the presence of atmospheric noise characteristic of Australia’s temperate south-east coast.

During the last decade, multi-dimensional SAR tomography (MD-TomoSAR) has been applied successfully in multicomponent motion estimation in urban environment. For measuring the deformation of architectural complex in natural environment, the application of MD-TomoSAR is limited by the low spatial density and non-uniform distribution of extracted persistent scatterers (PSs). In this paper, we present an improved MD-TomoSAR method based on persistent scatterer interferometry (PS-InSAR) to extract structural displacement information of ancient temples in Angkor site. By constructing the Delaunay triangulation network twice, some isolated single PSs and small networks yielded by rejecting unqualified arcs in first network can be connected within the reconstructed second network. This technique improves the spatial coverage and robustness of the connected network. Selecting the extracted single PSs as the reference points, overlaid double PSs can then be identified by constructing local star networks based on a local maximum ratio method. To robustly detect single and double PSs, beamforming are combined with M-estimator for parameter estimation, breaking though the intrinsic error of traditional beamforming estimation method without significantly increasing the computing burden. Besides, Angkor site is locates in rainforest environment in the tropical monsoon climate zone, the prominent atmospheric phase screen (APS) effect results in the high uncertainty of PSs detection and parameter estimation. To overcome this limitation, the proposed method can robustly detect single and double PSs without the need for prior removal of APS.

For analyzing the triggering mechanism of abnormally deformation for ancient temples in Angkor site, linear and seasonal deformation components are both taken in to consideration in the tomographic imaging model. 46 scenes of TerraSAR-X stripmap images spanning the periods 16/2/2011-16/12/2013 are used. Validated by field investigation, the results show that there is no ground subsidence tendency in the buffer and core area of Angkor site, while there are some structural instability trends located in parts of ancient temples. Compared to other ancient temples in Angkor, the seasonal deformation effect is relative prominence in the largest one - Angkor Wat. Moreover, the higher the PS locate, the larger the seasonal deformation amplitude is. Combined with other auxiliary data, we analyze the correlation between time series of non-linear deformation component and groundwater level. Finally, we propose several suggestions about the protection and sustainability of Angkor site in the future.

The study shows that the proposed MD-TomoSAR method are effective and reliable in deformation monitoring of Angkor site, and the results provide scientific supports for the sustainability of Angkor. Meanwhile, these approaches can be also extend to deformation monitoring of distributed artificial features in suburban areas.

Land subsidence, excessive groundwater withdrawal and thick haze, are the new unbearable pains of Beijing city.  Land subsidence is the disaster phenomenon of environmental geology with regionally surface altitude lowering caused by the natural or man-made factors. Many different factors can cause the land surface to subside. Beijing, the capital city of China, has suffered from land subsidence since the 1950s, and extreme groundwater extraction has led to subsidence rates of more than 100 mm/year. The first record of subsidence in Beijing was discovered by the survey and mapping departments in 1935, near Xidan. The measured maximum accumulative displacement was only 58 mm until 1952. Nevertheless, since 1950s, with the rapid development of the light industry in eastern Beijing, many wells were drilled and lots of water was unboundedly pumped from underground for industrial consumption. The local land subsidence area was gradually formed since then. As reported by the China Geological Survey (CGS), the groundwater level in Beijing has dropped rapidly and exceeded natural recharge since the 1970s, due to industrial development and the demands of a growing population. Especially for the years after 1999, the groundwater overdraft due to drought resulted in rapid development of the land subsidence in Beijing. 

Interferometric synthetic aperture radar (InSAR) has been widely used for monitoring local surface deformation associated with groundwater withdrawal. Temporal InSAR techniques such as persistent scatterer interferometry (PSI) and Small BAseline Subset (SBAS) have been successfully used to detect, map and analyze long-term and slow land subsidence for several years. And in several studies InSAR time series techniques have been used to investigate land subsidence in east Beijing Plain. 

In this study, we employ two SAR data sets acquired by Envisat and TerraSAR-X satellites to investigate the surface deformation in Beijing Plain from 2003 to 2013 based on the multi-temporal InSAR technique. The results show that land surface in eastern Beijing Plain deforms at different rates ranging from -12.86 cm/yr (subsidence) to 0.92 cm/yr (uplift) relative to the reference point, between Jun. 2003 and Nov. 2013. The InSAR-measured sinking rates and accumulated displacements agree with results estimated from ground leveling surveys. We use observation wells to provide in situ hydraulic head levels, in order to investigate the evolution of land subsidence and groundwater change in spatial and temporal. Then we analyze the accumulated displacement and hydraulic head level time series using continuous wavelet transform to separate periodic signal components. Furthermore, we implement cross wavelet transform (XWT) and wavelet transform coherence (WTC) to analyze the relationship between the accumulated displacement and hydraulic head level time series. Our investigation shows land subsidence in Beijing Plain has connection with hydraulic head level falling caused by over-exploitation of groundwater; the ability of multi-temporal InSAR technique is potentially important for valuating land subsidence and investigating its formation mechanisms.

Land subsidence of Chaoyang district develops quickly and seriously in recent years and there are too many subway lines in Chaoyang district. The land subsidence may influence subway constructions and operation. In this paper, 30 TerraSAR-X images acquired from April 2010 to September 2013 were used to get the settlement rate around subway lines by IPTA approach and the settlement rate is between -77.34mm/yr to 1.27mm/year in Chaoyang district. There are differences of land subsidence from north to south around the subway lines. The land subsidence is mainly occur in Dingfuzhuang, Fatou where in the east of study area and northeast of Wenyu River. We choose the section from Communication University to BALIQIAO station as a cause and predict the subsidence trend from 2013 to 2015 by ARIMA model. The result of this prediction shows there is a tendency of decline followed by a rise then decline from 2013 to 2015. The trend of settlement is slow down, the monthly rate is between 0.5mm/month to 2mm/month.

In the Cordillera Blanca, Peru, significant landslide hazards arise from different types of terrain motion processes. Such landslides have the potential to cause massive loss of life and damage to infrastructure, as documented by events in the past. At high elevations, in combination with numerous glacial lakes, such slope instabilities often act as triggers for potentially far-reaching glacier lake outburst floods (GLOFs). Numerous structural hazard mitigation works have been undertaken in the past decades, in particular related to glacier lakes (e.g. lake drainage or reinforcements of natural dams). Recently efforts are undertaken to complement these structural hazard mitigation works by non-structural measures, such as improved spatial planning and the installation of early warning systems. Such systems have pioneering character in high mountain areas and have to cope with extreme conditions, but also huge potentials for application in other regions and to other cases.

In contrast to hazards and risks related to glacier lakes, less attention has been payed in the Cordillera Blanca to terrain displacements and landslides at elevations below the glacierized peaks, despite numerous zones of instable terrain and even catastrophic landslide events. Radar remote sensing techniques such as repeat-pass differential SAR interferometry (InSAR) provide powerful tools for the mapping and monitoring of such land surface deformations at fine spatial resolutions. Tropical mountain regions, such as the Peruvian Andes, are particularly suitable for InSAR applications, since limited vegetation and snow coverage allows for slope stability analyses and monitoring throughout the entire year.

Within the ESA funded project S:GLA:MO (Slope Stability and Glacier Lake Monitoring) we compiled a landslide inventory on the basis of multi-source EO information. Slope deformations have been identified using radar imagery from ERS-1/2 SAR, ENVISAT ASAR, TerraSAR-X, ALOS-1 PALSAR-1 and Radarsat-2. The InSAR processing included differential interferograms with varying temporal baselines using TanDEM-X to remove the topographic related phase contribution as well as Persistent Scatterer Interferometry (PSI) applied on stacks of ENVISAT ASAR and ALOS-1 PALSAR-2 data. In addition to radar information, optical high-resolution imagery from different acquisition dates and TanDEM-X were used for the interpretation by a specialist in order to classify the type of movement and the underlying processes. Our landslide inventory contains five velocity classes (0-2 cm/yr, 2-10 cm/yr, 10-50 cm/yr, > 50 cm/yr and uncertain velocity) and several processes such as landslide, rockglacier, debris movement, unstable moraine, dead ice, rockwall instability, settlement and also mining. Information of the process timing (continuous, episodic, uncertain, unknown), a distinction between certain and uncertain delineation of the process and a column with comments from the interpreter are also included. In total 126 landslide zones were delineated.

In October 2015, a follow-up project was launched in the framework of ESAs Alcantara initiative with the aim to evaluate the potential of InSAR for the assessment of landslide hazards and the integration of EO information in monitoring and early warning systems in the Cordillera Blanca. In a first phase of the project, the recent state of activity of the mapped landslides was analysed with use of recent Sentinel-1 and ALOS-2 PALSAR-2 interferograms. Then, both geomechanical modelling of slope stability and in-situ measurements have been used for validating the inventory. A simple slope stability model has been applied to the study region, yielding a map of spatially distributed information of a factor of safety, which was compared to the landslide inventory with a focus on sites where terrain deformations are observed, despite a relatively high factor of safety. Terrain deformation due to human activity, such as mining, have been excluded from this analysis.

The information provided by the InSAR-based landslide inventory has not been available so far for this region and is of high interest for local authorities, on the one hand for identifying and prioritizing hot spots for hazard prevention efforts and disaster risk reduction measures, but also for the identification of illegal mining activities. In particular the combination of purely InSAR-based data products with other EO data and field measurements helped to generate value-added information and exploit the full potential of this technology.

Future activities aim at designing a prototype landslide monitoring service based on both spaceborne SAR information and in-situ measurements. Other efforts evaluate the potential for an operational integration of InSAR information in recently developed early warning systems for glacier lake outburst hazards. Despite being promoted by high-level international policy documents, such early warning systems still have a pioneering character in high mountain areas. Based on the gained experiences, InSAR seems to be able to provide early indication of potentially adverse developments, which could help to anticipate possible catastrophic mass movement events.

The recent availability of large amounts of remotely sensed data requires setting up efficient paradigms for the extraction of information from long series of multi-temporal, often multi-sensor, datasets. In this field, monitoring of terrain instabilities is currently performed through algorithms which estimate millimetric displacements of stable (coherent) objects, through analysis of stacks of SAR images acquired in interferometric mode. The result is generally a decomposition of at least part of the complete complex covariance matrix obtained from all possible pairwise combinations of the images in the stack, separating its spatially- and temporally-correlated parts.

The same SAR temporal data stacks can be used to apply change detection algorithms, to reveal, over potentially huge spatial scales and with high resolution, terrain surface changes due to e.g. environmental hazards (floods, fires, earthquakes). In this case, again, the temporal covariance matrix contains in practice all the information related to the environmental changes.

The covariance matrix, or its normalized version, known as coherence matrix, expresses thus all the information content related to a time series of remotely sensed, coherent data. In the case of SAR data, this kind of representation offers a unified framework for the study of phenomena linked either to the presence of “periods” of persistent scattering characteristics, or to changes of backscattering patterns, hinting to variations in the terrain characteristics.

The average operation, involved in the definition of the above-mentioned covariance and coherence matrices, has to be performed necessarily over “homogeneous” pixel sets. This homogeneity criterion can be intended in various ways, including the one connected to the covariance definition itself, thus leading to a sort of recursive estimation process. Moreover, such homogeneity measures are often used as a substitute for the classical Euclidean distance in nonlocal estimate implementation frameworks, used for instance in the design of effective SAR speckle filters.

The coherence matrix highlights the role of the interferometric phase. After having suitably modeled various phase contributions, due to topography, atmosphere, etc., it is possible to detect periods in which a target remains stable, and can thus be used as a benchmark for estimating ground deformations or other effects related to the variations of the signal optical path.

From the above discussion, it appears that a thorough, physically based modeling of the coherence over such long times series of SAR data constitutes a priority for efficient data exploitation.

We illustrate some of the inference which can be made starting from a time series of more than a hundred COSMO-SkyMed (CSK) images acquired in InSAR mode over the Haiti capital of Port-Au-Prince, spanning a period of almost 3 years with short repeat times. Such tight acquisition schedule can be obtained nowadays with latest-generation SAR constellations such as CSK or (at lower resolutions) Sentinel-1A/B. On the mentioned CSK dataset, some recently proposed models for coherence have been tested over selected regions of interest, covering different terrain types, from forest, to cultivations, to man-made smooth surfaces such as tarmac lanes, to built-up areas. Coherences are estimated over homogeneous pixel sets determined through a nonlocal criterion. Results may help shed some light on the nature of constant, decaying and periodic components of the InSAR coherence.

 

ACKNOWLEDGMENTS

Work supported by the Italian project “APULIA SPACE” (PON03PE_00067_6), PON Ricerca e competitività 2007-2013.

As a special kind of geo-hazard, landslide is characterized by increasing frequency, random distribution, strong concealment and complicated causes, and has become one of the most common and serious hazards in China. Accurate monitoring of the surface deformation associated with landslides is vital to protect people from the landslide hazards. Spaceborne Interferometric Synthetic Aperture Radar (InSAR) technique has great potential in this field due to its advantages of all-day, all-weather, spatial continuous, high precision and contact free. Especially with the successively launch of the SAR satellites and the great abundance of the SAR images, InSAR technique has now entered the era of time series analysis and multi-data fusion. Many successful applications with respect to the monitoring of landslides have been conducted by employing the well-developed multi-temporal InSAR (MT-InSAR) algorithms.

However, since landslides usually occur in the mountainous area characterized by heavy vegetation, steep terrain and complicated atmosphere, it is very difficult to eliminate the inherent errors of InSAR, which greatly degrade the capability and applicability of InSAR in monitoring landslides. For instance, InSAR measurements from single track are vulnerable to the geometric distortions in SAR images (i.e., foreshortening, layover and shadow). In addition, only the line-of-sight (LOS) projection of the three-dimensional (3-D) surface deformations can be provided by InSAR measurements, which could induce misinterpretation in the investigating of landslides that always behave as down-slope deformations. These limitations greatly degrade the applicability of InSAR in the landslide monitoring. 

In this paper we employ 16 L-band ascending SAR images acquired from ALOS satellite and 18 C-band descending SAR images acquired from EVNISAT satellite to investigate the slow-moving landslides in Zhouqu, China. Located in the eastern of Qinghai-Tibet Plateau, Zhouqu has been suffered from the landslides-related geo-hazards for several decades, including the giant debris flow occurred on August 7, 2010 that caused nearly 1800 deaths. The geometric distortions of the ALOS ascending and EVNISAT descending datasets are first assessed according to the SAR imaging geometries and the topography of the Zhouqu area. It is found that the performance of ALOS data is generally superior to that of the ENVISAT data in the study area. A new polynomial-based model is proposed to estimate the phase ramp in the single-look wrapped interferograms from the phase of the multi-look unwrapped interferograms. With the modified MT-InSAR method, the time series results retrieved from the ALOS data reveal that noticeable slow-moving landslides are found in the Luojiayu-Sanyanyu, Suoertou, Xieliupo and Nanshanqiaotou areas, which have strong correlation with the seasonal precipitation in the area as well as with the 2008 Wenchuan earthquake. Furthermore, the ALOS and EVNISAT data-derived independent line-of-sight deformation rate results are combined to retrieve the real down-slope deformation in the Xieliupo area on the basis of the surface-parallel flow assumption. The results reveal that the Xieliupo landslide is dominated by the east-west deformation, and moves toward the Pai-lung River as a result of the gravitational force. A distinct acceleration of the down-slope deformation occurs in the front end of the slope, with a maximum velocity exceeding 50 mm/yr, which poses the threats to the Pai-lung River and S313 provincial highway.

This work presents possibilities to improve the quality of land subsidence monitoring by supplementing analyzed data with Sentinel-1B images. In presented work DInSAR (Differential Interferometry SAR) technique was used. Results obtained with DInSAR technique provide information about small deformation of the surface. Using this technique the areas endangered by ground subsidence can be identified and appropriate steps in order to save the infrastructure, buildings, roads, and also human life can be taken. Ground subsidence monitoring is especially important in the areas where active underground coal mines occur. In Poland, the Upper Silesian Coal Basin is the active mining area.

Due to the intensity of mining activities in this region, resulting subsidence troughs are characterised by dynamic changes. Hence, it is necessary to use high temporal resolution images in order to ensure appropriate monitoring of ground subsidence in the dynamically changing surfaces. Until 24 April 2016 differential interferograms could be generated every 12 days by using only Sentinel-1A data. Since Sentinel-1B was launched, it is possible to obtain data every 6 days. Presented results show 12-days interferograms and 6-days interferograms. Complex analysis of Sentinel-1A and B data (acquired in Interferometric Wide Swath mode) allows detection of both areas: where ground subsidence occurs rapidly (using 6-days interferograms) and where ground subsidence occurs more slowly but regularly (using 12-days interferograms).

Belgium is renowned for its diverse collection of built heritage, visited every year by millions of people. Because of its cultural and economic importance, conservation is a priority at both federal and regional levels. Buildings may suffer from structural instabilities related to industrial and urban development, such as groundwater extraction, mining and excavation. The adequate protection and preservation of built heritage requires an integrated analysis of environmental, architectural and historical parameters. The GEPATAR project (GEotechnical and Patrimonial Archives Toolbox for ARchitectural conservation in Belgium) aims to improve management of the patrimony by integrating datasets from the archives of the Geological Survey of Belgium – Royal Institute of Natural Sciences (RBINS) and the Royal Institute for Cultural Heritage (RICH). Satellite radar interferometric techniques are used to assess ground stability risk of monuments.

PSI processing of ERS and ENVISAT archive data has provided countrywide deformation maps, showing regional trends of ground movements between 1992 and 2010. The observed movement patterns can be related to regional subsurface processes, such as compaction of soft alluvial sediments, mining activities and groundwater extraction.

In the Brussels study area, analysis of ERS and ENVISAT data has revealed evidence for continuous ground movements over the last decades. Uplift of the historic city centre (at rates up to 3 mm/yr) can be attributed to recharge of subsurface aquifers, after water-demanding industries were relocated from the city centre to the periphery since the 1980s. In addition, the city’s main infrastructure suffers from local stability problems, such as collapsing tunnels and cracks in buildings, highlighting the urgency for geotechnical monitoring schemes. These ground movements (especially if differential) are a potential risk for the built heritage in the historic centre.

In addition, an extensive time series of TerraSAR-X Stripmap archive data acquired between 2011 and 2014 over Brussels has been processed and analysed. The TerraSAR-X data ensure data continuity after the ESA missions ended and provide much higher point density. The applicability of TerraSAR-X Stripmap data for assessing the stability of individual buildings in urban zones is currently being evaluated. Patterns of differential ground movements derived from PSI processing are used to identify monuments that are potentially at risk.

In the following steps of the project, selected buildings will be assessed by our project partners through visual inspection, damage survey and stability analysis. These assessments will then be correlated with the evidence of ground deformation from PSI processing. Ultimately, the case studies should demonstrate the validity of PSI products in operational workflows for the management of cultural heritage on both local and regional levels.

The GEPATAR research project is funded by the BRAIN-be program of the Belgian Science Policy (BELSPO).

Warsaw agglomeration is characterized by complex geological structures and by various ground deformations. Some of them are geological-based or hydrological based large-area deformations. Additionally there are many places of local man-made subsidence caused by the construction of new apartments and office buildings, metro lines and other underground works. Many of local deformations are located close to the natural or artificial escarpments. History of movements of earth surface in Warsaw agglomeration in the period 2004 – 2016 was recreated; it is based on the various types of the radar data gathered within the project: “The integrated system of surface deformation monitoring based on Persistent Scatterer Interferometry, measurements from permanent GNSS stations and precise leveling” financed by the National Center for Research and Development in Poland.

The presented work shows the comparison of results obtained, based on the TerraSAR-X and Sentinel-1 data. The data from both satellites were registered from the ascending orbit during the period from November 2015 to October 2016 (c.a. 30 images in each set of the data). Both data sets were processed separately within the Gamma software using in very similar way. The results were validated using measurements from permanent GNSS stations and precise leveling. Next, the results from both data sets were compared taking into account PS density in various parts of the city characterized by different type of buildings, possibility to detect various types of deformations (due to the construction of metro line, subsidence of new buildings, deformation of the escarpment and subsidence of the terrain due to the changes of ground water level), and accuracy measurements for each type of the movement.

The Mosul Dam is well known for its instability, as the riverbed is made of unstable soft soil and gypsum, a mineral that dissolves as water runs through it. The dam structure has to be cemented daily in order to keep water from seeping through.

To monitor the structure’s stability, the U.S. installed sensors as an early warning system, as there are serious concerns that the dam could collapse and the consequences would be fatal: Between 500,000 and 1.5 Million people would be at serious risk. A wall of water – up to 14 m height – would be unleashed. Mosul would be devastated and even Bagdad, some 275 miles away, would be impacted. (source: http://edition.cnn.com/2016/04/08/middleeast/inside-the-mosul-dam/)

Airbus Defence and Space has helped to identify the risks by applying interferometric time series analysis in order to monitor the dam. More than 30 TerraSAR-X high-resolution SpotLight (1m resolution) scenes have been acquired, covering a one year time period between April 2015 and April 2016. Data has been processed with a SBAS approach to deduct information on surface deformation with millimetre precision. The high resolution WorldDEM™ digital elevation model was used to introduce accurate small-scale dam height information. We demonstrate that the use of TerraSAR-X SpotLight imagery and WorldDEM™ as elevation input is an ideal source for the reliable derivation of deformation features on critical infrastructure.

InSAR derived findings were correlated with satellite based radar altimetry measurements of water level in order to analyse the influence varying water pressure.

Significant surface movement effects of the dam crown (up to 9 mm per year) and of sinkholes within the dam’s vicinity are clearly detected and measured. The results show a number of surface movement anomalies, which were either visible through pure vertical displacements of the surface (e.g. through underground dissolution) or lateral movements of the dam crown induced by water pressure of the reservoir and / or underground instabilities: http://arcg.is/2cCYjc6

Interferometric time series analysis based on high-resolution TerraSAR-X data in combination with WorldDEM™ digital elevation model, is a sound complement to terrestrial surveying in general, whenever surface information or object movements is required. Nevertheless, it is also an alternative in adverse conditions such as remoteness, climate, wilderness or when the social / political environment cause unwarranted risks for operating staff. Furthermore, the space-based solution offers a cost-saving potential due to significantly reduced effort and time for staff mobilisation and related safety measures.

Monitoring of the vertical ground motions along section of the A-1 highway in the vicinity of Ruda Śląska using multi-temporal SAR data

 

Zbigniew Kowalski¹ e-mail: zkow@pgi.gov.pl

Marek Graniczny¹  e-mail: mgra@pgi.gov.pl

Maria Przyłucka¹ e-mail: msur@pgi.gov.pl

¹Polish Geological Institute (PGI) – National Research Institute , 00-975 Warszawa, Rakowiecka 4, Poland

 

Keywords: Upper Silesian Coal Basin, subsidence basins,  hard coal mining, DInSAR, A4 highway

 

            Construction and development of road infrastructure were among the major tasks that Poland had to take after joining the European Union. That did allow us to catch up with the West, as well as to improve the Polish economy. In the years 2007 - 2013, Poland received more than 10 billion euros for the construction and development of roads.

            Discussed pilot area is located in Upper Silesian Coal Basin (USCB) and covers the whole city of Nowa Ruda Śląska and parts of Mikołów, Zabrze, Katowice and Świętochłowice communities. The pilot area is in the shape of rectangle with sides 7.4 to 8.9 km, covering an area of 65.5 km². The highway was open to use in 2005. The pilot area is characterized by large deformations of the terrain caused by the exploitation of natural resources and in particular hard coal. Four coal mines are operating in this area: KWK “Pokój”, KWK Halemba – Wirek, KWK Bielszowice and KWK Ruda Śląska – Radoszowy. The longwall exploitation system is used in these mines with caving or hydraulic fill. KWK “Pokój”, which seems to have the greatest influence on the presence of deformation in the vicinity of highway undertake its activities in the highly urbanized area of the Ruda Śląska city. It operates three mining levels: 320, 600 and 790 m.

            Satellite radar imagery proved to be an effective tool to detect vertical deformations in the Earth ́s surface over very large regions. One single satellite image can cover areas up to 10 000 km². Deformations of the ground owing to subsidence as little as a few millimetres can be detected using the Interferometric Synthetic Aperture Radar (InSAR) technique.  This technique involves having multiple images over the same region that have been acquired over time.

            Three different satellites (ALOS L - band,  TerraSAR – X - band and Sentinel -  1 – C band) from three different periods of time were analyzed (Fig. 1).

·        ALOS time span 2008.02.25 – 2008.05.27 (92 days) ·        TerraSAR - X time span 2012.11.11 – 2012.12.03 (22 days) ·        Sentinel 1 time span 2016.10.27 – 2016.11.08 (12 days)  

The biggest numbers of subsidence basins were identified at the oldest ALOS images 8 (6 from the northern part of the highway and 2 from southern), following 4 basins (3 from the north and 1 from south) in TerraSAR – X and finally 5 basins (3 from north and 2 from south) in Sentinel 1 (October – November 2016). Most probably ALOS L – band is most suitable for identification of such basins due to the longest wavelengths and possibility of penetration through vegetation. However, the performed analysis revealed that vertical deformations in the vicinity of A- 4 highway still exists and pose a serious threat to road infrastructure.Początek formularza

 

Fig.1Subsidence basins and the maximum values of subsidence[mm]caused by the exploitation of hard coal in the area of ​​Ruda Śląska near the A4 motorway. A – ALOS time span 2008.02.25 – 2008.05.27 (92 days) B – TSX time span 2012.11.11 – 2012.12.03 (22 days) C – Sentinel 1 time span 2016.10.27 – 2016.11.08 (12 days)

Landslides are catastrophic events which may affect both human life as well as the man-made and natural environment. They are resulting from different causes such as natural causes or human activities, while they appear a variety of forms. Early prediction of landslides as well as landslide hazard analysis and mapping are essential in order to reduce property damage and loss of life. In this context, many methodologies have been developed aiming at the investigation of landslides and their better understanding. Recently more approaches utilizing multi-criteria operations by the computer as well as timely and high-quality information derived from space-borne observations, are exploited for landslide investigation with effective results. Such an approach is interferometry, a technique of acquisition earth surface height information through the measurement of the phase of the backscattered signal. These measurements are very dense so that the generated Digital Surface Model (DSM) may contribute to the effective determination of the topography.

This work is focused on exploitation of Sentinel-1 data for the detecting of a landslide in Klepa village, Aitoloakarnania Prefecture, West Greece. Sentinel-1 data were selected as they are timely, with global coverage and easily accessible. The specific landslide occurred in February of 2015 as a result of interaction of the geological formations with heavy rainfall, which led to intense relief changes. In particular, the wider area of Klepa is geologically structured by limestones, cherts and fluvial deposits. The intense and protracted rains caused in the lithological unit of cherts a great landslide with a length of 240 m. and a width of 130 m. The landslide has completely destroyed four houses of the settlement, while several houses located in the crown of the landslide have been seriously damaged and evacuated. Moreover many houses in distance up to fifty meters away from the slide were showed cracks as well as huge damages were appeared on the road network.

In order to recommend appropriate constraint actions of the phenomenon and protect the settlement, a study with geotechnical boreholes was materialized. Thereby, at an early stage it was implemented the mapping of the wider region, imprinting the lithological units, tectonic elements and the landslide zone. Additional several points were collected using a Differential GNSS receiver while the geotechnical boreholes contributed to collect data and monitor the area continuously. Regular measurements of the boreholes using inclinometer have demonstrated that the landslide phenomenon is still in progress. In that context, interferometric DSM generation was utilized aiming to identify and map any terrain change after the great landslide event in February of 2015. In particular, Sentinel-1 data from October of 2014 till April of 2015 were acquired and they were subjected to interferometric process resulting on extraction of two DSMs, one before the landslide and one after it. Those DSMs were compared to each other and height changes were mapped.

The advent of the Sentinel-1 constellation poses an opportunity for terrain deformation surveillance at unprecedented coverage and temporal resolution. Availability of free imagery over a study zone every 6 days facilitates subsidence monitoring at regional scale, generally unaffordable with existing X-band, high resolution sensors because of cost constraints.

However, the issue of sensitivity still remains. Is Sentinel-1 IWS spatial resolution enough for detecting and quantifying small-scale subsidence phenomena? Is C-band sensitivity to deformation good enough when averaged over this pixel size?

This work proposes a study of the subsidence phenomenon over two different areas in Catalonia, based on the application of PSI interferometric techniques (Persistent Scatterer Interferometry). The first area covers the city of Barcelona and its surroundings, where several subsidence phenomena have been detected and associated to public works, aquifer depletion, etc. The second area centers in the middle Llobregat river basin, where salt mining activity yields, direct or indirectly, important subsidence processes. To compare results from both satellites, the time-overlapping part of Sentinel-1 and Cosmo-SkyMed stacks in these areas of interest will be analyzed. It covers an observation time window of approximately 9 months, from February to October 2015, even though additional images outside this window will also be selected in order to obtain accurate deformation results. For their processing, the Differential Interferometric Software developed at the Institute Cartogràfic i Geològic de Catalunya (DIS-ICGC) will be used. The results of the processing of both satellites stacks will be compared, paying special attention to known subsidence phenomena already studied and quantified.

We estimate land displacement in the Perth Basin, Western Australia, using 109 scenes from the TerraSAR-X satellite SAR (synthetic aperture radar) mission, acquired over four integer years from October 2012 under the German Aerospace Center (DLR) science project LAN1499. Groundwater extraction has been occurring since circa 1975, increasing in some areas since 2000 to cater for the growing domestic and horticultural demands for water in the Perth region. Height time series from two continuously operating global navigation satellite system (CGNSS) stations in Perth’s north since the mid-1990s indicate subsidence of up to -6 mm/yr in the early 2000s, but which is non-linear in time.

The extent of the TerraSAR-X scenes is ~30 km by ~50 km, enabling a displacement map of the Perth metropolitan area to be produced. We use the combined small-baseline (SBAS) and persistent scatterer (PS) methods to estimate linear velocities at high spatial resolution. CGNSS are used to constrain the InSAR measurements, the latter of which only provides relative displacement values. Time series analysis at different locations indicates that spatially variable seasonal and other non-linear displacement is occurring.

This variability appears to be correlated with surface geological and local groundwater extraction volumes/rates. Subsidence in a narrow coastal band in the north west of Perth and also in the eastern bound abutting the Darling Scarp in the Swan River Valley is indicated. Parts of the Swan River Valley also display seasonal displacement of ~10 mm, which are attributed to expansive surface clays which shrink and swell in the distinct dry, hot and cool, wet seasons respectively. The high spatial resolution of TerraSAR-X also enables the identification of displacement of features in harbour areas, indicating small areas of relative subsidence.

The non-linear behaviour suggests that constraining the TerraSAR-X PS at a linear rate from CGNSS may not be appropriate. We test various constraints and the effect these have on the displacement map, and time series. Hence, areas that indicate particular displacement characteristics when constrained to a linear velocity may be influenced by the choice of constraint.

Time series from a repeat levelling profile approximately matching the TerraSAR-X 2012-2016 epoch is used as additional validation for the InSAR. Seasonal and long-term displacement from the TerraSAR-X correlates with the levelling at many levelling benchmarks along the profile. The magnitude of the displacement varies at some benchmarks, where the levelling suggests up to 30 mm of seasonal displacement compared to ~10 mm from the TerraSAR-X. The cause of this difference is assumed to be the higher spatial resolution of the levelled heights at a benchmark compared to the ~40 metre grid used for the TerraSAR-X displacement map.

Roza Khutor is the mountain cluster of the Big Sochi area - the place of the 2014 Winter Olympic Games. Now it is the biggest ski resort in Russia. This area has always been a region of high landslide risk due to widely spread clays and marls saturated by water from abundant rainfalls and snow melting. In recent years landslide risk assessment has become vital because of strongly increased human impact.

The conditions for InSAR are not favorable there. Mountains (heights 1000-2500m) are almost fully covered by vegetation and only tops of peaks are free from vegetation but covered with snow since November till May instead.

We have been studying the Big Sochi area since 2011. Earlier using the StaMPS software we made PS-InSAR inventory map for the coastal part based on Envisat (13 images, track 35D- 29.11.2010-23.03.2012), ALOS PALSAR (18 images, track 588A - 22.01.2007-17.09.2010) and TerraSAR-X ( 17 images, track 54A - 24.12.2011 -13.09.2012) data sets. All these images span the period before and during the construction of Olympic Games facilities. We found out that for the coastal landslides space adaptive filtering of amplitude of the ALOS PALSAR images by applying Kolmogorov-Smirnov test similar to the SqueeSAR approach considerably  increases  the number of PS. Applying also DePSI, NSBAS and offset tracking we analyzed in detail activity of many particular landslides in the coastal cluster of the Big Sochi. But reliable results for Roza Khutor were obtained only for the ALOS PALSAR data set. About 30 landslides were revealed within the study area 6 of them being not mentioned on the ground data based inventory map. The vital question is what happens there now?

The Roza Khutor area is covered by Sentinel-1  images from the four tracks: two ascending 43A, 145A and two descending 21D and 123D. This makes the idea of regular future updating of PS-InSAR landslide inventory map very inviting. Using SNAP we calculated different pair interferograms and it was found out that it is possible to get quite good average coherence (0.3-0.4) for pairs of images separated in time by 12 days (in case of no snow cover). Also coherence maps strongly correlate with precipitation and temperature data from the Tuapse weather observing station as they present information about snow cover appearance and melting. Thus, a set of 15 images from track 43A (since 06/04/2015 up to 10/03/2015) was selected for further investigation. Using the set of consequent pair interferograms it appeared possible to update the previously created PS-InSAR inventory map for Roza Khutor based on the ALOS PALSAR data. A number of pre-existed landslides were revealed also using pairs of S-1A interferograms. However, displacement maps for almost all of the pairs are far from being reliable due to unwrapping errors and atmospheric artifacts. Even displacement maps from SBAS module of SARSCAPE using 45 interferograms can not be considered fully reliable. This needs further improvements, probably using another PS-InSAR techniques and more accurate DEM.

The work was partly supported by the RFBR grant 16-05-00937.

Acknowledgements

We are very grateful to Prof. Andy Hooper for the StaMPS/MTI software, assistance and  very helpful discussions and advice.

We would like to thank the SNAP developers team for their assistance.

 Authors acknowledge the European Space Agency ESA (project C1-7991), the Japanese Space Agency JAXA and the Deutsches Zentrum für Luft- und Raumfahrt DLR (project LAN1247) who kindly supplied us with the SAR data for this study.

 

 

Frequency-Modulation Continuous-Wave InSAR is the combination of Frequency-Modulation Continuous-Wave ( FMCW ) technology and Interferometric Synthetic Aperture Radar ( InSAR ). It not only has the capability of topographic mapping with high accuracy under all-day, all-weather conditions, but also has small-volume, light-weight and cost-effective advantages due to the large time-bandwidth product and small transmit power of frequency-modulation continuous-wave[1-5]. Small-volume, light-weight InSAR system helps to reduce space and load requirement of platforms and can be installed on smaller, more flexible and less expensive platforms to satisfy increasing application demands in many areas such as topographic mapping and deformation monitoring. Therefore, FMCW InSAR technology is a meaningful direction in the future. This manuscript briefly describes the system, signal acquisition, signal processing and results of our FMCW InSAR.

       We have been doing research on FMCW InSAR system and signal processing for several years, and have accomplished system design, implementation and several flight experiments of the Ka and Ku-band FMCW InSAR, as shown in Figure 1. Our FMCW InSAR system mainly consists of three components: radar host, Mini-POS and Antennas. The overall dimensions of radar host is , and the radar host weighs only 1.6 kg. The total power consumption of the radar system is just 65W.

 

 

 

( a ). Radar host

( b ). Mini-POS

( c ). Antennas

Figure 1. FMCW InSAR system. ( a ) is the radar host, ( b ) is the Mini-POS, and ( c ) is the Antennas

       The Ku-band radar system uses frequency-modulation continuous-wave as the transmit signal with 600 MHz bandwidth at 14.5 GHz carrier frequency, and it receives echo signals in dechirp mode using two antennas separated in cross-track direction. The radar system works in strip mode with positive side-look.

       The key issue in signal processing is motion compensation. For the FMCW InSAR system mounted on small platforms, the trajectory and attitude errors can be considerably high due to atmospheric turbulence and platform properties, such as its light weight and low flight height[4, 6]. Moreover, we cannot use conventional large-volume, heavy-weight and high-accuracy inertial measurement units due to the limit of platform’s capability. Therefore, residual motion error is required to be estimated and compensated to obtain high-accuracy topographic mapping result. In the data processing stage of our FMCW InSAR system, we use the compensation scheme by combing the POS data and estimated residual error from echo data. Firstly, we carry on preliminary compensation using POS data, and then improve the image quality by using Doppler parameters estimated from echo data. Secondly, we compensate the high order baseline error by using multi-squint technology and estimate the low order baseline error according to corner reflectors. The processing result demonstrates the effectiveness and reliability of our motion compensation scheme.

       We carried out an airborne flight FMCW InSAR experiment in Shanxi province, China. Finally, we obtain topographic mapping result with 0.5m height accuracy using our signal processing scheme. The interferogram and DEM are demonstrated in Figure 1. The interferometric results demonstrate the effectiveness and reliability of FMCW InSAR for high-accuracy topographic mapping.

Figure 2. FMCW InSAR result. ( a ) is multi-look Interferogram, and ( b ) is the final Height Map

 

Keywords: FMCW, Mini-InSAR, Signal processing

 

References:

[1].   Cumming, I.G. and F.H. Wong, Digital Signal Processing of Synthetic Aperture Radar Data: Algorithms and Implementation. 2004.

[2].   Fu, K., P. Siqueira and R. Schrock. A university-developed 35 GHz airborne cross-track SAR interferometer: Motion compensation and ambiguity reduction. 2014: IEEE.

[3].   Scannapieco, A.F., A. Renga and A. Moccia, Preliminary Study of a Millimeter Wave FMCW InSAR for UAS Indoor Navigation. Sensors, 2015. 15(2): p. 2309-2335.

[4].   Xing, M., et al., Motion Compensation for UAV SAR Based on Raw Radar Data. IEEE Transactions on Geoscience & Remote Sensing, 2009. 47(8): p. 2870-2883.

[5].   Aguasca, A., et al., ARBRES: Light-Weight CW/FM SAR Sensors for Small UAVs. Sensors, 2013. 13(3): p. 3204-3216.

[6].   Zaugg, E.C. and D.G. Long, Theory and Application of Motion Compensation for LFM-CW SAR. IEEE Transactions on Geoscience and Remote Sensing, 2008. 46(10): p. 2990-2998.

 

 

Persistent Scatterer Interferometry (PSI) is a valuable technique for studying ground deformation in Belgium and providing more information on their spatial and temporal patterns. In this work, about 600 SAR images of different tracks from the European satellites ERS and ENVISAT were processed. TerraSAR-X data covering the time span 2012-2014 and the area of Brussels were processed as well. A mapping of the annual average velocity of the PS of the entire period highlighted already known ground movements discovered during country scale levelling campaigns. But thanks to its millimetre precision, it permitted to highlight new movements that were not identified before.

Seven large subsidence or uplifting areas can be spotted from the velocity map. From north to south, the regions facing ground movements are: The West-Coast, Antwerp and along the Schelde river and estuary, the Limburg Campine coal mines basin, Merchtem-Londerzeel, the cities of Brussels and Liège as well as a large area related to the Tournai-Mons-Charleroi coal mines basin. The purpose of this work is to make an overview of the situation (observations, first interpretations) of these ground movements. The highest positive (uplift), 20 to 25 mm/year and negative (subsidence) -17 to -20 mm/year velocity values are recorded around the former Limburg coal mines areas. The movements are closely linked to the groundwater extraction needed during the exploitation time and the recharge of the mining aquifer occurring at the end of the pumping.

Among the seven highlighted zones, in Merchtem 25 Km NW of Brussels, a ground subsidence (-3 mm/year on average) is occurring since the beginning of the ERS acquisitions in 1991. Through the time (ERS, ENVISAT and TerraSAR-X) the subsidence pattern reduces his extent and is replaced by an uplift due to the raise of the water table. Piezometers located in the deep Cambro-Silurian aquifer show a clear recharge of this aquifer since the late 1990’s. The subsidence is finally reduced to a zone where three breweries are very active and pumping groundwater for the production in the Ledo-Paniselian aquifer and in the Cambro-Silurian as process water. The Stella Artois brewery located in Leuven 30 Km E of Brussels is facing as well a ground subsidence clearly visible with the PSI data. It seems that the Belgian beers production may induce ground deformation.

The coastal areas of East and Southeast Asia are some of the most densely populated in the world. Due to the combination of over-exploitation of groundwater, removal of peat forests, oil and gas extraction, and natural sedimentation processes, many coastal areas are suffering from rapid subsidence. Subsidence of coastal areas increases relative sea level, leading to flooding and increased susceptibility to tropical storms and typhoons. In some areas, the known presence of natural subsidence has led to resistance against imposing limitations on resource extraction, while in others successful management strategies have been enacted. In combination with in-situ data, high resolution InSAR-based maps of subsidence [e.g. Aopbaet et al., 2009, Chaussard et al., 2013] provide a key tool for understanding the relative contribution of these sources of subsidence, as well as a potential management tool to prevent further sinking.

In this study we have selected a number of highly populated and rapidly developing coastal areas of East and Southeast Asia - Singapore, Jakarta and Semarang (Indonesia), Yangon (Myanmar), Bangkok (Thailand), and Hong Kong and Macau (China) - in which to study their ongoing subsidence. The locations were selected due to the availability or planned availability of in-situ observations from research being carried out by scientists at the Earth Observatory of Singapore (EOS) and our collaborators in neighbouring countries. We combine InSAR data from a set of previous and current missions (ALOS, ALOS-2, Sentinel-1, and TerraSAR-X), processed using a combination of SBAS and persistent scatterer interferometry, to build a picture of subsidence in these regions with high spatial and temporal resolution. In some areas, we detect subsidence rates exceeding 10 cm/yr, posing a significant threat to populations along the coastline. The results are compared with optical satellite observations, GPS, and other ground-based measurements to provide an independent validation of the observations.

Coal mining usually causes drastic large deformation gradient in the earth’s surface in a short period, which leads to the formation of collapse basin or collapse pit. The drastic large deformation gradient caused by coal mining activity is non-linear in time, non-continual spatially, which is presented as a series of intensive interferometric fringes, making the image sampling rate unable to meet the requirement of Nyquist sampling principle, causing aliasing of phase, and debilitating the restoration ability of deformation phase in the phase unwrapping process.

    In this paper we propose using amplitude based offset tracking technique to monitor large deformation gradient in the subsidence center. To obtain the small deformation information around the basin and collapse pit, we use phase based Small Baseline Subset (SBAS) interferometry technique to monitor the small deformation , after which we integrate the results under the two methods to obtain complete monitoring results. We take Daliu Tower mining of Shanxi Province in China as an example, using TerraSAR-X satellite data of 13 scenes spanning from 21/11/2012 to 02/04/2013 to conduct deformation monitoring experiment. The experiment result shows that the proposed method has the capability to detect and measure the large deformation gradient, meanwhile, small deformation information is preserved. Moreover, by comparing the experiment monitoring results to continuous GPS measurements, we estimate that the accuracy of deformation monitor is about 8 cm in the range direction.

Keywords:  mining subsidence, subsidence monitoring, large deformation gradient, InSAR, offset tracking

Landslide monitoring is a crucial issue since its manifestation directly affects both the human life and the environment (loss of life, destruction of construction works, destruction of the natural and cultural heritage, etc.). Persistent Scatterer Interferometry (PSI) is an ideal technique for monitoring the earth's surface, since it is able to measure small-scale movements in the ground surface as well as to identify the landslide behavior, to examine the relationship with the triggering factors and to evaluate the effectiveness of containment measures. The technique utilizes corner reflectors for monitoring the deformation of ground due to their stability and their strong radar backscatter. In particular, Persistent Scatterer Interferometry (PSI) requires large stacks of SAR images acquired over the same area, aiming to create several single pair interference reference to a master image and thereby identify the displacement history of corner reflectors.

The specific work discusses the use of corner reflectors for monitoring an active landslide in the village of Analipsi, Ilia Prefecture, Greece. The first landslide event occurred in March 2014, creating mass movement in the rural areas of the village and influencing a house. The main and most critical landslide event took place a year later in March 2015, where two houses were completely destroyed, the main street of the village was collapsed and several cracks appeared across the wider area of the landslide’s crown. The complicated geology of the region and the presence of significant tectonic activity in combination with the intense and prolonged rains were the major factors which contributed to development of the landslides. In particular, the geological bedrock of the area consists of irregular alterations of banks of finely to medium grained brittle sandstones with interpolation of sandy and clayey marls, while the landslide materials include loose sandy-silty soils which have slipped over the underlying sandy marls.

In the specific area it implemented extensive geological research aiming at a detailed description of the phenomenon, its full understanding and its future monitoring. Initial steps were focused on geological mapping of the area and on the detailed mapping of landslide using Differential GNSS receivers. Additionally two geotechnical boreholes as well as a piezometer one were constructed in order to measure the displacement and the piezometric surface periodically. The installation of the corner reflectors and Sentinel-1 data processing through PSI, represent a further tool for landslide monitoring, understanding its geometry and its kinematics. Sentinel-1 data were selected due to mission’s short revisiting period, which fits with the purpose of the survey. Thus after the installation of corner reflectors, it was started the collection of large stacks of Sentinel-1 data, covering the area of Analipsi. These data were processed using appropriate software and the results are reported in the current paper.

The western branch of the East African Rift System, in Central Africa, is a region naturally prone to landslides due to the geomorphology of the area and to the occurrence of earthquakes and heavy rainfall events. The city of Bukavu (DR Congo) is located within the Rift, on the southern shore of Lake Kivu, in a tropical environment. Little is yet known about the current kinematics and the processes that drive large slow-moving landslides that continuously affect highly populated slopes of the city. Here we use InSAR time series to monitor ground deformations associated to these slope instabilities. Using 100 Cosmo SkyMed SAR images acquired between March 2015 and June 2016 with a mean revisiting time of 8 days in both ascending and descending orbits, we produced and compared displacement rate maps and ground deformation time series derived from two techniques: the classical Small Baseline Subset (SBAS) and the Multidimensional Small Baseline Subsets (MSBAS). With the MSBAS technique, the ground deformations observed by the two satellite orbits are jointly inverted, increasing the sampling frequency and allowing to measure higher ground deformation rates. The study focuses on the largest landslide in Bukavu (1.5 km²) that mostly moves in the E-W and vertical directions. InSAR results show that the landslide is divided into blocks that move with different velocities (up to 25 cm/yr), which is consistent with field observations and DGPS measurements taken at several benchmarks in the area during the same period.  The combination of this data with rainfalls, seismicity and field observations should help us to better understand the mechanisms (of natural and/or anthropogenic origins) that control the evolution of this landslide.

In Iran, arid and semi-arid climate and low rainfall lead to more consumption of underground water resources. Thus, in the last two decades; the groundwater overextraction has been led to the significant groundwater depletion and water level decline in most parts of Iran. Consequently, the large decline of water level has been led to the significant land subsidence in many aquifers throughout Iran. Qom plain, 1841 km² in the area, with the arid climate (160 mm, the annual mean rainfall) is located in the central part of Iran. Despite the low annual rainfall, the groundwater of the plain is continuously pumping by 559 production wells to be used for agricultural activities. This conducted to the significant groundwater decline in the plain. The observed data in the middle parts of the plain reveals the water table decline of 34.4 meters during the years 1973 to 2014 (i.e. the decline of 0.84 m/y in average). This significant decline led to land subsidence. The main goal of this study is achieving the zone of subsidence and its rate utilizing radar interferometry technique. Moreover, we aim to investigate the hazard of infrastructure resources which influenced by subsidence using Geospatial Information Systems.

The spatial coverage, suitable resolution and precision of Synthetic Aperture radar Interferometry (InSAR) technique makes it a prevailing technique to identify the rate and coverage of subsidence. In this study, we use ENVISAT ASAR images from 2003 to 2010, in descending orbits. The SLC images are processed via open source software StaMPS/MTI (Stanford Method of PS/Multi-Temporal InSAR) to produce interferograms. The interferograms are corrected for the phase signature due to the orbital separation (flattening) using precise Doris orbital data for ENVISAT satellite. The topographic phase contribution is evaluated from the 90m SRTM DEM (Shuttle Radar Topography Mission Digital Elevation Model). The remaining differential inteferogram phase can be associated to phase change contributions from deformation signal (here subsidence), tropospheric delay and orbital error and noise. The Medium-Resolution Imaging Spectrometer (MERIS) is used to estimate the magnitude of the atmospheric signals in interferograms. Once all interferograms are corrected, the PS time series analysis method will be employed to obtain the mean velocity map of the Qom plain region. Finally, vulnerability of infrastructures resources will be investigated using analysis in Geospatial Information System.

Urumqi city is located on the edge of Urumqi depression in northern slope of the eastern segment of the North-Tianshan Mountains. Coal is rich in Urumqi city with the coal reserves of ~100×10⁸ tons, so the city is known as "the city on the coal field". After decades of mining, ground collapse areas in Urumqi have reached dozens of square kilometers, and many collapse belts were formed in urban areas with a number of kilometers. Ground collapses not only destroy the ecological geological environment and threaten the surface construction facilities, but also make roads and urban planning blocked. The problem of ground collapse has seriously affected the economic development and social progress of Urumqi City, which has a negative impact on the normal life of the surrounding residents.

Urumqi mining area is located between the Urumqi River and Tiechanggou River, where Liudaowan mine, Weihuliang mine, Jiangou mine, Xiaohonggou mine, Dahonggou mine, and Tiechanggou mine distribute from west to east. In order to reduce the impact on the ecological environment and urban construction, since November 2014, the government of Urumqi begins to close coal mines, and plans to gradually shut down all coal mines in the area by the end of 2015. In the work, ascending and descending Sentinel-1 TOPSAR images from November 2014 and ascending ALOS-2 PALSAR-2 images from February 2015 are used to monitor ground subsidence based on InSAR time-series analysis technique. Our purposes are that: 1) understanding spatial-temporal characteristics of ground subsidence in Urumqi mining area; 2) preliminarily determining whether or not all coal mines have been closed.

Various natural phenomena has a significant impact on the quality of human life from the past. Investigation of this phenomenon, identifying the pattern and understanding how to prevent damages caused by them is necessary to provide comfort and improvement of the quality of man's life. One of these natural phenomena, phenomenon of earth surface movement and deformation such as the subsidence phenomenon.

Land subsidence is a worldwide phenomenon, where there is a sudden sinking or gradual downward settling of the earth's surface with little or no horizontal motion. This phenomenon is noticed to take place in many urban areas such as Tehran, the capital of Iran.

Tehran is facing the problem of the instability of terrain and structures against phenomena like sub-ways excavation, groundwater extraction, high-rate urbanization, railways construction, fault lines, and so on. Tehran is located at the southern foot of the Alborz Mountains. A large fault is located in this mountain range and there are several fault lines in the plains south of the city of Tehran. This city, virtually surrounded by faults, has suffered large earthquake disasters in cycles of approximately every 150 years. The North Tehran and Mosha faults situated towards the northern side of Greater Teheran and the Ray Fault on the southern limits of the city have the potential to generate MW = 7.2 and 6.7 respectively. At the same time, Tehran has experienced the highest urbanization process of any city in Iran in the recent years. Large influx of people into Tehran took place after 1979. Furthermore Tehran metro system consists of seven metro lines. Five lines (lines 1, 2, 3, 4 and 5) of the metro are operational while two lines (lines 6 and 7) are still under construction. The metro network is currently 170km long that carries more than 3 million passengers a day. Upon completion of all the lines, the system will reach a length of 430km.

Generally, alluvial basins of arid and semiarid zones are the places with excessive groundwater withdrawal, and also they have a high potential for land subsidence. Tehran Alluvium and Kahrizak Formation, dominating the central part of the Tehran plain, represent potential aquifers with good hydraulic conductivity. Therefore, decrease in water level have caused severe land subsidence in Tehran. Monitoring surface deformations related to this phenomenon can be used to assess the stability, and therefore safety of above ground infrastructure especially in residential areas.

In this study, the surface deformation measurements were obtained using the procedures implemented in SARPROZ. we applied the InSAR technique to detect and characterize surface deformation to find the exact relationships between subsidence and phenomena like sub-ways excavation, high-rate urbanization, and ground water exploitation. In this study, in addition to Tehran Basin, the urban areas was also investigated.

Due to atmospheric disturbance signal and high density of Persistent Scatterers (PSs) in urban areas (up to 700 PS/kmsq in urban areas), the persistent scattering interferometry (PSI) technique is chosen for this research. In PSI technique, the coherent radar reflectors of a certain area of interest are exploited for overcoming the difficulties related to conventional synthetic aperture radar (SAR) interferometry (namely, phase decorrelation and atmospheric effects), achieving millimeter accuracy in monitoring relative displacements of the targets. Also, atmospheric and orbital errors are essentially removed and linear and non-linear deformation patterns are identified.

To compute the LOS movements, 71 Sentinel-1A images in the IW mode, from November 2014 until October 2016, and due to keep track of the deformation in a long time series, 75 ENVISAT-ASAR images from 2003 until 2010 are used. Also we used both ascending and descending geometries to decompose the LOS movements into vertical and horizontal displacements.

According to the obtained results, the coincidence of the spatial pattern of the subsidence area with the cultivated area strongly suggests that the observed deformation in Tehran Basin is due to groundwater exploitation for use in agriculture and industry. Furthermore, in some areas close to the urban sub-way lines, and as well as in the region 22 due to construction related to the project “ A thousand and one city” earth settlement has been seen. The results depict a large vertical signal and a smaller horizontal signal, as expected for subsidence.

We validated the estimated rates and accuracies using GPS measurements and leveling method. The RMSE differences obtained from comparison to these methods lie within the expected range.

【Abstract】The over-exploitation of the groundwater in the North China Plain has caused severe  subsidence in this area, which  has compromised the sustainable development of China’s national economy. SBAS time-series analysis was  employed in this paper to research the Tianjin Plain. Information was gathered about the surface deformation of a large area of the Tianjin Plain based on comprehensive ASAR data. This paper first, analyzes the distribution of subsidence in Tianjin City on a macro-scale, then quantitatively studies the distribution  of  subsidence funnels, and finally analyzes the degree of coincidence of actual, recorded groundwater cones of depression and subsidence sites in the  Beichen District . The results show that: 1) The leading factor of subsidence in Tianjin City is the depletion of groundwater  caused by overuse, which directly leads to subsidence. 2) The center of subsidence in the area of over-exploitation approximately coincides with the groundwater funnel center and shows a tendency of slightly moving towards the northwest as a whole, due to the thickness and properties of the rock and soil, specifically the solidification speed of the soft soil layer is slower than that of hydraulic head changes of underground water after groundwater exploitation. 3) Elastic and inelastic water storage coefficients were inversed by combing InSAR and groundwater level data.This study lays a foundation for the control of groundwater exploitation and the establishment of a model of sustainable exploitation of groundwater in the future.

Late-stage continental rifting in Afar, Ethiopia, provides the opportunity to observe magmatic spreading centres analogous to that of slow-spreading mid-ocean ridges. Recent observations of a long-lived shallow axial magmatic system at the Erta Ale magmatic segment indicate that current thermal models for slow spreading centres, which would predict a deep, isolated magma chamber, may need revision. I use new Sentinel-1, and existing ERS and Envisat SAR data to observe magmatic spreading centres in Afar in order to determine the depths and geometries of these systems, using elastic models. Here, I present Sentinel-1 InSAR observations, using 2 descending tracks and 1 ascending track to cover all magmatic segments in the Afar region. Recent surface deformation for volcanoes in Afar, applied alongside constraints from GNSS, seismicity, petrology, and magnetotellurics, acquired during the Afar Rift Consortium, constrain current models of magmatic systems at spreading centres. I also analyse previous observations from spreading centres in Afar, the Main Ethiopian Rift, and Iceland, to establish whether the depth of magmatic systems in subaerial rift zones varies systematically with parameters such as spreading rate, crustal thickness, and magma injection rate. The results of this study have implications for the evolution of continental rift zones into segmented oceanic spreading centres, models of magmatic systems in rift zones, and geohazards in the Afar region.

Understanding the dynamics of landslides and predicting their occurrence is a common theme in the study of natural hazards. This paper demonstrates the monitoring of mass-movement caused by Europe’s largest coastal landslide complex, Black Ven, on the UNESCO designated Jurassic Coast World Heritage Site (JCWHS) in the south of the United Kingdom, using a coupled InSAR, airborne laser scanning (ALS) and Differential Global Navigation Satellite System (D-GNSS) approach.

Mass-movement on the JCWHS represents a “coastal conflict” between the delivery of fossils to the beach, by landsliding and mud flows etc, which are a draw for professional fossil hunters and tourists alike, and the need to protect the property and infrastructure on which visitation depends. These important fossil sites and the classic coastal geomorphologic features which feed them are also the main reasons for UNESCO World Heritage Site designation. Reliable and frequent monitoring of the site is therefore essential for providing insights into the dynamics of the system, especially the acceleration and/or migration of displacement, and any subsequent management decisions.

Interferometric SAR (InSAR) techniques have been applied to time series of Sentinel-1 (C-band) and ALOS-2 (L-band) to detect and quantify displacement. Both conventional InSAR and offset-tracking have been applied according to local ground conditions.

Repeat airborne laser scanning (ALS) overflights by the Environment Agency (England & Wales) have provided a time series of Digital Elevation Models (DSM). These have fulfilled two functions, namely 1) monitoring changes in surface volume over time; and 2) providing contemporary data for the removal of topographic phase in InSAR analysis, thereby improving accuracy.

D-GNSS provides field reference data for InSAR observations and results show success is variable, depending on magnitude of movement, wavelength of the sensor and temporal baseline.

The permanent scatterer (PS) technique is a powerful multitemporal tool in the context of InSAR. In the original PS analysis the interferometric phase is generated by referring all images to a common master acquisition. In the normal-baseline–temporal-baseline space, this configuration can be represented with a star graph, where each point (Node) indicates an image and each connection (link) indicates an interferogram.

The main objective of this study is to investigate how using different image graph will affect the PS-InSAR analysis results. In this framework three principal properties of image’s graph were considered: 1- Connectivity: A graph is connected whenever, for all distinct pairs of nodes, a linking path exists. This property is needed in order to unwrap the phase time series. 2- Number of links: the minimum number of links to guarantee connectivity is N − 1, and the number of links to make a complete graph (each node is connected with all others) is N (N − 1)/2. Whenever the number of links is> N − 1, the graph is redundant. 3- We assign a weight to each link in order to quantify its goodness and to compare different graphs. The optimum choice would be the complete graph. However, complete graph has massive computational cost, especially when the large number of images is available. Moreover, the links of a complete graph are not necessarily all informative. Due to the principal properties mentioned and spatial coherency as a weight to each link, optimum graph has been generated by adding some coherent links to MST (Minimum Spanning Tree). The inconvenience of an MST could be the limited lever arm of the baselines for the estimate. A possible general solution is to add a number of best links to the MST that maximizes the coherence. Then, Interferometric phase has been extracted by optimum (super-MST) graph and other ones like: Star, SBAS (Small Baseline Subset), Delaunay, MST and complete graph. PSInSAR procedure has been followed in 3 step: initial candidate points selection (PSC, partially coherent targets), estimation of unknown parameters through spatial network between candidate points, final estimation of unknowns after APS removal. A stack of 33 images acquired by Sentinel-1A from ascending orbit 28 during October 2014 to October 2016 was employed in this study. The data covering Tehran, capital of Iran that contains both urban and non-urban areas. Different graphs were compared in terms of the variances of the height, the deformation trend and the density of final PSs. The graph connectivity assures the temporal continuity of the deformation measurements and, thus, the possibility of unwrapping the phase time series.

Experimental results showed that the star graph allows the exploitation of large baselines (thus getting precise 3D PS positioning that is essential for estimating the atmospheric delay) but does not allow the exploitation of radar targets exhibiting PS behavior only on a sub-set of images. In SBAS graph due to small spatial and temporal baselines condition, some of the available images were not used and the resulting graph was often disconnected thus preventing the correct motion measurement without an available a priori model and hence the small baseline condition did not allow a precise 3D location in space. This prevent a good separation of elevation, motion and atmospheric phase components. Delaunay and complete graphs enable the PSInSAR to work with a small number of images, if compared with the requirements of the PS technique and Preliminary experiments provided a more accurate rate of deformation from a reduced set of images. In super-MST graph, by adding the most coherent interferograms not yet used, in spite of a limited growth of computational costs, a reasonable expected accuracy is reached.

Identification of areas that prone to subsidence and estimation of its rate plays an important role in the control management of this phenomenon. Differential interferometry radar technique (D-InSAR) with very high accuracy is one of the most suitable ways for identify and measure the rate of subsidence. In this study, to identify and measuring of subsidence in Najafabad Plain (central Iran) have been used differential radar interferometry techniques in the period 2003 to 2012. For this purpose, some pair images of time series was used from ASAR and PALSAR sensor in C-band and L-band radar in ascending and descending passages. The method will be used in this study is based on laboratory-Field surveys. For validation of technique will be used from survey data, maps of Land use and geology and data of observation wells in the region. Assessment results of two sensors will compared with together. As a result the maximum rate of annual subsidence average in the area will assessed per year. The results will compared with the highest of subsidence amount occurred in areas under cultivation and due to excess extraction of groundwater and subsidence of aquifer surface. Also rate of subsidence will be obtained for dropping of water table by relationship between subsidence and the changes of piezometer wells surface.

In this paper landslide creep deformation in Vanak city, center of Iran, will be investigated by using C-band and L-band InSAR measurements. A total of 20 SAR images acquired by ASAR and PALSAR satellite between 2003 and 2012, the period when GPS ground-based truth and ASAR data are available, will be processed using the Small Baseline Subset (SBAS) approach to produce time-series of displacement. Persistent Scatterers extracted by Small-Baseline method are homogenous and have high spatial density in the landslide area. However, tropospheric artifact in interferograms make it challenging to separate displacement pattern of the creeping landslides from the background noise. To reduce the effect of troposphere we calculated and eliminated the correlation between displacement and topography in each unwrapped interferogram before least-square inversions. InSAR time-series results from GPS ground-based truth and ASAR data indicate pattern and rate of motion on two reactivated landslides in the region, that it is called Vanak landslide, near the west of Semirom City. The ALOS time-series results will be compared with Envisat time-series results to investigate the performance of C-band and L-band radar data for landslide monitoring.

Ground Based Synthetic Aperture Radar (GBSAR) is a field based imaging system offering users enhanced capabilities in mapping topography and monitoring ground displacements. Funded by NERC, Newcastle University purchased a MetaSensing FastGBSAR in October 2015. The associated commercial software package uses a near-real-time persistent scatterer interferometry methodology. This technique is more suitable for artificial surfaces with sufficient strong back scatterers rather than natural terrain and the tool relies on meteorological data to correct atmospheric variations. This research therefore aims to improve GBSAR interferometry by developing an automatic processing chain to process FastGBSAR imagery for near-real-time geohazard monitoring purposes.

 

In most cases, FastGBSAR functions in continuous operation mode with a perpendicular zero-baseline between acquisitions. A new method, presently termed a near-real-time GBSAR zero-baseline subset algorithm, has been proposed for deformation monitoring applications. The first step of the methodology involves the detection of coherent pixels based on a redundant network of interferograms constructed from a small subset of images (typically 10-15). Within this image subset, “siblings” of each pixel are identified within a large window (the window size is usually set in the range of 25-40 pixels), based on similarity of time series amplitude characteristics. The coherence of every pixel over each interferogram is then computed based on the complex correlation of its “siblings”. A coherence threshold (typically 0.7) is defined to determine whether or not a pixel is coherent over an interferogram. As coherent pixels are changeable over different interferograms, an algorithm has been developed whereby a matrix is constructed via the coherence indicator for each pixel. Provided that the matrix has a full rank the pixel is considered as coherent both in space and in time. The detected coherent pixels can be treated as a basic library and regarded as persistently stable. For applications with a fast changing surface, it can also be treated as temporarily stable with periodic updating e.g. every 30 minutes to several hours.

 

A subset of pixels is selected by spatial gridding based on each having the best total coherence within a grid unit. These coherent pixel candidates are updated when a new image is added. A Delaunay triangulation of coherent pixel candidates is used for temporal and spatial phase unwrapping. Triangles in which the maximum arc length exceeds a certain threshold (typically 30-50 meters, which depends on applications) are removed in order to avoid unwrapping errors. The recursive temporal unwrapping is conducted by the use of Multiple Model Adaptive Estimation (MMAE) for GBSAR data. The MMAE uses a number of parallel Kalman filters, one for every arc and ambiguity set, each implementing different models. The spatial unwrapping is then performed using a minimum cost flow algorithm to ensure that the temporal solution is spatially consistent (i.e. the sum of the unwrapped double differenced phase within each triangle must be zero). Once a consistent ambiguity set is determined for each arc, all detected coherent pixels are integrated into the triangulation network by spatial unwrapping.

 

After unwrapping the phase at all coherent pixels, the atmospheric component and the phase difference resulting from displacement are separated for every interferometric phase observation. This is achieved by imposing a temporarily linear model on displacement velocity. The atmospheric delay and deformation inversion are estimated simultaneously in a linear least squares sense.

 

In this paper, a couple of applications of FastGBSAR will be demonstrated, including the Hollin Hill landslide, coastal erosion (Tynemouth) and dune movement (Changli). Further work will focus on extended applications with the proposed deformation monitoring method.

This paper is focused on the monitoring of subsidences and uplifts using Sentinel-1 (S-1) interferometric data and Persistent Scatterer Interferometry (PSI) and Differential Interferometric SAR (DInSAR) techniques. DInSAR and PSI have undergone a remarkable development in the last two decades. Most of the developments have been based on C-band data from the ERS-1/2, Envisat and Radarsat missions. A major step has been experienced in 2007, with the advent of very high-resolution X-band data (TerraSAR-X and CosmoSkyMed). A further significant improvement is now provided by the C-band sensor on-board the S-1A and S-1B satellites. S-1 has an enhanced data acquisition throughput with respect to previous sensors, which considerably increases its deformation monitoring potential. A key characteristic of the S-1 data is their higher coherence with respect to ERS and Envisat, which is due to the short revisiting time (6-days using the twin satellites S-1A and S-1B) that results in a reduced temporal decorrelation and the small orbital tube that results in less geometric decorrelation.

The authors have gained experience with the processing and analysis of different types of S-1 data. Several case studies have been studied so far, which include, among others, Mexico City, the salt lake of Atacama, Catalonia, Canary Islands, Southern Spain (Granada, Murcia, etc.). The paper will discuss different key aspects related to subsidence and uplift monitoring, including: (i) the PSI performances over wide areas, considering different land covers and land uses; (ii) the monitoring over local areas, discussing advantages and limitations; (iii) some validation results will be discussed; and (iv) a comparison with TerraSAR-X results will be addressed.

Conventional DInSAR techniques have been frequently used in the past for deformation mapping including the mapping of landslide activities. However, several difficulties arise when attempting to apply DInSAR in areas with steep slopes and rugged topography, high humidity and dense vegetation cover such as over the Three Gorges Region. In addition to these difficulties, it is shown that the maximum detectable displacement gradient of DInSAR can be exceeded in the case study area even when using the high resolution TerraSAR-X data. 

A sub-Pixel Offset Tracking approach (sPOT) is applied to monitor slow-moving landslides in densely vegetated and steep terrain. This approach is shown of being capable of measuring centimetre-level landslide rates by using natural scatterers in densely vegetated terrain in line with measurements derived from corner reflectors.

The potential and limitations of TSX-ST (TerraSAR-X Staring Spotlight) data on measuring surface deformation using DInSAR and offset tracking techniques are assessed through case studies in two landslide sites on the southern banks of the Yangtze River, in particular whether the improvement of the resolution of Staring Spotlight mode helps to address some of the issues that were encountered previously.

In addition, the TanDEM-X Coregistered Single look Slant range Complex (TDX CoSSC) data are employed to produce a 6 m resolution DEM. The impact of using different sources of DEMs is then assessed on deformation measurements via offset tracking and DInSAR.

Finally, the relationship between landslide occurrence and possible hydrological driving factors is assessed to infer possible landslide mechanisms. 

This work is partially supported by the CSC and UCL through a PhD studentship at UCL-MSSL.

Sinkholes and sinkhole-related land subsidence constitute a severe geo-hazard along the Dead Sea in Israel and Jordan, affecting both human activities and infrastructure. In order to discriminate between potential subsidence triggers (dissolution, viscoelastic creep, consolidation) and to constrain some of the mechanical properties of the subsurface granular sediments, we examine a 4-year long subsidence record using high-resolution InSAR measurements from the COSMO SkyMed satellites. In particular we study: (a) sinkhole precursory subsidence, which show gradual acceleration before (and in places, also after) sinkhole collapse; (b) Land subsidence in response to surface loading in the mudflats environment, which is characterized by an initial rise and a quasi-exponential decay; and (c) Subsidence following flash-flood events, which is characterized by an abrupt increase immediately after the floods due to enhanced salt dissolution, and a quasi-exponential decay thereafter. Precursory subsidence duration correlates with the sediment properties, in agreement with previous numerical simulations, and can thus be used to constrain sediment viscosity. For the loading experiments in the mudflats, we explore consolidation and viscoelastic creep as possible subsidence mechanisms. Quasi-exponential subsidence decay after flash-floods can be explained by: (a) decay of salt dissolution rate due to an exponential drop of the groundwater hydraulic head after a flash flood; (b) Viscoelastic creep; (c) A combination of dissolution and creep. The Kelvin-Voigt creep model can explain the entire observed subsidence decay pattern, constraining the viscosity and elastic modulus of the consolidated gravel to 10¹⁵ - 10¹⁶ Pa s and ~175 MPa, respectively. However, if the elastic moduli are limited to values reported in previous studies (600-4700 MPa), only 10-30% of the subsidence can be explained by viscoelastic creep, implying that more than 70% of the post-flood subsidence decay should be attributed to decreasing dissolution rates due to the observed exponential drop of groundwater head. The viscosity values obtained by our calculations agree well with numerical simulations of sinkhole formation along the Dead Sea, whereas the elastic moduli are generally on the lower end of previous estimates. 

The combination of ascending and descending Persistent Scatterers Interferometric (PSI) data by means of resampling and/or spatial interpolation, separately for each Synthetic Aperture Radar (SAR) geometry, is a commonly followed procedure, limited though by the reduced spatial coverage and the introduced uncertainties from multiple rasterisation steps. Herein, an alternative approach is proposed for combining different PSI line-of-sight (LOS) observables in the vector domain, based on the geographic proximity of PS point targets. An alternative procedure is presented herein, using PS data in their initial vector format and by exploiting vector manipulation capabilities as well as the geostatistical modules already available in many GIS software packages. In the proposed post-processing scheme all necessary analysis steps are performed by means of attributes transfer and calculations between features geodatabases, prior to any rasterisation. By increasing the number of input point vectors during subsequent interpolation, the overall error budget coming from spatial interpolation is being reduced. The increase of output surface resolution together with the reduction of interpolation error budget for the combined PSI results is of significant importance during modelling of ground motion or integration of PSI measurements with GNSS observations. The benefit of working with vector data, instead of raster layers, especially in saving storage space should be more pronounced when large stacks of PSI or wide area processing results are involved. The proposed approach are applicable to any PSI dataset independently of the processing scheme. The advantages of the proposed vector-based approach compared to the commonly used grid-based procedure is being demonstrated using real data.

Persistent scatterer interferometry (PSI) [1,2] is a SAR-based technique for the measurement of surface deformations that may have been caused by anthropogenic or natural geophysical processes such as earthquakes, landslides, mining, construction, etc. PS are by definition single scatterers i.e. they comprise backscatter from a single point-like source. In suburban or natural terrains, such as the alps and the adjoining valleys, generally there is a low prevalence of targets exhibiting point-like behavior over long time spans. Moreover, layover of scatterers on mountainsides over those in the valleys occur frequently. Therefore, the deformation sampling based only on PS identification may remain inadequate, especially for a small-scale analysis. In our work we explore the combined use of PSI and SAR tomography [3,4,5,6,9] to resolve such critical layovers in alpine areas i.e. separating the different scattering centers, and thereby improving the coverage. At the same time, since the objective is to measure deformation, we incorporate advanced differential tomographic techniques [10] that allow us to estimate an average deformation velocity for the individual scatterers besides their location in the elevation (perpendicular to the line of sight (PLOS) axis).

We perform experiments with an interferometric data stack comprising of 37 Cosmo-Skymed (CSK) stripmap scenes acquired over Mattervalley in the Swiss Alps, which has many active landslides and rockfalls [11]. The acquisitions are spread over a time span of around 6 years. We use the interferometric point analysis (IPTA) toolbox for the PSI processing. An estimate of the atmospheric phase screen (APS) is iteratively obtained. In case of alpine areas, the APS estimation is often more involved. The local atmospheric conditions and propagation paths through the troposphere may strongly vary spatially due to the extremely rugged topography which can change by even more than a km between the valley floor and the mountain top.

Prior to tomographic inversion, the APS isolated in the PSI solution is extrapolated over the entire scene and compensated for in the differential phases. We then apply a 2D maximization of a single-look beamforming-based merit function for the simultaneous estimation of the unknown parameters viz. the elevation and linear deformation velocity of the dominant scatterer. A second maximization is applied to look for the parameters for a potential second scatterer. A generalized likelihood ratio test (GLRT)-based detection strategy is used. The extent of the possible values of these unknowns determines the overall 2D parameter space in which the solution is searched. In our case, we expect layover of scatterers which may be separated by as much as a few kilometres (along the elevation axis). An extensive parameter space is needed to resolve such layovers. Therefore, the search for the solution becomes computationally far more expensive compared to similar computations in case of urban areas (where the elevation extent is generally not more than a few hundred meters). We address this problem as follows. A fine DEM in map coordinates is transformed into range-azimuth coordinates using a lookup table containing the shits between the coordinate systems. For each bin in the range-azimuth coordinate system, we use the lookup table to evaluate if it has a one-to-many correspondence with bins in map coordinates. In case of such correspondences, the multiple DEM values getting transformed from map geometry into the same range-azimuth bin allow us to estimate an approximate value for the height/elevation difference for the scatterers in layover. For wider differences, we search for the solution in a correspondingly wider parameter space. In this way, the parameter space is adapted for each range-azimuth bin in the layover-affected regions.

Fig. 1 shows the SAR intensity image, DEM/DTM and an optical orthophoto of the observed area. Layover cast by the alps is clearly visible in the SAR image. Fig. 2 shows the distributions of the spatial and temporal baselines of the interferometric stack. The baselines are quite irregular. Fig. 3 shows the PS obtained in an initial PSI solution. The color scheme represents the estimated linear deformation velocity. Fig. 4 shows the scatterers detected with a subsequent beamforming-based differential tomographic inversion, and projected in 3D in Google Earth; the colors correspond to the estimated deformation velocity.

References

[1] A. Ferretti, C. Prati, and F. Rocca, “Permanent scatterers in SAR interferometry,” IEEE Trans. on Geosc. and Remote Sens., vol. 39, no. 1,pp. 8–20, 2001.

[2] C. Werner, U. Wegmüller, T. Strozzi, and A. Wiesmann, “Interferometric point target analysis for deformation mapping,” in Proc. IEEE Int. Geosci. Remote Sens. Symp., 2003, pp. 4362–4364.

[3] O. Frey and E. Meier, “3-D time-domain SAR imaging of a forest using airborne multibaseline data at L-and P-bands,” IEEE Trans. on Geosc. and Remote Sens., vol. 49, no. 10, pp. 3660–3664, 2011.

[4] D. Reale, G. Fornaro, A. Pauciullo, X. Zhu, and R. Bamler, “Tomographic imaging and monitoring of buildings with very high resolution SAR data,” IEEE Geosci. Remote Sens. Lett., vol. 8, no. 4, pp. 661–665, Jul. 2011.

[5] M. Siddique, I. Hajnsek, U. Wegmüller, and O. Frey, “Towards the integration of SAR tomography and PSI for improved deformation assessment in urban areas,” in FRINGE Workshop, 2015.

[6] G. Fornaro, A. Pauciullo, D. Reale, and S. Verde, “Multilook SAR tomography for 3-D reconstruction and monitoring of single structures applied to COSMO-SKYMED data,” IEEE J. Sel. Topics Appl. Earth Observ. in Remote Sens, vol. 7, no. 7, pp. 2776–2785, Jul. 2014.

[7] G. Fornaro, S. Verde, D. Reale, and A. Pauciullo, “CAESAR: An approach based on covariance matrix decomposition to improve multibaseline-multitemporal interferometric SAR processing,” IEEE Trans. on Geosc. and Remote Sens., vol. 53, no. 4, pp. 2050–2065,Apr. 2015.

[8] A. Ferretti, A. Fumagalli, F. Novali, C. Prati, F. Rocca, and A. Rucci, “A new algorithm for processing interferometric data-stacks: SqueeSAR,”  IEEE Trans. on Geosc. and Remote Sens., 49(9), pp. 3460-3470, 2011.

[9] M. Siddique,  U. Wegmüller, I. Hajnsek, and O. Frey, “Single-Look SAR tomography as an add-on to PSI for improved deformation analysis in urban areas,” IEEE Trans. on Geosc. and Remote Sens. 54(10), pp. 6119-6137, 2016.

[10] F. Lombardini, “Differential tomography: A new framework for SAR interferometry,”  IEEE Trans. on Geosc. and Remote Sens., 43(1), pp. 37-44, 2005.

[11] T. Strozzi, H. Raetzo, U. Wegmüller, J. Papke, R. Caduff, C. Werner and A. Wiesmann, “Satellite and terrestrial radar interferometry for the measurement of slope deformation,” Engineering Geology for Society and Territory, vol. 5, 2014.

 

We present surface displacement results at the Geysers Geothermal Field in northern California derived from Interferometric Synthetic Aperture Radar (InSAR). Ground deformation at this site results from the effects of fluid extraction for power generation and injection of recycled wastewater back into the reservoir. A total of 72 scenes from RADARSAT-2 Ultra-Fine mode stacks from opposing look directions were acquired from January 2013 through April 2016. For our analysis we use MDA’s Homogeneous Distributed Scatterer (HDS-InSAR) method, which generates higher spatial densities of measured coherent targets compared to point based methods by exploiting both persistent point and coherent distributed scatterers by adaptively multi-looking statistically homogenous pixel neighbourhoods which improves the spatial resolution of the resulting deformation maps. 

Results show significant spatial variations in the observed deformation signal over an elongated area of nearly 15 by 5 km. The deforming areas are in agreement with the locations of the geothermal power plants and injection wells, and these signals are in agreement between the ascending and descending stacks. Furthermore, LOS deformation estimates are decomposed to derive 2D (vertical + east-west) deformation maps over the area of interest. Results show a combination of vertical deformation (subsidence and uplift) with cumulative motion of -2 to 2 cm, and horizontal deformation over the same area of vertical movement. Time series histories reveal both, long term (over three years) and short term, temporal deformation fluctuations and agreement of deforming areas compared to previous InSAR studies.

 

Abstract:

 

The 2008 Wenchuan earthquake resulted in thousands of geo-hazards including landslides, and debris flows. In this paper, we will attempt to use time series InSAR techniques to monitor the activities of two post-seismic geohazards, namely the Daguangbo landslide and the Wenjiagou debris flow.

 

Caused by the 2008 Wenchuan earthquake, the Daguangbao landslide is one of the largest landslides in the world. The whole mountaintop collapsed in the 2008 earthquake, leading to a height change up to 500 meters. TanDEM-X data is used to generate a high resolution post-seismic DEM in this paper. The high gradients of topographic errors and the decorrelation caused by vegetation in mountainous areas make the processing of Tandem-X data challenging. To solve this, we propose a re-flattening iterative method to generate a post-seismic DEM. 15 Sentinel SAR images were acquired during March 2015 to March 2016. The time series displacements of the Daguangbao landslide are obtained using our advanced InSAR TS+AEM method with the Sentinel-1 SAR images and high-resolution post-seismic TanDEM-X DEM. Four active zones are observed with a maximum displacement rate of 8 cm/year, suggesting that this landslide is still active even 8 years after the earthquake.

 

The Wenjiagou debris flow was the second largest landslide in the Wenchuan earthquake. The volume of the deposits is determined by comparing the post-seismic DEM derived from tandem-x data and the pre-seismic SRTM. In this site, a complex project has been set up to prevent potential debris flows by separating water and mud. Our InSAR results from Sentinel-1 observations (2015~2016) suggest that the source area of this debris flow is still active due to steep topography and rainfalls, whilst the project site and its downstream are stable. The impacts of this debris flow prevention project are to be discussed.

Copahue-Caviahue volcanic complex is located in the province of Neuquen, Argentina, at 37.5°S, 71.1°W. It is constituted by the Caviahue caldera and the Copahue volcano edifice at the border between Argentina and Chile. Copahue volcano is one of the most active volcanoes in Argentina. Historical eruptions have been reported on 1992, 1995 and 2000. Two villages are situated at 5 and 9 km from the crater, increasing risk level. The active crater host an acidic hot lake with ph<1 and temperatures up to 60°.

Previous works addressing crustal deformation in this area were performed by using ENVISAT, ERS and COSMO-Skymed images acquired between 2002 and 2013 [1]. They show a deflation process centered at the edifice’s northern flank, which acted between 2004 and 2010 at a rate of 2 cm/year. Inverse modeling of these results placed the deformation source at 4 km depth below the volcano, suggesting a relationship with the oscillations of the brittle-ductile boundary where leakage of brines and steams would produce depressurization of the system.

After Chilean 8.8 Mw earthquake on February 27 2010, several changes where observed, mainly related with an increment in local seismicity inside Caviahue caldera. Furthermore, two eruptions produced during 2013 and 2014 [2]. SBAS-DinSAR time series show that, since 2011, the volcano inflates at a rate of around 7 cm/year until at least middle 2014. Inflation pattern is located exactly in the same region where deflation was previously detected [3] [4].

In this work we analyze the suitability of ALOS2/PALSAR data for characterizing deformation related with the Copahue volcano area. Acquisition frequency in WBD (ScanSAR mode) is about one scene per month, which makes it potentially useful for volcano deformation characterization via time series processing.

We processed a dataset composed of scenes acquired in WBD (ScanSAR) mode between November 2014 and November 2016. We were not able to compute coherent interferograms between scenes acquired during 2014 and scenes acquired after February 2015. However, interferograms between scenes acquired after May 2015 show very good degree of interferometric correlation.

Results show that, after the uplift phase between 2012 and 2015, the system seems to stabilize and no clear deformation signal associated with the volcano is detectable.

 [1] María Laura Velez, Pablo Euillades, Alberto Caselli, Mauro Blanco, Jose Martínez Díaz. Deformation of Copahue volcano: Inversion of InSAR data using a genetic algorithm. Journal of Volcanology and Geothermal Research 202 (2011) 117–126.

[2] Simthsonian Institution – Global Volcanism Program. http://volcano.si.edu/volcano.cfm?vn=357090. Accesed on October 16 2015.

[3] M.L. Velez, P. Euillades, M. Blanco and L. Euillades. Ground Deformation Between 2002 and 2013 from InSAR Observations. Copahue Volcano, Active Volcanoes of the World, DOI 10.1007/978-3-662-48005-2_8. Springer-Verlag Berlin Heidelberg 2016.

[4] Rayner De Ruyt, Pablo Euillades, Leonardo Euillades. Identificación del Campo de Deformación Cortical del Complejo Volcánico Copahue (CVCC), mediante el Algoritmo DinSAR-SBAS. XXIII Jornadas de Jóvenes Investigadores Grupo Montevideo. 2015.

Simultaneous land subsidence and sea level rise pose great economic and social risks in the river deltas all over the world. The Pearl River Delta (PRD) is one of the most important economic regions with the highest population densities in China. Population and economy has increased dramatically during the last 30 years, as a result, considerable land subsidence occurs here together with sea level rise posing the area to risk of flooding and building damage. Very limited ground measurements have been conducted in this region yet, while satellite-based InSAR data are ideal for regional investigation of land subsidence. Using Envisat data, our study reveals apparent coastal subsidence with a mean velocity of about 2.5 mm/yr in Shenzhen and Dongguan, two megacities in the PRD region. We then evaluate the feasibility of measuring land subsidence in the PRD region using data acquired by the ALOS-1/2, COSMO-SkyMed, TerraSAR-X and Sentinel-1 satellites. Based on a multi-track time series analysis approach which assumes the vertical deformation is negligible, we combine radar imagery from multiple sources to estimate land subsidence in the region. The results give much higher accuracy and temporal resolution than that is able to be obtained from the traditional approach.

Landslides and periglacial terrain motion processes such as debris-covered glaciers, rock glaciers, push-moraines and solifluction slopes are very common in the Swiss Alps, particularly at high altitudes. With interferometric data from ERS-1/2, JERS-1, ENVISAT, PALSAR-1, Radarsat-2, TerraSAR-X and Cosmo-SkyMED more than 1500 single locations were identified in the past years and organized in a landslide inventory. The landslide inventory marks the location of ground movements, the process type and additionally there are possibilities to add more information on the potential impact on infrastructures. The determination of the objects contained in the landslide inventory is done using a combination of differential SAR interferometry (InSAR), optical images and interferometric point target analysis for slow landslides with only few millimeters per year deformation rates. Knowledge on the exact location and the deformation behavior of landslides during time is important for the hazard assessment and the risk analysis of areas potentially affected by the processes. The potential damage to infrastructures can be either direct by the process itself due to terrain motion or indirect by secondary processes, such as hydrological processes (debris-/mudflows) triggered by the landslides. The creation of a landslide inventory is very time consuming, because processes have to be identified manually by an operator. But, since most of the processes are stable in location, we propose a temporal tracking of the deformation rates of selected objects in the landslide inventory taking advantage of the high temporal resolution of Sentinel-1 A/B.
Recently, we observed over the Swiss Alps that the velocity of several periglacial terrain movements increased from centimeters per year in the 1990's up to several centimeters per day. Even though precise quantification of deformation rates of several centimeters is no more possible with InSAR,  the unique spatio-temporal coverage of Sentinel-1A/B offers a very good opportunity to monitor changes in the displacement rate at an early stage of the potential acceleration. In addition, there is also value in continuing the observation of fast processes because changes in location and extent and possibly slower movements can still be quantified. The key to our processing strategy lies in the use of the landslide inventory as a-priori information for the monitoring of single process areas through time. In our processing work flow we use Sentinel-1A and 1B terrain corrected interferograms of 6, 12, 24, 48, 96 and 360 days and compare the extent and velocity of the motion in comparable times throughout the year. Using different time intervals, we can track motion with velocities of ~1 cm/a on the lower end and ~10 m/a on the upper end. Using flags in the landslide inventory that point to a regional stable reference point, the interferograms are cropped to the extent of the landslide of interest and phase-normalized to the corresponding stable reference to minimize atmospheric phase shifts in the extracted interferograms. Then, a comparison of consecutive interferograms and the interferograms juxtaposed to the same intervals but 360 days apart, quickly reveals if the deformation behavior changed significantly over time. With this approach, seasonal changes that are common especially for periglacial processes are compared accordingly and do not lead to the misinterpretation of exceptional increase in velocity. In addition, not only changes in the velocity of the processes but as well changes in the spatial extent can be monitored. The interferograms are terrain geocoded with 5 m ground resolution using a terrain model of initially 2 m spatial resolution (SwissAlti3D), which allows detecting landslides with a width of only 30 m and areal changes as small as 5 to 10 m.
With our approach, fast landslides and periglacial terrain movements are traceable trough time and in a relatively quick way. From the moment of the Sentinel-1 SLCs data availability, the processing chain will add the scene to the stack and calculate the interferograms according to our predefined scheme. The 6 day revisit time of the Sentinel-1A/B constellation gives the opportunity for differential interferometry with a high temporal resolution. Additionally, in case of overlaps in adjacent interferometric wide swath (IW) tracks, the observation time for landslide objects is shortened again to 3 days, although with different observation geometry. When those observations are treated accordingly, the detection of changes in rates and extent of slopes deformations can be done in an exceptional short time, which is of high value for hazard and finally emergency management. However, the same limitations are valid as for every InSAR investigation. In particular, seasonal snow cover at high altitudes and increased vegetation growth at low altitudes sometimes hinder the formation of coherent interferograms.

Bhutan is a relatively small Himalayan country, landlocked between India and Tibet (between 88° and 92° east and 26° and 28° north) and characterised by a very rugged terrain and high topographic gradients, with elevations ranging from around 100 to 7500 m a.s.l.. The steep terrain, as well as geological and geomorphological settings, make this country highly susceptible to landslides phenomena of different size and typology. Given the low population density, secondary hazards related to landslides, such as landslide damming and floods, are particularly important for hazard assessment considerations, since their effects can propagate several kilometres downstream.

We here present the results obtained to date in the framework of a large research effort aimed at generating a broad picture of slope instabilities in north-western Bhutan. In this context, we exploit Synthetic Aperture Radar Differential Interferometry (DInSAR) techniques for the identification of formerly unknown landslides in north-western Bhutan and to assess their activity. We focus on the potential and limitations of DInSAR techniques for landslide identification on a regional scale across a high mountain terrain with little initial information. We generated displacement velocity maps and time series through the Small Baseline Subset (SBAS) processing approach of available ENVISAT ASAR and ALOS PALSAR acquisitions for the area of interest. These acquisitions cover respectively the periods between 2003 and 2010 and between 2006 and 2011, and thus provide valuable information for a back analysis of ground movements in the region. Furthermore, we combine the results obtained from the DInSAR analyses with a preliminary inventory of landslides and field observations. The initial inventory was generated through the analysis of optical images and a high resolution Digital Surface Model (ALOS World DEM, 5m GSD). The inventory was initially composed of around 900 polygons, including rock slides, rock avalanches deposits and deep seated gravitational slope deformations. A field campaign carried out in October 2016 allowed to refine this preliminary catalogue of slope instabilities, by jointly interpreting data collected in the field with respect to the results obtained with the DInSAR processing.

This work stems from an evident lack of information and knowledge regarding landslides distribution in north-western Bhutan and from the absence of a landslide hazard assessment. This dearth of information on such hazardous phenomena is related to the intrinsic inaccessibility of the terrain, to the current lack of ground based data and, in part, also to existing travel restrictions.

The StaMPS/MTI is a non-commercial persistent scatterers interferometric (PSI) software package, which is widely used by the scientific community for estimation of slow geodynamic or geomorphologic surface changes. We present here a new package, DAISY, which takes results from StaMPS for overlapping (in time and space) ascending (ASC) and descending (DES) image series. Integrating the ascending and descending LOS velocities we can get direct information on the nearly vertical and east-west velocities, which together with different thematic layers of geoinformation systems (GIS) can help the experts to properly investigate and interpret the derived results.

In the first processing step ASC PSs are selected, that have at least one DES PS within a specified chosen distance, and vice versa. In the second step those clusters are selected where at least one ASC and one DSC PS can be found within the chosen radius. In these clusters imaginary dominant scatterers (DS) are estimated so their sum of weighted distance squares with respect to ASC and DES PSs are minimized. The ASC and DES LOS velocities of DS are estimated as weighted means using the distances from the DSs. In the third step there are two possibilities. Since the velocities are referenced to the mean of cropped images, it may be reasonable to select those DSs which velocities can be treated as zero values (if any) in GIS environment. Another possibility is to specify one reference area, which mean value is subtracted from the DSs data. The last step is the integration of ASC and DES LOS velocities to derive the two unambiguous velocity components in the observation plain and the nearly vertical and east-west velocities, which may be biased by the unknown north components. The selected PSs and DSs data can be uploaded into the GIS software, and the derived quantities as attribute data can be visualised.

Based on the accepted ESA scientific project proposal (30142) we have 23 ascending and 32 descending ENVISAT raw images covering the Carpathian bend interior and the volcanic area in Romania. We present our first results of DAISY applied to this region in an ArcGIS environment.

The presented study consisted in processing ALOS-2 SM1 images by sub-pixel correlation and conventional differential SAR interferometry in order to assess the potential of these new data in the context of landslide deformation mapping. We selected the test site on la Réunion Island because of the presence of the well known Hell-bourg landslide, studied in the past by Remote Sensing methods, using both X, C and L-band interferometry.

The study showed that even with a reduced data set, consisting of 7 images with time spans up to 120 days, ALOS-2 PALSAR data represent a unique tool for landslide displacement monitoring in the highly vegetated context of La Réunion Island. Both the L-band and the high spatial resolution (1.8m) allow better performances in the motion measurement than other SAR data such as C-band SAR,  medium resolution L-band data  or high resolution X-band with weeks  time-span). 

We showed that the total deformation occurred between January 2016 and august 2014 reached up to 1.5m in azimuth and 70cm in slant range. The motion presents partitioning in different regimes. From the offset results, we can clearly delineate the trace of the scarps and the sectors with different velocity regimes. Time Series suggested us that significant motion occurred on a relatively short period between November 2014 and end January 2015. We suggest that this could be related to heavy rains occurring during the wet season.

The use of L-band (or larger wavelength when available) with the higher possible resolution therefore appears as the more adapted Space-borne Earth Observation tools for monitoring of very slow to slow landslides in a tropical semi-vegetated context. In this perspective the current ALOS-2 has obvious interest, provided that we access to at least bimestrial acquisitions.

Displacement maps from Multi-Temporal InSAR (MTI) techniques are usually noisy and fragmented. Thresholding on ensemble coherence is a common practice for identifying the surface scatterers that are less affected by decorrelation noise during post-processing and visualisation of the final results. The parameters of velocity, height and others, sought as the ultimate MTI estimates, are commonly considered reliable when their ensemble coherence is exceeding a certain threshold of, e.g. 0.7, and reaches the value of 1. Visually inspecting sets of highly coherent points, usually only the eyes of InSAR experts are then searching for groups of scatters in which deformation occurs and evaluate those possibly dangerous.

Simple selection of the points with coherence greater than a specific value is, however, challenged by the presence of spatial dependence among observations. If the discrepancies in the areas of moderate coherence share similar behaviour, it appears important to take into account their spatial correlation for correct inference. Low coherence areas thus could serve as clear indicators of measurement noise or imperfections in mathematical models. Once exhibiting properties of statistical similarity, they allow for detection of observations that could be considered as outliers and trimmed from the dataset.

Thresholding on ensemble coherence value might cause loss of information over areas undergoing more complex deformation scenarios. If the measurement noise of any source is present in low-coherent areas, imperfections in mathematical models should be addressed in professional expertise first. Having information about the location of defective areas, the expert user can steer his focus in order to retrieve real deformation profiles. By applying data mining strategies, it is possible to support routine procedures and extract additional information contained in the datasets.

What more, visual inspection of millions of scatters with wide area coverage capabilities of Sentinel-1 is not more sufficient in providing useful insights into the actual nature of undergoing processes. Nation-wide monitoring initiatives are making this task even more complicated.

With shorter revisit intervals of Sentinel-1 it is also of interest to reconsider, more closely, the practice of imposing simple threshold on ensemble coherence value and assess its full informative character. Coherence itself has been recognized as a valuable parameter in range of thematic mapping applications such as change detection.

Although, lot of advances have been achieved in exploiting low or partially coherent targets, all effort in evaluating higher level data products often remains in the hands of end-users, causing common concerns about the reliability of InSAR results by simply looking at the locations with extreme velocities. In this work, we are presenting different case studies from applying an approach based on renowned data mining and statistical procedures for mitigating the impact of outlying observations in final results.

 

Transport networks across Europe face significant challenges in monitoring and predicting ground deformation along their transport infrastructure.  Incidents related to landslides and subsidence on road and rail transport systems can cause significant disruption, particularly during winter periods.  Therefore, it is in the interest of owners and operators of transport infrastructure to understand and manage their exposure to geological hazards to minimise impact.  Current monitoring practices provide reactive rather than proactive information on landslide and ground deformation events across a transport network, and primarily consist of in-situ sensor technologies and site visits.

LiveLand, an ESA ARTES 20 IAP funded development project led by CGG, aims to assist the transport networks initially in Scotland but with the potential to expand further into Europe.  The service will provide improved intelligence to facilitate the proactive management of landslide and ground deformation events and support transport owners and operators with their hazard and asset management systems. Initial user discussions have identified a requirement for geohazard information at regional and local scales along specific routes and sections of rail networks.  As such, LiveLand will demonstrate an integrated approach to monitoring transport networks at a range of scales using a three tiered system of information provided by experts within their field: earth observation satellites; geology; weather forecast modelling technologies; and in-situ low cost GNSS units.

Throughout 2017 LiveLand demonstrators will be available to the rail and road network participants, for review and integration into their systems.  During this period a detailed analysis and validation of the service will be undertaken, with the key aim to ensure that the end user knowledge of geohazards and the level of confidence in the probability of landslide occurrence along transport networks are significantly improved.  LiveLand will demonstrate the benefits of an integrated approach and provide an easily accessible geohazard information service to support an opportunity identified within the transport industry.

The long term societal and economic benefits of reduced disruptions and improved commercial performance of the transport network will have an impact not only on the transport companies themselves, in terms of reduced fines and closure of the network, but also on the day to day lives of the general public.

In order to integrate and homogenize the big amount of data derived by several acquisition platform (i.e., remote sensing, geological geophysical and in situ measurements) we propose an innovative approach so called: Chain Rule Optimization Procedure (C.R.O.P.). In detail the proposed methodology is an optimization tools developed in a numerical environment able to simulate the natural phenomena in a multiphysics context; this approach allows taking in account the complex physical interactions that occurred during the evolution of a natural system. In this context, the development of physically-based model is made possible because each optimization steps of the chain represents the input of the subsequent step (figure 1).

Following this approach, the achieved best-fit model represents a calibrate solution that takes into account the impact of the most relevant physical parameters involved in the observed phenomena. More Specifically, to demonstrate the capability and versatility of our proposed methodological approach, we select as case of study the Ischia Island (Southern Italy), a volcanic area characterized by very high values of heat flow, a high temperature geothermal system and relevant active long term- ground deformation processes detected via C-BAND Multiorbit DInSAR data acquisition. Accordingly, we applied the multi-sensor SBAS approach to the whole archive of SAR data (about 300 images) collected over the Ischia Island from ascending and descending orbits during the 1992-2010 time interval by the ERS-1/2 and ENVISAT sensors.

In order to achieve a unitary physically based model of the active long-term ground deformation phenomena, we applied the proposed methodology starting from the collection and integration of the available geological and geophysical information acquired in the last decades by the scientific community. For the Ischia volcano case study, the first step is represented by the realization of the 3D geological model of the Ischia volcanic-hydrothermal system. More specifically, we build a 3D geological-structural and petrophysical model of the whole crust beneath the Ischia volcano by integrating geological and geophysical. This geological model represents the numerical domains for the subsequent numerical optimization procedure of the available temperature measurements (i.e., shallow and deep borehole geothermal measurements). At this stage, we also evaluate the amount of the conductive and/or convective thermal regime, in order to explain the complex status of the hydrothermal system of the island of Ischia. The 3D optimized thermal field results allow to reconstruct the 3D imaging of the B/D transition for a defined time window. Finally, a 3D fluid-dynamic model, in viscous flow approximation, is performed to highlight how the viscosity contrasts between the rocks of the ductile and brittle region modulate the long-term subsidence of the Ischia Island. This fluid-dynamic model is optimized by exploiting component of the ground deformation pattern detected via satellite multi-orbit C-Band SAR (Synthetic Aperture Radar) data acquired in the time interval 1992 - 2010.

Finally, we point out that the application of the C.R.O.P proposed methodology at Ischia volcanic island highlights that the driving forces that modulates the spatial and temporal evolution of 1992-2010 long-term subsidence phenomenon, detected via advanced DInSAR Interferometry, are controlled by the coupling effects of crust rheology, which are governed by existence of the thermally anomalous igneous intrusion and by the gravitational loading of the volcano edifice.

As one of the largest normal faulting events in the northern Tibetan Plateau, the 2008 Mw7.1 Yutian earthquake occurred in a snow-covered mountain area between the Ashikule basin and Guliya Ice Cap in the western Kunlun Mountains, also known as a stepover zone bounded by three major sinistral strike-slip faults (e.g. Altyn Tagh, Karakax and Longmu Gozha Co), followed by four Mw>5.3 earthquakes at a distance of ~100 km to the north-east of the mainshock. In this study, the coseismic rupture of the 2008 Mw7.1 Yutian earthquake was revisited using InSAR data from Envisat ASAR and ALOS PALSAR in ScanSAR mode. The explicit fault structures and the coseismic slip distribution of the mainshock determined in this study provide critical information on the mechanical behaviors of the local fault system. We also investigated its post-seismic motion in the following ~2.5 years after the 2008 mainshock using Envisat ASAR in ScanSAR mode and generated a time-series of afterslip based on the InSAR observations. In order to investigate the recent postseismic deformation patterns, we employed Sentinel-1A data with Interferometric Wide Swath mode collected since October 2014, which covered the postseismic motion of those four Mw>5.3 events between the year of 2011 and 2014. We finally discuss the implications of the coseismic and post-seismic phases of the 2014 Yutian earthquake on the local tectonic stress evolution by calculating the Coulomb stress changes caused by the 2008 event, which could offer new insights on how the failure of extensional stepover fault potentially triggers earthquakes along bounded strike-slip faults.

Although the traditional instruments used in structural monitoring e.g. total station and levelling can determine the ‘health’ of man-made structures, observation and results analysis using such methods is time-consuming, expensive and inefficient. The rapid development of InSAR techniques has motivated many researchers to investigate the potential of InSAR for automated remote monitoring of infrastructure stability. In this study, we attempt to use multi-platform SAR data to monitor the deformation of Mosul dam in Iraq.

Built in the 1980s with aims for irrigation, flood control, and hydropower, Mosul dam is one of the most important structures in Iraq. The dam is located in the north of Iraq within the Arabian tectonic plate which is relatively stable, but it is built on soft, dissoluble gypsum, anhydrite and karstified limestone bedrock. Water reservation started in 1984 and since then, soluble foundation problems have been observed due to the geological setting of the dam. Continuous monitoring of this dam is vital to ensure the safety of the people living downstream. The State Commission on Survey of Iraq has been monitoring the dam stability using a 3D terrestrial geodetic network every six months, and the leveling observations can be used to calculate the vertical displacements and mean velocity of the dam.

In this study, four SAR data sets (i.e. Envisat, Sentinel-1 Cosmo-SkyMed [CSK], and TerraSar-X), collected during the period between March 2003 and September 2016, were used to monitor and evaluate the stability of this dam. The GAMMA software was used to interferometrically process all the SAR data. Two constraints were applied to select the interferograms: (i) the perpendicular baseline should not exceed 400 m, and (ii) the time separation of the two SAR acquisitions should be less than 900 days.  These constraints were only applicable for CSK and Envisat data because Sentinel-1 and TerraSAR-X have good orbital control with a relatively short time span. In contrast to other data sets, an additional step with a spectral diversity method was introduced to deal with phase jump issues between subsequent bursts caused by the Doppler differences in two acquisitions of Sentinel-1 data. When the 30 m SRTM DEM was used to remove topographic phase, obvious phase residuals can be observed on CSK interferograms as a result of long perpendicular baselines and DEM errors. To address this issue, a DEM with a spatial resolution of 4 m and an RMS difference against ground truth of 2.1 m was produced using a pair of TanDEM-X images. Surface displacement time series and mean velocities were separated from Atmospheric Phase Screen (APS) and other noises using the in-house InSAR TS + AEM software.  

For the period between 2003 and 2010, only one ascending and one descending Envisat dataset are available. Under the assumption that there was no movement along the dam axis, a model was constructed to recover the 2D dam movements for this period: one component in the vertical direction and one horizontal component perpendicular to the dam axis. For the period between 2014 and 2016, 3D displacements were generated taking into account the different geometries of Sentinel-1, COSMO-SkyMed and TerraSAR-X.

Comparisons between InSAR and levelling displacements showed RMS differences of 3.7, 3.9, 4.2, 3.5, 4.1 and 4.1 mm for Envisat (ascending), Envisat (descending), COSMO-SkyMed (ascending), TerraSAR-X (descending), Sentinel-1 (ascending) and Sentinel-1(descending), respectively.

In northern China, the problems of ground subsidence induced by mining activities have caused widely attention in recent years. With large-scale of resources exploitation in some mining areas in the North China and the Northwest China, the natural ecological environment has been seriously destroyed at the same time, which might result in small regional-scale subsidence in surrounding regions, and even more serious cases in some environmental geological hazards, including soil erosion, ground fissure, mine collapse, and seismicity. Handan-Fengfeng Mining (HFM) area is located in the southern part of Hehei Province in China, which is one of the earliest coal exploitation region in China. With over five decades of coal exploitation, a large number of coal mine gob areas have been formed in this region after many years of mining. Due to the complex geological environmental conditions and the intensive coal mining activities in HFM, the geological hazards, such as ground fissure, sink, and deformation, can be easily triggered, especially after the heavy rainfall. According to an official investigation in 2016 [1], the ground fissure has extended to about 2.34 km in length, and the deformation area has increased to 12.5606 km² in HFM, caused by the underground mining development.

Interferometric synthetic aperture radar (InSAR) is an extremely powerful technique to measure the ground subsidence over a large area from the geophysical processed remotely. In monitoring of mining-induced subsidence, there are some problems may limit the application of InSAR, such as the diverse topography, the large-gradient ground subsidence, and the dense vegetation coverage. All of these limitations are possibly resulting in a decorrelation of the radar interferometric signals, and may lead to data misinterpretation. In this study, TerraSAR-X data with stripmap mode from February, 2014 to April, 2016 are applied to investigate and retrieve the deformation information over HFM, using temporarily coherent point InSAR (TCPInSAR) [2] associated with non-local means filtering [3]. As a comparison, Sentinel-1A data with IW (Interferometric Wide swath) mode from September, 2015 to April, 2016 are also processed. To validate the time series results from InSAR, near-monthly leveling measurements (eight times) over two of the mining areas (Jiulong and Wannian) were performed from April, 2015 to March, 2016, with a total number of 147 points. The main conclusion of this study can be summarized as follows: 1) Comparison with Goldstein filtering [4], the non-local means filtering is performed better to reduce noise and preserve details in mining regions with low coherence. 2) The total areas of deformation caused by the mining activities in HFM is approximately 18.96 km², which is 50.9% larger than the announcement from the report of government. 3) The results from Wannian indicate that the location of deformation in ground surface has some deviations with the actual coal exploitation location under the ground. The reason is due to the fact that the coal seams are inclined to the horizontal plane in this mining area. 4) The InSAR results from both TerraSAR-X and Sentinel-1A show a good agreement with the leveling measurements in the period with high temporal density of SAR images. While, the results in some periods with low temporal density of SAR images have different levels of underestimation, due to the issue of decorrelation caused by the large gradient deformation. That is to say, the temporal density of SAR images is an essential precondition for the active mining-induced deformation monitoring. With the recently launched Sentinel-1B, the revisit time of Sentinel-1 constellation has increased to 6 days, and the decorrelation caused by the large gradient deformation will be significantly improved.

References:

[1] Government Office of Fengfeng Mine, “Operation plan for prevention and controlling of geological hazards in Handan City in 2016,” Hebei: Government Information Opening Platform of Handan City [online]. Available: http://hdzfxxgkpt.hd.gov.cn:81/content.jsp?code=K01376451/2016-08401.

[2] Zhang, L., X. L. Ding, and Z. Lu, “Ground settlement monitoring based on temporarily coherent points between two SAR acquisitions,” ISPRS J. Photogram. Remote Sens., vol. 66, no. 1, pp. 146-152, 2011.

[3] C. Deledalle, L. Denis, F. Tupin, A. Reigber, and M. Jäger, “NL-SAR: A unified nonlocal framework for resolution-preserving (Pol)(In)SAR denoising,” IEEE Trans. Geosci. Remote Sens., vol. 53, no. 4, pp. 2021–2038, Apr. 2015.

[4] R. M. Goldstein and C. L. Werner, “Radar interferogram filtering for geophysical applications,” Geophys. Res. Lett., vol. 25, no. 21, pp. 4035–4038, 1998.

Coastal areas are often more densely populated than the hinterland and exhibit higher rates of population growth and urbanization. Taking into account that the vulnerability of these areas is at the same time increasing due to sea-level rise and coastal hazards such as storm surges or extreme rainfall events accompanied with floods, the importance of safety structures such as dikes is increasing as well. Hence, a spatial distributed dike monitoring should be part of a sustainable adaptation strategy. SAR- interferometry could be an essential tool to ensure this kind of required monitoring, due to an increasing amount of SAR-data from various satellites with high spatial and temporal coverage. Given this prospect, multi temporal Interferometry techniques will be very suitable monitoring techniques for dike structures to identify dike stability in terms of deformation with the accuracy of few millimeters. This procedure focuses on pixels that show a stable scattering behavior in a sequence of multiple SAR-scenes. In opposition to ground-measurements, the spatial coverage of this technique provides comparable results for different parts of dikes, where weak segments with particular high movements could be identified in advance. Furthermore the use of multiple scenes helps to overcome inaccuracies due to atmosphere or temporal decorrelation effects which is one of the major error sources in InSAR. The proposed monitoring tool allows preventing future dike crevasses and helping the risk reduction in high-populated areas. This paper focuses on the potential of the single master multi temporal interferometry technique for a spatial distributed dike monitoring of the coast in northern Germany. Therefore 56 Envisat-scenes were analyzed which have been acquired during different tidal conditions in the period between 2004 and 2010. In addition 44 Sentinel-1a scenes in the time frame between 2014 and 2016 were analyzed to detect the further deformation development of the dike structures on the one hand and to assess the advantages of the Sentinel mission for the dike monitoring. Both datasets cover an area of a sea shore dike including a flood regulation barrage. The results point out the potential for a spatial distributed dike monitoring tool. Although there are only small deformation rates in the whole area, the spatial heterogeneity is visible and a part of the dike with higher deformation rates could be detected. Due to its very dense temporal coverage the Sentinel-1a data shows hardly any data gaps which could affect the unwrapping results due to temporal decorrelation. Taking the temporal coverage of only 6 days over Europe in combination with Sentinel-1b into account, these huge datasets could contribute substantially to a spatially distributed dike monitoring and in this way reduce the risk of severe flood induced catastrophes in coastal areas. But even though the high temporal and spatial coverage of the Sentinel system guarantees some advantages for dike monitoring, the detection of Persistent Scattering pixels on natural surfaces (grass covered dikes) is still not satisfactorily elaborated. For the future a more sophisticated method to achieve more reliable results also on natural surfaces still has to be investigated.

Seasonal soil freezing affects approximately 51% of the land mass Northern Hemisphere, while 24% are located in the permafrost zone (Zhang et al., 1999). Stress and frost heaving caused by seasonal freezing and thawing, including the increase of the permafrost active layer, constitutes one of the main sources of physical damage affecting Northern infrastructure and transportation networks, including roads, railroads and airfields. While dedicated building techniques can be applied to mitigate for damage from frost heaving (e.g. Uhlmeyer et al., 2002), the expense involved typically limits their applicability especially in remote regions, necessitating robust monitoring capabilities for correct allocation of resources to e.g. road network maintenance.

Interferometric SAR techniques have been applied successfully in the past to monitor both large scale land deformation and displacement (Bustin et al., 2004; Tofani et al., 2013) and deformation of individual structures such as buildings and bridges, as well as roadways (Tarchi et al., 2012; Shan et al., 2012; Yu et al., 2013). With the recent increase of availability of suitable SAR sensors by both commercial and public operators (TerraSAR-X/TanDEM-X; Cosmo-SKYMED, RadarSat-2, Sentinel 1 A/B) capabilities for operational monitoring of Northern transportation networks is becoming a possibility.

We report on a dedicated experiment involving the assessment of different InSAR techniques at X-band in capturing the signatures caused by typical seasonal frost damage on a paved road. The experiment was carried on an airfield in Sodankylä, Finland (67.4N, 26.6E), during the autumn and early winter of 2016-2017.

The experiment involved creating discrete artificial damage structures (potholes, cracks and mounds, which are typical damages caused by frost heave on roads) of varying size on a 75 m stretch of an airport taxi road. Damages were created with an asphalt cutter and drilling chisel. Cracks were made parallel and perpendicular to the satellite track. The spatial resolution of the induced damages were on par with (the highest) resolution of satellite images; cracks were 5,10  and 30 cm wide and about 5 meter long, while dimensions of the potholes and mounds were 80 and 55cm, respectively. One TerraSAR-X Staring Spotlight image pair (spatial resolution of ~1m) and two COSMO-SkyMed Himage image pairs (spatial resolution of 5m) in VV-polarization were taken of the airport area prior and after the damage. Fixed corner reflectors were positioned in the airport to ensure geolocation accuracy as well as to monitor possible large scale rise or depression of the land surface. We report the detection capability of road frost damage of the two sensors, assessing in particular the effect of scale of damage on possible omission errors in the detection.

This work has been conducted in parallel with project preparations of the Sod5G project, in which the main objectives are the 5G-network, road weather services development and piloting environment for the dedicated special needs of multi-authority -, intelligent traffic and vehicle winter testing services. As the result of this project, accurate location-based road weather information and forecast service are implemented throughout the testing area road network and entity, delivered through advanced 5G-development network to all authorities, vehicles, drivers and the rest of the traffic actors in real time.

The Sodankylä pilot environment will be tailored for the needs of local vehicle winter testing, multi-authority co-operative services piloting as well as Finnish Meteorological Institute’s (FMI) intelligent transport/traffic systems and local road weather services development work. FMI’s satellite receiving systems are employed also for the estimation of the airport runway slipperiness.

 

Bustin, A., R. D. Hyndman, A. Lambert, J. Ristau, J. He, H. Dragert, and M. Van der Kooij, 2004. Fault Parameters of the Nisqually Earthquake Determined from Moment Tensor Solutions and the Surface Deformation from GPS and InSAR. Bulletin of the Seismological Society of America, 94:363-376; doi:10.1785/012003007

 

Shan, W., C. J. Wang and Q. Hu, 2012. Expressway and Road Area Deformation Monitoring Research Based on InSAR Technology in Isolated Permafrost Area, 2012 2nd International Conference on Remote Sensing, Environment and Transportation Engineering, Nanjing, 2012, pp. 1-5. doi: 10.1109/RSETE.2012.626057

 

Tarchi, D.,  H. Rudolf, G. Luzi, L. Chiarantini, P. Coppo, A.J. Sieber, 1999. SAR interferometry for structural changes detection: a demonstration test on a dam Proc. International Geosci. Remote Sens. Symposium, IGARSS, Hamburg Germany (1999), pp. 1522–1524.

 

Tofani, V., F. Raspini, F. Catani, N. Casagli, 2013. Persistent scatterer interferometry (PSI) technique for landslide characterization and monitoring Remote Sens., 5 (3), pp. 1045–1065.

 

Uhlmeyer, J.S.; Pierce, L.M.; Lovejoy, J.S.; Gribner, M.R.; Mahoney, J.P. and Olsen, G.D.  (2002).  Design and Construction of Rock Cap Roadways – A Case Study in Northeast Washington.  Proc. 2003 TRB annual meeting.

 

Yu, B., Liu, G., Zhang, R. et al., 2013.  Monitoring subsidence rates along road network by persistent scatterer SAR interferometry with high-resolution TerraSAR-X imagery. J. Mod. Transport. (2013) 21: 236. doi:10.1007/s40534-013-0030-yng

 

Zhang, T., R. G. Barry, K. Knowles, J. A. Heginbottom and J. Brown, 1999. Statistics and characteristics of permafrost and ground-ice distribution in the Northern Hemisphere. Polar Geogr., 23(2), 132 - 154.

 

Many applications of synthetic aperture radar differential interferometry (DInSAR) lead to a set of sparse phase measurements, e.g. in the processing of long multitemporal stacks of SAR differential interferograms through persistent scatterers interferometry (PSI) techniques. Often, sparse phase data have to be unwrapped, and then interpolated on a regular grid to be useful for subsequent processing steps. This step is necessary for instance in the reconstruction of the so-called APS (Atmospheric Phase Screen). Atmospheric artifacts superimposed on DInSAR measurements have the potential of hindering the accurate estimation of deformation signals. Indeed, sometimes the spatial frequencies of the atmospheric phase contributions can overlap those of deformation signals, so that such artifacts can be misinterpreted as deformation features.

For the phase unwrapping stage, the solutions are directly dependent on the PS network density; moreover, phase aliasing, which appears when the signal sampling does not satisfy the Nyquist condition, especially in presence of noise, increases when passing from regular-grid to sparse data. This is because the phase sampling conditions get usually worse. 

An improvement of the APS estimation step has been proposed, by investigating from the empirical point of view an alternative procedure, which involves an interpolation of the complex field derived from the sparse phase measurements. Unlike traditional approaches, the proposed method allows to bypass the PU step and obtain a regular-grid complex field, from which a wrapped phase field can be extracted. Under general conditions, this smooth phase field can be shown to be a good approximation of the original phase without noise. Moreover, the interpolated, wrapped phase field can be fed to state of the art, regular grid PU algorithms, to obtain a smoother absolute phase field.

The performances of this empirical approach are evaluated here over a real dataset, that is composed by 30 ascending SAR X-band COSMO-SkyMed images. The images cover the urban area and outskirts of the capital of Haiti, Port-au-Prince.

The accuracy of the reconstructed phase fields is analyzed by the local value of the final inter-image phase coherence (γ_int), a quality figure related to the residual phase noise after subtraction of all modeled contributions. Its values are taken on points (PS) not used in the interpolation, using different spatial densities and random subsampling patterns in a test area characterized by a strong subsidence bowl.

The obtained results may be applied into a broader context than the one specific to the PSI technique, considering the few assumptions on the initial phase field, i.e. its smoothness and good sampling conditions.

 

Acknowledgments: CSK® Products © of the Italian Space Agency (ASI), delivered under a license to use by ASI, in the framework of a research activities of the Department of Physics (DIF) of the University of Bari (Italy).

Abstract: The stability of slope is a problem that needs to be paid attention to in the mining, and the effective monitoring and analysis of the stability of the slope can provide decision-making basis for safe production. In this paper, based on the ground synthetic aperture radar (GB-SAR) monitoring slope stability program, the use of small baseline set technology to deal with monitoring data. For the continuous monitoring data of GB-SAR, the space between the images is zero. Therefore, this paper divides the data into several subsets, and uses the correlation between images to determine the threshold of time base line. Images in each subset is interfered, and the interference pattern with redundant observation is obtained. The relation between adjacent subsets is obtained by a repeated image. This paper uses spatial coherence as a reference index to select high coherent points, prevent the leakage of selected targets large deformation. And for the small baseline spaceborne SAR set (SBAS) technology can be compared to GB-SAR interferograms with redundant observation, by using the least square method is used to calculate the deformation phase high coherent points. Through experiments, the feasibility of GB-SAR slope monitoring data is verified by the small baseline set technology.

Key words:  slope; ground based synthetic aperture radar; small baseline subset; deformation monitoring

Central Florida's thick carbonate deposits and hydrological conditions make the area prone to sinkhole development. Sinkhole collapse is a major geologic hazard, threatening human life and causing substantial damage to property. Detecting sinkhole deformation before a collapse is a difficult task, due to small and typically unnoticeable surface changes. Most techniques used to map sinkholes, such as ground penetrating radar, require ground contact and are practical for localized (typically 2D, tens to hundreds of meters) surveys but not for broad study areas.

In this study we use Persistent Scatterer (PS) time series analysis of Interferometric Synthetic Aperture Radar (InSAR), which is a very useful technique for detecting localized deformation while covering vast areas. We acquired SAR images over four locations in central Florida in order to detect possible pre-collapse or slow subsidence surface movements. Our data consists of TerraSAR-X and COSMO-SkyMed images with pixel resolutions ranging between 25cm and 1m. To date, we have obtained four datasets, each covering a period from March of 2015 to June of 2016 over a total of roughly 2200 km². We generate PS time series for each of the four datasets using DORIS and StaMPS software packages. Preliminary results indicate localized deformation in the range of 5mm/yr in some houses and commercial and apartment buildings in two of the sites. Deforming areas vary in size from approximately 10m x 20m of a single house to 60m x 60m for a commercial building. Future work will include the expansion of the PS time series beyond June 2016 and ground truth surveys with ground penetrating radar for verifying the space-based sinkhole activity detection.

The Aswan High Dam, Egypt, was built in the 1970s and is one of the biggest dams in the world. It stopped the seasonal flood of Nile river, allowing the urban expansion of cities/villages, full year cultivation and produces $10 milliard KWH annually. The dam is located in an area where several earthquakes (< M_L 6) occurred from 1981 to 2007.

Here we want to identify any potential damage that could be caused to the dam, and assess its overall structural stability using Persistent Scatterer Interferometry (PSI). We used Envisat data, from a descending orbit and acquired between 2004 and 2010.

Preliminary results show that small rates (maximum around 3mm/year in the satellite Line-Of-Sight) of deformation can be identified, which implications must be further investigated. In addition, the results indicate that the Aswan High Dam presents two different behaviors. The western part shows differential subsidence relative to a reference point selected outside the dam, while the  eastern part, corresponding to the electrical power plant, shows a slight up-lift. Both phenomena need further investigation to assess if the detected movements correspond to the expected vertical behavior for this kind of mega-structures.

The opportunity to monitor opencast mining areas has increased significantly once with the launching of the Sentinel 1 mission. This kind of information allows us to gather and consolidate the knowledge related to the pattern of distributions, function and interaction of natural resources with other spatially explicit factors (e.g. land cover, human development and environmental disasters). Advanced differential SAR interferometry techniques, small based subset methodology and coherent change detection were applied to characterize the ground motion over Oltenia Coalfield, Romania. Geometrical and geological factors that play a role in the ground movement were considered in this analysis. Therefore the slope instabilities and flat accumulation areas were investigated. First, the researches were conducted at the highest spatial resolution through 23 TerraSAR-X data acquired in Starring Spotlight mode between 2014 and 2015 over Rosia Jiu opencast area. Secondly, Sentinel 1 C-band capabilities to assess ground deformation at regional scale in the Oltenia Coalfield basin was inspected. The dominant features of the observed motion are a relative large spatial areas of landslide in the tailings areas and subsidence in nearby villages. Finally, a detailed analysis of these results has enabled to present the advantages and disadvantages of applying these techniques for mapping ground degradation in the opencast mining areas.

The area around the central train station in Oslo, Norway, has been the focus of several major urban development projects over the past decades. Developments include a new highway tunnel passing under the city, bridges, a new opera house and a number of office and residential buildings. Being a historic landfill area including layers of sawdust deposits from upstream sawmills, a natural amount of subsidence was expected. In order to ensure the stability, all new constructions were built on pile foundations, reaching down to the underlying bedrock. However, a reduction in pore pressure, caused by increased drainage as a result of construction activities, triggered increased subsidence in certain areas, observed by ground-based measurements. In order to monitor the effect on existing infrastructure, primarily nearby railroad tracks, the Norwegian Railroad Authority started a subsidence monitoring project including ground-based measurements and satellite-based InSAR measurements.

In order to map the subsidence over the area, four different InSAR stacks with different resolutions (multi-looked) have been acquired: (a) Sentinel-1 (~20m); (b) Radarsat Standard Mode (~20m); (c) Radarsat-2 Fine Mode (~7m); and (d) TerraSAR-X Staring Spotlight (~0,7m). From late 2014 all four stacks cover the target area and a direct comparison of the different resolutions was possible. All stacks were processed with the SBAS workflow as well as with the Permanent Scatterer (PS) workflow, both using SARscape. While the PS gave better coverage over constructions like buildings or bridges, the SBAS workflow gave mostly better coverage over the wide track area east of the train station. The processing results show that the density of PS' is heavily impacted by the resolution, especially for the track area, while the coverage with coherent SBAS pixels is similar for all four stacks.

As expected, do high resolution results show a more focused image of the subsidence patterns. Especially for SBAS the results from the 20m resolution stacks show are very blurry and only able to roughly indicate location of subsidence patterns. Some small scale patterns are missed out entirely. The PS workflow provides more focused results even for the 20m stack. Still, with increasing resolution the subsidence patterns get more focused. While for larger subsidence pattern the determined subsidence rates agree for all resolutions, small scale features appear weaker in coarser resolutions.

A comparison with ground based measurements shows that in especially for low resolution data the InSAR results are often lower than the ground truth. For Sentile-1 and Radarsat-2 standard mode this is most likely due to the pixel size is approx. 400m^2, which may be larger than the actual footprint of a small but strong subsidence pattern. For finer resolution this effect can be ruled out, however, it appears that filter operations of the InSAR workflows also is a contributing factor.

InSAR time series analysis methods, including PS InSAR or SBAS, have been wide used for mapping crustal movement, ground surface subsistence and so on. In past decades, one problem the InSAR world often faced is how to collect enough many InSAR data with proper spatial and temporal baseline, that is really difficult in some areas not hot. Since launch of Sentinel 1A, of which the interferometric wide mode has been set as regular duty mode, 12 days of revisit cycle made acquisition InSAR data became easy. Along with Sentinel 1B started work, the revisit cycle furtherly shorted as 6 days. For most of interferometric SAR application, short of data is not any longer problem.

For long term deformation monitoring with progressive data accumulation, classic methods such as PS InSAR or SBAS, which treat selected InSAR data set and gave out deformation velocity map with linear or nonlinear model, or timely deformation behavior history for some high coherent stable points. If more new acquisition data has been added along with monitoring going on, we usually process these existed data plus new data as a whole set to get new result. Obviously, there are vast redundancy computation in such case. Another way is only processing a proper InSAR data set including new acquisition and part of old data which could connected with new acquisition in good interferometric condition, and also fully combined with result derived from the old data set. However, there is no existing theory and methods dealing with latter.

In this paper, we try to propose a novel InSAR time series analysis method based on Short Baseline Subset (SBAS) method to treat progressive data accumulation. Firstly, we would using SBAS method for initial InSAR data set to get average deformation velocity during corresponding time span. Along with more new acquisition data has been input, new interferometric pairs between new data and old data with good condition have been generated. Based on the initial result and these new interferometric pairs, innovated time series analysis method would been used to retrieve timely deformation information.

The Sentinel-1A/B TOPS data for Yellow River Delta of China would be used for experiment. In this area there are severe ground subsistence associated with petroleum exploration and ground water pumping, which is maybe harmful to safe of the hydraulic engineering, the Canal of South-to-North Water Transfer and the reservoir. We would compare results between classic SBAS and the novel method.

The territory of the Upper Nitra catchment belongs to the Slovak regions heavily affected by geohazards – namely slope failures and territory subsidence due to undermining.

The mass movement processes are generated by a specific geological setting of the Vtáčnik mountain range. Rigid volcanic rocks comprising epiclastic volcanic conglomerates and breccias, pyroclastic and lava flows of Older to Middle Sarmatian, are forming the summit parts and flanks of the mountain range. They overlay softer plastic complexes of Neogene sediments, prevailingly clays/claystones (Koš Fm. – Younger Badenian) and sandstones and siltstones (Handlová Fm. – Younger Badenian). The Badenian complexes are known for mighty seams of brown coal present at depths of ca 200-300 m below surface.

The slope deformations have a strong representation here, with almost all basic types of slope failures. In the summit parts of the mountain range block ridges are present. Down the slope the detached blocks form block fields, and finally, towards the Upper Nitra Depression, landslides have evolved. Since 90-ies of the previous century the most dangerous landslides have been gradually included in the project of Monitoring of Geological Hazards.

The unfavourable slope stability conditions have been impaired by underground mining of brown coal which at several sites led to acceleration of mass movements or to the subsidence of the territory above mined-out underground spaces. Although the coal mining is gradually ceasing out, the subsidence still represents an ongoing process.

The monitoring methodology of affected sites was based on classical methods of terrestrial geodesy, piezometric and inclinometric measurements and observation of climatic parameters. The onset of up-to-date technologies of remote sensing has opened new ways of geohazards observations also in this part of Slovakia.

In this work, we demonstrate the applicability of satellite SAR interferometry (InSAR) for identifying slope movement activity in urban areas, semi-arid regions or over low vegetated slopes. Presented are the results from processing of Sentinel-1A/B data using Persistent Scatterers InSAR and advanced multi-temporal InSAR techniques implemented in SARPROZ software. The reliability of the InSAR measurements is verified via comparison with terrestrial measurement techniques.

While exploiting former ERS and ENVISAT missions, many slope failures in Upper Nitra region remained unnoticed in InSAR results due to rapid changes that were undetectable utilising revisit periods of 35 days (and more). Since the deformation phenomena are persistent over recent days and the area is under vital monitoring efforts, the availability of shorter revisit intervals of Sentinel-1A/B (up to 6 days) underlines the operational capability of Sentinel-1A/B observations for routine monitoring and regular updates of active landslide maps.

Ground deformation has been a significant problem in reclaimed land from see, as the reclamation is usually carried out by dumping uncompacted fill materials onto a seabed of unconsolidated marine sediment. In particular, the settlement variability is crucial to performance assessment of the reclamation development because this differential settlement, rather than total settlement, can lead to damage of ground constructions (such as buildings, bridges, runways and highways) and underground facilities, with possible consequences in terms of human and economic losses[1]. In addition, large-scale constructions easily cause land subsidence because of the loading of buildings. Timely and precisely monitoring the subsidence of reclaimed lands will help to prevent geological hazards and economic loss. Thus, monitoring and assessing of reclamation subsidence in coastal areas is a necessary work.

Radar interferometry is a reliable remotely-sensing technology of measuring subtle deformations of both natural and man-made structures. It can provide ground deformation information faster and more economically than traditional ground-based observation techniques. Recent advances in SAR interferometry have demonstrated the robustness and precision of some advanced InSAR approaches to overcome the intrinsic limitations of conventional InSAR. In 2001, Persistent Scatterers InSAR (PSInSAR) was proposed and had been proven as an effective tool for deformation multi-temporal analysis in urban areas[2]. However, the study on ground deformation in reclaimed areas proved to be a big challenging for PSInSAR because of the complexity of the deformation regimes, non-linear subsidence and lacking of angular man-made targets in non-urban (rural) areas. In order to improve PS network density, a distributed scatterer (DS) interferometric analysis has been utilized in some advanced PSInSAR process, such as SqueeSAR[3]. It extremely enhances the spatial density of measurement points by means of jointly processing PS and DS. However, this methods does not take into account the interference between different scattering mechanisms and results in an unreliable estimation especially for small sample sizes. Moreover, this method is sensitive to data stacks due to the variability of temporal samples and is computationally intensive.

In this paper, an improved DS-InSAR approach is developed to solve the above problems, which is characterized with FaSHPS-based DS detection and eigendecomposition phase optimization[4],[5]. First, we explore the confidence interval for each pixel by invoking the central limit theorem, and select SHPs using this interval for distributed targets. Then, we estimate the sample coherence matrix taking advantage of the SHP families and perform eigendecomposition on the coherence matrix in order to estimate the optimized phases and select the DS corresponding to the different scattering mechanisms. Finally, the selected DSs are processed together with the PSs using the conventional PSInSAR algorithm for estimating the displacement time series for each measurement point. Compared with SqueeSAR, the identification of homogeneous pixel seems more robust and reliable. Also, the estimated phases are more accurate because of eliminating the impact of different scattering mechanisms. Particularly, both the identification and optimization can significantly improve the computational efficiency. A real data analyses over the Hong Kong Science Park reclamation area and Hong Kong International Airport verifies the efficiency and robustness performance of the proposed method superior to other DS-InSAR implementations. The result demonstrates that it is effective to improve target density, accuracy and efficiency in monitoring ground deformation particularly over non-urban areas.

References
[1]Jiang, Liming, Hui Lin, and Shilai Cheng, Monitoring and assessing reclamation settlement in coastal areas with advanced InSAR techniques: Macao city (China) case study. International Journal of Remote Sensing, 2011. 32(13): p. 3565-3588.
[2]Ferretti A, Prati C, Rocca F. Permanent scatterers in SAR interferometry[J]. Geoscience and Remote Sensing, IEEE Transactions on, 2001, 39(1): 8-20.
[3]Ferretti A, Fumagalli A, Novali F, et al. A New Algorithm for Processing Interferometric Data-Stacks: SqueeSAR[J]. IEEE Transactions on Geoscience & Remote Sensing, 2011, 49(9):3460-3470.
[4]Jiang M, Ding X, Hanssen R F, et al. Fast Statistically Homogeneous Pixel Selection for Covariance Matrix Estimation for Multitemporal InSAR[J]. IEEE Transactions on Geoscience & Remote Sensing, 2015, 53(3):1213-1224.
[5]Cao N, Lee H, Jung H C. A Phase-Decomposition-Based PSInSAR Processing Method[J]. IEEE Transactions on Geoscience and Remote Sensing, 2016, 54(2): 1074-1090.

The Tatun Volcano Group (TVG) consisting of over twenty Quaternary volcanoes, is located in northern Taipei Basin.  Although there has been no Plinian  eruption occurring in human history, but a recent study shows that a phreatic eruption may have occurred at approximate 6000 years ago (Belousov et al., 2010). Moreover, the highly active geothermal and seismic activities imply that TVG might be dormant active and of potential to erupt in the future. Since TVG is only a few kilometers north of the Taipei metropolitan area and northeast of two nuclear power plants, the possibility of eruption necessitates monitoring precursor phenomenon related to the volcanic activities.

Interferometric Synthetic Aperture Radar (InSAR) has been successfully applied in mapping ground surface deformation in volcano zones to analyze the latent geophysical dynamics (Lu et al., 2007). In this study, we make use of the newly developed Temporarily Coherent Point InSAR (TCPInSAR) (Zhang et al., 2012) to exploit the spatial and temporal surface evolution over TVG. Taking into account of the densely vegetated and mountainous cover, 17 L-band ALOS/PALSAR-1 images spanning from 2011 to 2013 are used retrieve the displacement history.  The stratified tropospheric delay is well considered and corrected by employing a patch-based method based on joint estimation.

The derived deformation results are validated by comparing with GPS measurements and have a good agreement. The observed uplifts indicate that TVG in recent years has undergone active geothermal activities. Meanwhile, the comparison with precipitation data reveals that the seasonal variation of displacement is related with rainfall. The results demonstrate that InSAR has the capability of monitoring surface deformation over volcanic zones.

References

Belousov, A., Belousova, M., Chen, C.-H. and Zellmer, G.F., 2010. Deposits, character and timing of recent eruptions and gravitational collapses in Tatun Volcanic Group, Northern Taiwan: Hazard-related issues. Journal of Volcanology and Geothermal Research, 191(3): 205-221.

Lu, Z., Dzurisin, D., Wicks, C., Power, J., Kwoun, O. and Rykhus, R., 2007. Diverse deformation patterns of Aleutian volcanoes from satellite interferometric synthetic aperture radar (InSAR). Volcanism and subduction: the Kamchatka region: 249-261.

Zhang, L., Lu, Z., Ding, X., Jung, H.-s., Feng, G. and Lee, C.-W., 2012. Mapping ground surface deformation using temporarily coherent point SAR interferometry: Application to Los Angeles Basin. Remote Sensing of Environment, 117: 429-439.

Measuring and monitoring deformations of man-made structures such a bridges is a key task of applied geodesy and geomatics. However, these deformation measurement techniques are time consuming and thus expensive. The rapid development of space technology occurred in the last decades has allowed the detection of the displacement of earth surface from space with high precision and unexpected benefits for earth observation and related global studies. This progress has been possible thanks to microwave images obtained through Synthetic Aperture Radars (SAR) systems mounted on satellites and the development of Multi-Temporal Interferometry (MTI) techniques. It is very important to develop effective bridge (and others civil infrastructures) monitoring approaches that can help identifying structural problems before they become critical and endanger public safety. By applying InSAR processing techniques to a series of radar images over the same region, it is possible to detect movements of infrastructure systems on the ground in the millimetre range, and therefore identify abnormal or excessive movement indicating a potential problem that needs more detailed ground investigation. In this paper we investigate the behaviour of the Carranza and the recently open 1812 Constitution bridges over the bay of Cádiz (SW Spain) using satellite radar interferometry.

Multi-temporal InSAR methods are effective tools for monitoring and investigating surface displacement on Earth based on conventional radar interferometry. These techniques allow us to measure deformation with uncertainties of one millimeter per year, interpreting time series of interferometric phases at coherent point scatterers (PS). Over the last decades, coastal areas in many parts of Spain have undergone a continuous urban expansion because of the growth of cities and development of new residential areas. The transgression of the sea, as a consequence of sea level rise and the subsidence of populated areas, may result in serious problems to many constructions situated in the coastline. This has an important impact on the economy, environment and society, representing a considerable natural hazard. We use ERS-1/2 and Envisat data in the period 1992-2010 to detect subsidence areas over the southern Spanish coast using time series analysis of SAR data. 

This work is focused on geohazard mapping and monitoring using Sentinel-1 (S-1) data and the DInSAR (Differential interferometric SAR (Synthetic Aperture Radar)) techniques. Since the first results achieved in 1989, the DInSAR techniques have experienced a continuous growth mainly related to two main aspects. First, an important research and development effort has generated a wide number of data processing and analysis tools and methods, including DInSAR and PSI. The second aspect is the satellite data availability. Most of the DInSAR and PSI developments have been based on C-band data acquired by the sensors on-board the satellites ERS-1/2, Envisat and Radarsat. These data cover long periods of time, which is a key aspect to guarantee a long-term deformation monitoring. A significant improvement is given by the new C-band sensor on-board the S-1A and B satellites, launched on 2014 and 2016 respectivelly. S-1 has improved the data acquisition throughout, with respect to previous sensors, increasing considerably the DInSAR and PSI deformation monitoring potential.

 

This work describes a procedure to generate geohazard activity maps in a relatively rapid way over wide regions by exploiting the high coherence and temporal sampling provided by the S-1 satellites. This procedure has been developed in the framework of the project Safety (www.safety.es) which aims at providing Civil Protection Authorities (CPA) with the capability of periodically evaluating and assessing the potential impact of geohazards (volcanic activity, earthquakes, landslides and subsidence) on urban areas. 

We present and discuss first results in Spain achieved on the application of Sentinel-1 mission for the monitoring and survey of relatively small and isolated hydrological critical infrastructures like a dam and its surroundings. We describe the results obtained for the Siles dam, located in the province of Jaén (Spain), during its settlement and initial load process. ESA´s Sentinel-1A Interferometric Wide Swath (IW) SAR data are used and processed using the Coherent Pixel Technique (CPT) algorithm. A comparison of the A-DInSAR results with in-situ measurements will be also given. First results for this comparison shows a good agreement between A-DInSAR and levelling for the maximum value of subsidence. The results show that Sentinels program can be a very good choice for this type of studies due its short repeat cycle and spatial resolution as well as the open data Access policy which permits to obtain a high number of images without charge. The last facilitates to define a sustainable operational monitoring system.

 Long-term small deformation is an important index of slow-moving slope hazard (SMSH) recognition, and fault zones are usually the regions of serious slope hazards because of strong tectonic movement, frequent earthquakes, and cracked rock and soil. Taking the sustaining active Xianshuihe Fault (XSF) zone in the eastern Tibet Plateau as the study area, this paper collected 53 scenes of ALOS/PALSAR to perform the multi-temporal Interferometry Synthetic Aperture Radar (InSAR) process; yielded millimetric deformation precision; and by combining with field investigation, remote sensing interpretation and geological setting analysis, identified the types, deformation features and spatial distribution of SMSHs, as follows: 1) Creep landslide, debris flow and slow-moving moraine are the three main types. 2) Creep landsides have the geomorphic characteristics of garbled horizontal steps, a rough and uneven longitudinal profile, vegetable covered landslide beds, etc., which can be taken as the features of field validation. 3) Most creep landslides distributed in the northern section are paleo-landslides, co-seismic landslides and shattering slopes, of which those intersecting with XSF are obviously movable and those without intersection are stable. 4) In the areas of debris flows source, the slow and disperse deformation of rock and soil commonly develops, which can be considered as the index for identifying the debris flows, and two types of debris flow sources are found: "soil-stone forest" and "trench loose deposition". 5) In the alpine area above 4200 m a.s.l., there is widely distributed moraine slope movement, with huge body and faster velocity, which is the main erosion method in this area. The study results also indicate that the combination of InSAR and geological setting analysis can identify SMSH efficiently, being preferential for work involving a great amount of hazards and inconvenient field investigations in mountainous areas.

We present new results on the deformation field for La Palma Island in the Canary Archipelago using A-DInSAR analysis of C-Band SAR images from ENVISAT satellite for a six-year period between years 2004 and 2010. The recent volcanic activity registered in La Palma (with seven eruptions in ca. A.D. 1480, 1585, 1646, 1677, 1712, 1949 and 1971) and the absence of any geodetic monitoring system implemented in the Island at the 90’s converts A-DInSAR studies in a basic information source for the study of surface displacements. From late 90’s several geodetic studies using GNSS, gravimetry and InSAR techniques have been carried out in the island (see Fernández et al., 2015 for a summary). In this study we updated previous radar satellite results by processing both geometries (ascending and descending), increasing the temporal coverage up to 2010 and, using a recently developed A-DInSAR processing technique, including an error estimation for the LOS mean velocity and deformation time series (“Subsidence” software, developed by the “Universitat Politècnica de Catalunya, see e.g., Blanco-Sánchez et al., 2008, and Centolanza, 2015). Our A-DInSAR results are compared with previous InSAR ones and with GNSS results obtained from several campaigns. A discussion and geological interpretation including the inversion of the results to obtain the deformation sources is also presented. Radar images have been provided by ESA through Cat.-1 13933 project.

REFERENCES

Blanco-Sánchez, P., Mallorqui, J. J., Duque, S., and Monells, D. (2008), The Coherent Pixels Technique (CPT): An Advanced DInSAR Technique for Nonlinear Deformation Monitoring, Pure Applied Geophysics, 165, 6, 1167–1193.

Centolanza, G. (2015), Quality Evaluation of Dinsar Results from the Phase Statistical Analysis. Universitat Politècnica de Catalunya.

Fernández, J., P.J. González, A. G. Camacho, J.F. Prieto, G. Bru, 2015. Volcano geodetic research in the Canary Islands: A summary of results and perspectives. Pure and applied geophysics, 172/11, 3189-3228. DOI: 10.1007/s00024-014-0916-6.

In the frame of the EC FP7 MED-SUV project (call FP7 ENV.2012.6.4-2), and thanks to the GEO-GSNL initiative, GPS data and SENTINEL 1A/1B TOPSAR acquired on Mt. Etna between October 2014 and November 2016 were analyzed. The SENTINEL data were used in order to combine and integrate them with GPS, and detail the ground deformation recorded by GPS on Mt. Etna, during the last two-year’s volcanic activity.

The Sentinel data were processed by GAMMA software, using a spectral diversity method and a procedure able to co-register the SENTINEL pairs with extremely high precision (< 0.01 pixel). In order to optimize the time processing, a new software architecture based on the hypervisor virtualization technology for the x64 versions of Windows has been implemented. The DInSAR results are analysed and successively used as input for the time series analysis using the StaMPS package.

On December 28, 2014 eruptive activity resumed at Mt. Etna with a fire fountain activity feeding two lava flows spreading on the eastern and south-western upper flanks of the volcano, producing evident deformation at the summit of the volcano. GPS displacements and Sentinel-1A ascending interferogram were calculated in order to image the ground deformation pattern accompanying the eruption. The ground deformation pattern has been perfectly depicted by the GPS network, mainly affecting the uppermost part of the volcano edifice, with a strong decay of the deformation, according to a very shallow and strong dyke intrusion. The Sentinel 1A SAR data, covering the similar time spanning, confirmed that most of displacements are related to the dike intrusion, and evidenced a local gravity-driven motion  of the western wall of the Valle del Bove, probably related to the dike intrusion.

To monitor the temporal successive evolution of ground deformation, we performed an A-DInSAR SENTINEL analysis using the Small BAseline Subset (SBAS) approach included with the StaMPS processing package. The April 2015-December 2015, SBAS Time series, shown a volcano inflation, with an uplift of about 28 mm localized in the central and western area of the volcano. Suddenly, in the first days of December 2015, volcanic activity abruptly restarted at the central crater with a very strongly explosive eruption; this kind of activity continued, with a decreasing intensity, with other episodes at the same crater and then involving, in turn, all the other three summit craters of the volcano. On December 8, when the eruptive activity was concluding, a seismic swarm affected the uppermost part of the Pernicana fault where it joins the NE-Rift. The SBAS time series have then been integrated by the SISTEM algorithm with the ground displacements measured by two GPS surveys carried out on the NE flank of the volcano at the end of April and in mid-December 2015. Results of this data integration provide a very detailed picture of the ground deformation pattern on the volcano, preceding and accompanying the vigorous eruption and the seismic swarm; besides the general inflation of the edifice during the pre-eruptive period.

The January 2016 – November 2016 is the last period analyzed, characterized by the kinematic of the eastern unstable flank, with displacement involving both the Pernicana fault and the other structures dissecting this sector of the volcano.

Persistent Scatterers (PS) method is a modification of radar interferometry technique which is used in the case of high temporal decorrelation. This method is based on the idea of using "permanent scatterers" which are objects that preserve the stability of the backscatter level and location of the scattered signal phase center. The primary criterion for the PS identification is the level of scattered signal stability expressed in terms of the variance of the signal amplitude. The threshold level of the amplitude variance in this study was taken 0.2.

In this study we used RADARSAT-2 polarimetric SAR data obtained in Fine QuadPol mode between 2013 and 2014, the cycle of interferometric observations covered all seasons. Volgograd city was chosen as a test site.

Therefore it was possible to investigate the behavior of PS candidates at signal different polarizations and to identify the dominant mechanisms of backscatter. Information about the scattering mechanisms can be obtained using coherent polarimetric decompositions. We used the  Pauli polarimetric decomposition.

Pauli decomposition expresses the scattering matrix of each pixel as the combination of the responses of the plate, dihedral and diplane oriented at 45 degrees, which can be interpreted as surface, double-bounce and volume scattering. We studied the relative contribution of these mechanisms in total PS scattering. The result was that the PS dominant scattering mechanism is surface scattering (48% of PS backscatter is provided by this mechanism), the second – double-bounce scattering (30%), the third – volume scattering (22%).

The majority of the PS candidates is characterized by double-bounce and surface scattering, or a combination of these mechanisms. In this study, we have found, that the contribution of volume scattering to entire scattering exceeds the threshold 30% in 12% of PS. An increase of the threshold leads to fast decrease of the number of PS with domination of volume scattering. Volume scattering typically corresponds to scattering from vegetation. Therefore by selecting PS with volume scattering, we may identify PS with a diffuse or volume scattering mechanism specific to vegetation. As the phase of signal scattered from these PS is a random process, it cannot be used for mapping of the ground deformation, and such PS should be excluded from further processing.

Radarsat-2 data were obtained for this study in the framework of the international educational program of the Canadian space Agency SOAR-EI, project No. 5137.

Abstract: Landslide is a kind of natural geological disaster which can cause enormous damage [1]. Landslide deformation monitoring is important for geological disaster prevention. Synthetic Aperture Radar (SAR) techniques, especially Interferometric SAR (InSAR), have been widely used for extracting landslide deformation [2-3]. However, for the landslide with high moving velocity, the traditional D-InSAR technique is limited by its maximum detectable displacement [4]. In this case, the intensity-based offset tracking methods (e.g., Normalized Cross-Correlation method) is widely used for landslide displacement estimation. The normalized cross-correlation (NCC) method, based on single-channel SAR images, estimates azimuth and range displacement by finding peak statistical correlation between the matching windows of two SAR images [5]. However, the matching windows, especially for the boundary area of landslide, always contain pixels with different moving characteristics, affecting the precision of displacement estimation [4]. Based on the advantages of polarimetric scattering properties, this paper proposes a fully polarimetric SAR (PolSAR) offset tracking method for improving the precision of landslide displacement estimation. The method uses the polarimetric information to evaluate their similarity. A pair of high resolution SAR images covering the Slumgullion landslide located in southwestern Colorado, USA are used for experiments. The results indicate the the effectiveness of the proposed method. In comparison with traditional NCC method, the proposed method presents better perfomance in sub-pixel estimation.

 

References:

[1] Dai, F.C.; Lee, C.F.; Ngai, Y.Y. Landslide risk assessment and management: An overview. Eng. Geol. 2002, 64, 65–87.

[2] Crosetto, M.; Gili, J.A.; Monserrat, O.; Cuevas-González, M.; Corominas, J.; Serral, D. Interferometric SAR monitoring of the Vallcebre landslide (Spain) using corner reflectors. Natl. Hazards Earth Syst. Sci. 2013, 13, 923–933.

[3] Colesanti, C.; Wasowski, J. Investigating landslides with space-borne Synthetic Aperture Radar (SAR) interferometry. Eng. Geol. 2006, 88, 173–199.

[4] Wang, C.; Mao, X.; Wang, Q., Landslide Displacement Monitoring by a Fully Polarimetric SAR Offset Tracking Method. Remote Sens. 2016, 8, 624.

[5] Debella-Gilo, M.; Kääb, A. Sub-pixel precision image matching for measuring surface displacements on mass movements using normalized cross-correlation. Remote Sens. Environ. 2011, 115, 130–142.

Monitoring the structural stability of urban areas and observing the settlements affecting large infrastructures are emerging as dominant socio-economic issues for the safeguarding of the population since they a primary role in setting up mitigation and prevention actions. In large urban areas, the problem is accentuated by the age of the constructions that make them more exposed to increasing risks as a results of the material deterioration and loss of loading capacity. These activities become a civil protection issue when the structures are threatened by critical evolution of natural and man-made ground deformation processes.

 

Nowadays the evaluation of risks associated to subsidence and soil movements is based on the use of ground based methods, which are able to measure displacements at the surface or in boreholes, and on direct analyses such as in-situ inspections/investigations.

These methods, although accurate at a local scale, require placing devices on the structures (destructive method) which is expensive and not always feasible due to accessibility and logistic constraints. In addition, due to the extension, capillarity and frequency required for the monitoring of large urban areas, critical infrastructures and networks (road, railway, airport), approaches based only on in-situ measurements would require huge resources, not available today.

 

To guarantee a systematic and comprehensive control of structural stability over large areas, satellite remote sensing can be effectively adopted. However, the processing chains and the specifications for analysing and reporting the outcomes of satellite-based investigations still need to be better defined, validated and made coherent with the standards expected by the reference user community. Among the different methods based on passive and active satellite sensors, the Differential Interferometry SAR technology today represents an effective solution in terms of precision reliability and cost sustainability.

 

I.MODI project (Implemented MOnitoring system for structural Displacement), funded by Horizon 2020's SME Instrument Phase 2 program, aims at exploiting EO data to create a value added service where the integration between EO observation technologies, ground based data and ICT represents the core of the system in order to facilitate the access of a different types of users. The implemented service is based on a web-based customized platform that presently permits the full integration of archived EO data (COSMO-SkyMed and ESA archive SAR data from ERS-1/2 and ENVISAT) into the standard operational procedures based on in-situ technologies. The service is also designed to include Sentinel1 datasets. In addition I.MODI proposes an innovative approach for assessing the damage of structures and infrastructures by means of the exploitation of DInSAR derived displacements. Starting from the DInSAR displacement time histories, relevant parameters linked to the structural behavior are adopted to calibrate numerical methods for damage assessment. The adopted modeling approaches range from empirical estimates of the strain parameters to detailed finite element models.

In the framework of the project, the implemented procedures have been tested on real cases by performing back-analysis. The results have been validated by comparing the observed damages with that resulting from  the numerical models. In particular, the presented test cases deal with areas in the city of Rome affected by relevant settlement processes and large infrastructure in Italy.

 

Due to overextraction of groundwater resources for agricultural and industrial purposes in the last decades, most of the plain aquifers in Iran are subject to sever land subsidence, at locations exceeding 20-30 cm/yr. For some of them, field measurements or remote sensing methods, particularly InSAR, have been used in the past to investigate the cause, extent and rate of displacement, and hydrogeological characteristics of aquifers. However, as this problem has been spread out all over  country, it is very important  to develop a precise and near-real time land subsidence inventory to better understand the link between ground deformation data and hydrological parameters at large scales.

Thanks to Sentinel-1 wide footprint and regular and large coverage all over the world, we are able to generate continental-scale interfergorams to analyze geological processes occurring at any place at anytime using InSAR. However, with its moderate resolution and large coverage, Sentinel-1 provides a huge amount of data that might be difficult to deal with in terms of processing. To reduce the processing load and time, we suggest a method that automatically identifies the extent of subsidence basins from InSAR results by correlating the results with other parameters such as DEM, slope, and land-use. Then, a resampling method is used to keep the high density of pixels in subsidence areas for time-series analysis. In this way, processing load is reduced dramatically while the highest possible density of pixels are kept for deformation analysis. The time-series results can then be used to assess short-term and long-term variations of surface deformation in response to hydrological parameters in near real-time . We present examples from several basins in Iran including Tehran, the capital sate of Iran with 12 million inhabitants, and Rafsanjan Plain, Iran's center of pistachio cultivation in the southeast, and investigate the advantages and performance of Sentinel-1 in comparison to other SAR sensors for a long-term assessment of subsidence hazard and characterizing the mechanism of aquifer compaction that controls such hazards.

The agricultural expansion that has been taking place in the United Arab Emirates since the 1990s is still actively growing every year. However, since all these farms are situated in an arid region with a hot desert climate and extremely sparse precipitation events, they mostly rely on groundwater for their viability, which consequently leads to aquifer resources depletion and terrain subsidence phenomena. Al Wagan, being located on the eastern part of the Abu Dhabi Emirate, is an expanding agricultural area where large amounts of groundwater are being extracted, with the geological background mainly consisting of alluvial and aeolian deposits.

The region was inspected for land subsidence phenomena with DInSAR and SBAS techniques in order to quantify the terrain movements over a larger spatial extent than permanent GPS stations could cover and provide accurate estimations of the displacement history and velocities. The available ascending and descending ENVISAT images over the area were used for this study covering the time span 2003 – 2010 and the SRTM DEM was used for the topographic phase removal.

Despite the temporal decorrelation in the sand dune areas that obstructed the interferometric processes the results over the coherent alluvial sediments revealed an extended rapid subsidence pattern which affected all the farming sub-regions of the study area. The extent and the magnitude of the deformation pattern was consistent in both the ascending and descending stacks and the cumulative subsidence over the period of observations exceeded 1 meter in some occasions.

Even though the time series of the displacements seem to have a linear trend in general an accelerating quadratic trend was also observed in specific areas. The rate of subsidence reached 18 cm/year in some regions during the sampling period, while a small seasonal effect was observed in the time series with the phenomenon peaking in the summer season.

The water levels of the monitored wells of the study area were highly correlated with the observed subsidence pattern, as they exhibited declination in the examined time interval. The cumulative groundwater table retreat was 30 meters on average, varying between 10 – 60 meters depending on the well locations. This study highlights the negative consequences and the potential hazards of the excessive groundwater exploitation and it will be continued with the use of contemporary C-band and L-band data.

Madrid and surrounding cities make one of the larger urban areas in Europe. Monitoring of large urban areas is a priority challenge in this century. Remote sensing and, more precisely, DInSAR are promising tools for long term accurate monitoring of world’s urban environment. This work presents the potential of Sentinel-1 SAR sensors in Madrid metropolitan area. A set of 35 S1 Interferometric Wide Swath SLC images were processed with SARPROZ software based on PSInSAR methodology. Available permanent GNSS stations in the studied area were used for validation. A first analysis evaluated the performance of PS density and quality in different land use classes present in Madrid urban area. Critical areas were identified in the deformation maps highlighting clusters of PSs with similar subsiding or uplifting tendency. These cases were studied and validated with field visits. Sentinel-1A allowed for the identification of critical areas in Madrid area, mostly related to the construction works (river works, urbanization areas, new structures, etc.) and some other ground-structure interaction processes. Radar remote sensing is evolving quickly, being ready for an almost continuous monitoring of complex urban landscapes.

Objectives

The main objective is to determine the long term deformation trend for two earthquake-prone areas close to the seismic region Vrancea in Romania: Buzău and Focșani. For this scope, we are exploiting the continuity between the ERS-1/2 and Sentinel-1 European Missions. The identification of linear deformation rates of urban areas is based on the Persistent Scatterers Interferometry (PS-InSAR) technique. The study compares recent linear deformation rates extracted from Sentinel-1 data with the linear deformation estimated from historical ERS data spanning 5 years. The historical dataset used to retrieve the deformation rates from the past consists of 31 ERS-1 and ERS-2 images acquired from 1995 to 2000. Pixel spacing of those images is equal to 7.9 m in range direction and 3.97 m in azimuth direction. The current deformation rates are estimated from Sentinel-1 images in TOPS mode (2.32 m range spacing and 13.94 azimuth spacing), between 2014 and 2016.  Digital elevation model generation step is included in the proposed processing chain. Algorithms are implemented using the interferometric and interferometric point targets analysis packages of Gamma Remote Sensing software. 

Dataset's Pre-Processing

A reference image must be defined for each dataset. This image is selected near
the center of dataset’s acquisition temporal interval, considering the minimization
of dispersion of dataset’s baseline values. Slave images must be resampled in
geometry of master image. This co-registration step must be precisely executed,
especially in case of images acquired by Sentinel-1, otherwise the specific
Doppler frequency shifts in azimuth may affect the further processing steps.
Initial offsets between master and slave images are estimated using the
trajectories of satellite’s orbits. Non-linear offsets are calculated by
computation of interpolation polynomials, whose coefficients are estimated by
calculation of amplitude’s correlation index

Digital Elevation Models Generation

A digital elevation model was generated for each of the two datasets. The DEM is estimated from two acquisitions with small temporal baseline (to minimize the effects of decorrelation) and with large perpendicular baseline (to increase the sensitivity of topographic phase component with terrain’s height).

The processing algorithm contains the following steps: interferogram computation, interferogram flattening, filtering of interferometric phase and phase
unwrapping. In case of Sentinel-1 images, interferogram computation is realized
without common band filtering in azimuth direction (because of presence of
Doppler frequency shifts). The filtering process is adaptive to interferogram’s
power spectral density. Phase unwrapping is computed using minimum cost flow
algorithm

Persistent Scatterers Candidates

Deformation rates are estimated in points which present stable electromagnetic proprieties in time, and in which the residual component of interferometric phase is not significant. The identification of Persistent Scatterers candidates was
implemented considering both amplitude’s statistics (mean per sigma ratio) and
spectral coherence

Linear Deformation Rates Estimation

The interferograms between each master-slave pair of the dataset are computed at the locations of identified PSs. The topographic component is adapted and subtracted from each interferogram. After the unwrapping of interferometric phase, a phase regression analysis is conducted in each PS. Considering the linear dependency of topographic component with perpendicular baseline, the DEM’s heights are re-estimated, and phase regression analysis step is re-iterated considering the refined heights. This analysis estimates the linear deformation rates and the residual components of interferometric phase.

Future Work

The Sentinel-1 images dataset will be extended for a better estimation of recent linear deformation rates. A comparison between the deformation estimated in the 20 years-old timespan (from ERS dataset) and current linear deformation rates will be conducted. Deformation maps will be geocoded.

The Dead Sea (DS) area is one of the best examples of the significant impact that the uncontrolled exploitation of natural resources may have on the territory and the environment. In the last decades, the DS territory faced a profound change as consequence of the human activities that interfered with the delicate equilibrium of its ecosystem. In the 1960s, the potash industries started to heavily exploit the salty-rich water of the DS, pumping huge amount of water in the desalinization ponds located in the southern section of the lake. Furthermore, most of the freshwater that was coming from the Jordan River and from the other main feeders was diverted for urban and agricultural purposes in the region and as far as the Negev desert, in southern Israel. This, in combination with the unique climate of the territory, characterized by high annual evaporation rates (1500 mm) and very low average annual precipitations (60 mm) resulted in a strong negative water balance that caused the water level to drop with increasing speed. The decline rate, calculated as 17 cm/yr in the period from 1930 to 1973, has reached 100 cm/yr and exceeded 120 cm/yr nowadays. In the last 40 years, the level dropped by 36 m (as of March 2017, the DS water level is at -431 m m.s.l.) and the lake shrunk by more than one-third. The consequent change in the hydrogeological settings of the entire basin caused the seaward and downward migration of the fresh/saline groundwater interface forcing the freshwater to flow through the underlying evaporite layers constituted mainly by salt and gypsum. Dissolution-related phenomena such as subsidence and sinkholes started to appear all along the DS shoreline bringing heavy damage to the territory, the infrastructures (bridges, roads, earthen dikes of the desalinization ponds) and the buildings (houses, hotels, resorts, factories). The most active subsidence occurs in the areas surrounding the Lisan Peninsula (LP), located in the southern part of the DS in Jordanian territory. The Peninsula is also characterized by uplifting areas mainly as consequence of the upward movement of the underlying Lisan salt diapir.

This work presents and analyses the results obtained from Differential Interferometric Synthetic Aperture Radar (DInSAR) techniques applied the monitoring of an area of about 18 km x 22 km that entirely covers the LP. The available SAR datasets consist of: 24 ERS-1/2 covering the period 06/1992-06/2000 and 31 ENVISAT covering 01/2003-06/2010, both acquired in C-band by the European Space Agency (ESA); 10 L-band ALOS PALSAR acquired by the Japan Aerospace Exploration Agency (JAXA) from 11/2007 to 02/2011; 20 X-band COSMO-SkyMed (CSK) for the period 12/2011-05/2014 acquired by the Italian Space Agency (ASI). Finally, we exploited 32 Sentinel-1A images acquired by ESA and covering the period 10/2014-05/2016. In addition, another Sentinel-1A dataset made of 30 images acquired in ascending geometry over the same period was processed and used to assess the predominant component of the movements in the area. The image processing has been carried out using the Small Baseline Subset (SBAS) technique. The removal of the topographic component of the phase was carried out using the Shuttle Radar Topographic Mission (SRTM) digital elevation model (DEM) with a resolution of 30 m x 30 m. All the datasets were multi-looked differently in order to obtain the same ground resolution with a pixel size of 20 m x 20 m. The areas along the shore that were exposed year by year by the DS lowering, were masked out in all the images using the -415 m m.s.l. contour line of the SRTM DEM, that refers to the water level in February 2000 at the time the DEM was produced. The deformation values calculated along the line of sight (LOS) of the satellites were projected to the vertical direction considering the incidence angle of each point measured with the different sensors. The novelty of this work comes from the integration of three different wavelenghts (X, C and L) to study the dynamics of a salt dome. Five displacement maps have been produced, carefully checked and then fused to provide a total cumulated map of the displacements in the area. The ground movements have been analysed by comparison with in-situ tectonic observations collected by various authors since the mid-1980s. The results show an increase in the displacement rates starting from the 2000. The uplift occurring in the north part of the peninsula is probably caused by the combination of different factors such as tectonic, diapirism, elastic rebound. Semi-circular depressions occur around minor uplifting areas in the southern part of the peninsula as consequence of the salt diapir upward movement. Furthermore, the study shows an episodic rising of the Lisan diapir. The Sentinel-1A satellite used in this study demonstrated its great potential as a tool for continuous monitoring activity over areas affected even by very fast displacements. The obtained results updated the knowledge of the complex karst dynamics in the Lisan Peninsula, and could be used as the starting point for further studies in the area.

Speckle tracking offers complementary information for InSAR motion monitoring as it can detect absolute motion (does not require unwrapping) in both line of sight and azimuth directions and can detect fast motion in areas that appear to be incoherent due to local misregistration; however, there is at least an order of magnitude difference in measurement precision between standard speckle tracking algorithms and InSAR; speckle tracking precision is on the order of decimeters whereas InSAR analysis can reach millimeter precision. In this paper we try out the following four speckle tracking algorithms in order to estimate disparity maps and test the limits of speckle tracking precision: correlation-based dense correspondence with a number of different correlation metrics (i.e. normalized cross correlation, mutual information, and the maximum likelihood estimator of patch similarity derived from the speckle noise distribution); a simultaneous maximum a posteriori based network inversion of offset correlations using probabilistic filtering techniques; a brute force few variable (range/azimuth velocities and topography) optimization using the sum of all pairwise normalized cross correlations in a small temporal baseline network as the cost function; as well as semi-global matching followed by network inversion with the L1 norm and a fully 3D oriented kernel implementation of the non-local means filter.

We demonstrate that for the latter algorithm we are able to get time series that match well with the InSAR analysis obtained on the borders of a fast-moving slope imaged with TerraSAR-X StripMap data, including some overlapping measurements. Our findings suggest that as the size of the correlation window shrinks (necessarily so, to reduce motion underestimation), the main factor that limits motion estimation precision is local distortions (on the scale of the correlation kernel) of the speckle pattern due to baseline geometry, surface cover, temporal decorrelation, severe atmospheric conditions, and other factors that affect the stability of the speckle pattern. This indicates that modeling the impact of these factors on sub-pixel scatterers may be required to make improvements to spatial-temporal resolution and motion estimation precision, in order to finally close the gap in coverage between InSAR and speckle tracking. To rule out the possibility of method noise being the limiting factor, as well as to compare differences between algorithms and make suggestions for future improvements, we test the algorithms on a suite of simulated motion fields applied to real data. We also discuss implementation details and some of the progress that we have made in operationalizing speckle tracking for monitoring fast motion.

The main purpose of this research is to examine the phenomena of land subsidence in the region of the Amyntaio basin of the prefecture of Florina (North Greece), due to overpumping of groundwater resulting from the activity of Amyntaio lignite mine, but also to examine the impact of the shutdown of older mines in the region in previously subsiding settlements.                                                                                                                                                                   The phenomenon of land subsidence can result from numerous causes while there are many examples in Greece as well as abroad. In the case investigated in this study, overexploitation of aquifers phenomena resulting from the activity of the Amyntaio lignite mine, lead to a drop in piezometric surface causing land subsidence.                                                                                                                    The study area is located on the west of the Amyntaio mine and the settlements included and also affected are Anargiroi, Fanos, Valtonera, Aetos, Pedino, Amyntaio and Xino Nero. In order to investigate the phenomenon, the technique of differential SAR interferometry (DinSAR) was applied. Also an estimate of the average annual value of land subsidence as well as the detection and mapping of the affected regions were achieved by the implementation of the Interferometric Stacking technique. Both methods were applied using ENVISAT satellite data for the period 2003-2010. The results of applying the above techniques indicated the existence of land subsidence with average annual value in the range of 5mm per year affecting the settlements located on the west of the Amyntaio lignite mine, showing a pattern of deformation which agrees with the locations of faults and the stratigraphy of the area. Also some settlements, which were previously subsiding, now appear to be stable or even showing some uplift phenomena. In the broader region also more land subsidence phenomena were detected but as they were not related to the mining activity they were not examined further. 

This paper is about the use of TerraSAR-X and Sentinel-1 data for the monitoring of the Historic Centre of Sighisoara, a World Heritage site that is part of the Romanian cultural patrimony. Built in the 13^th century, the town has faced both natural and anthropogenic events (i.e. earthquakes, floods, landslides, fires, epidemics, incursions) that shaped the landscape of the town and the construction style of the defence towers and the surrounding fortification wall. Nowadays, the ground and structural stability of urban areas can be investigated and monitored using synthetic aperture radar data. The state-of-the-art microwave remote sensing technologies enable the detection of millimeter-level displacements measured along the line-of-sight. The results of the study support the national and local authorities responsible for the protection of the cultural heritage by rising awareness and indicating the potentially unstable buildings that require a more detailed in-situ analysis.

Considered an "irreplaceable source of life and inspiration" (UNESCO, 2015a), the World Heritage contains cultural and natural properties that are of "outstanding universal value" from the point of view of history, art or science (UNESCO, 1972). An exquisite definition states that the World Heritage List "reflects the wealth and diversity" of our planet's cultural and natural heritage (UNESCO, 2008). Of high interest for numerous scientific disciplines, the World Heritage has been extensively studied throughout the years (Harvey, 2010). Both natural and cultural properties are important for the society, as they mutually influence each other in the sense that the human action has an impact on the environment, while the environment affects the creations of the humankind (Lowenthal, 2005).

Furthermore, the heritage sites have a major significance for the local community. The bond between heritage and community represents a bridge that connects the past and the present and it symbolizes the legacy that will be passed on to future generations. A recent study (Croitoru and Becut, 2014) shows that the link between heritage and community "generates values like identity, tradition, the feeling of belonging, social cohesion". Moreover, the same study reveals that the local communities acknowledge the benefits of heritage that primarily reside in tourist development followed by the safeguarding of the local identity and traditions and knowing of the past. Likewise, another study (Mydland and Grahn, 2012) underlines the critical value of heritage in building and cultivating the local identity.

The undeniable process of climate change increases the vulnerability of the World Heritage sites and the complexity of the geological and geomorphological mechanisms that need to be properly understood in order to limit site deterioration (Howard, 2013).  Considering the long-lasting efforts carried out to protect and preserve the cultural and natural properties, a special emphasis is put on heritage management and sustainability (Keitumetse, 2014).

In this context, the knowledge of the current state of conservation is essential. Starting with 2013, UNESCO has implemented an information system that contains geospatial data, reports and threats for each World Heritage site (UNESCO, 2015c). Suitable for both cultural and natural heritage, satellite data enables the generation of custom-made monitoring products, such as land use and land cover maps, land cover change maps, multi-temporal analysis and change detection, digital elevation models, displacement maps, etc. At this moment, the great advantage of using Earth Observation for World Heritage resides in the monitoring of large areas with a high or very high level of detail, depending on the spatial resolution of the satellite images. The type of satellite data is also very important. Optical multispectral imagery provides information regarding the changes in landscape or in the World Heritage site itself, while the outputs derived from synthetic aperture radar (SAR) data enable the detection and measurement of ground and structural displacements/deformations. In addition, the fusion between optical multispectral and SAR satellite data significantly improves the quality of the monitoring results (Stramondo et al., 2006).   

The use of Earth Observation satellite data for heritage monitoring was intensively promoted by the European Space Agency (ESA). In 2003, UNESCO and ESA signed the "Open Initiative on the Use of Space Technologies to Support the World Heritage Convention" that has the goal to protect, monitor, document, present and share the World Heritage sites (UNESCO, 2015d). Additionally, the initiative aims to support policy making and governance and to contribute to the development of the digital heritage concept. Within the initiative, the term "space technologies" refers mainly to Earth Observation and secondly to other technologies such as navigation, positioning, communication, etc.

Nowadays, ESA and the European Commission (EC) provide access to the data acquired by the recently launched Sentinel-1A satellites, in the framework of the Copernicus Programme (ESA, 2015). Sentinel-1A is equipped with a C-band SAR that has the capability to cover a 400 km wide swath from a near-polar, sun-synchronous orbit. Numerous World Heritage sites have been successfully monitored using space technologies. Examples include, and are not limited to, the assessment of the climate change impact on the Great Barrier Reef (Bouma et al., 2011), the identification and monitoring of the land subsidence/uplift in the Venice Lagoon (Teatini et al., 2005), the modeling and analysis of the Great Wall (Li et al., 2008), (Chen et al. 2010), the prospection of monitoring of the Nazca Lines (Tapete et al., 2013), and so on. 

This work focuses on monitoring the ground motion and infrastructure stability in urban environment in Romania, with the aim of promoting and increasing the use of SAR data in geotechnical engineering applications relevant to measurements of ground subsidence/motion in diverse areas from Romania. Convincing results for very local areas will raise interest for applications in monitoring local breakdowns in infrastructure that could lead to environmental disasters. The application of satellite surveying in the field of civil engineering represents a niche research in Romania and, although several studies have been published, the remote methods are not common practice.

A number of four sites in Romania that were monitored using Sentinel-1 data will be presented. PSInSAR deformation maps and profiles derived from the Sentinel-1 SLC data between October 2014 – October 2016 were computed, together with an interpretation report based on analysis of the PSI results, geological maps and in-situ measurements/field visits, where available.

The first studied site is Bucharest city, the capital city of Romania, located in the southeast of Romania and covering an urban area of about 285 km2. Due to its position on the banks of Dambovita River and high underground water levels, the risk of subsidence in the area is significant. Moreover, its closeness to Vrancea seismic area increases the risk of seismic induced deformation in the area.

Bucharest is a fast developing city with the average construction rate of 8-20% new buildings with respect to the existing ones. Consequently, the civil engineering industry faced new challenges related to the need of having taller buildings with deeper underground levels, a developing network of subway lines and more bridges with large diameter piles’ foundations. All these new works have an important impact upon the upper ground stability.

The next two sites are the cities Buzau and Focsani, located in the South-East and East of the Carpathian Mts. Curvature. The instability phenomena can be favored by the variations of the hydrostatic level, the extraction of some commodities in the ground or by the influence of the earthquakes occurring in the neighboring zones, in the Vrancea region. It is necessary to also take into account the fact that all these processes take place on a background of general subsidence of the whole area, caused by the still active tectonics.

The fourth site is the city of Constanta, located immediately at the South of the Capidava-Ovidiu fault (which separates the Central part from the Southern part of Dobrogea), placed on a structure which is characteristic for the Southern Dobrogea. The surface sedimentary deposits belong to the Pleistocene: greenish and reddish clays with gypsum concretions in the base, sandy reddish clays with calcareous concretions, then loessoid deposits at the upper part. Loess thickness can reach 40 m in some parts and contains several levels of fossil soils (palaeo-soils). Due to its characteristics of being highly compressible and soluble, the loess is a risk factor of the ground. Another risk area is the Black Sea shore, which can be affected by processes of mechanical abrasion and solubilization of the sea cliff, leading to cliff collapse and suffusion zones.

The results of the PSI analysis were carefully studied and a preliminary correlation of the areas identified as being affected by subsidence with possible causes is presented.

Part of the presented results were obtained within the research project URMA (Multi-layer Geohazards Information System Concept for Urban Areas), founded by ESA through the 3^rd Call for Outline Proposals under the Romanian Industry Incentive Scheme programme, between November 2015 - November 2017.

New generations of radar satellites (e.g. TerraSAR-X, Cosmo-SkyMed, Sentinel-1A/B) with short repeat-pass cycles and high spatial resolutions have enhanced the capabilities for acquiring data over large areas shortly after major landslide events and for monitoring landslide activity at regular intervals. Over the past two decades, several studies have demonstrated the potential of differential synthetic aperture radar interferometry (DInSAR) for detecting and quantifying land surface deformation. The main challenges of the DInSAR technique include spatial and temporal decorrelation, an accurate estimation of the phase ambiguity in the phase unwrapping step and the presence of atmospheric artifacts. Permanent Scatterer Interferometry (PSI) and Small BAseline Subset (SBAS) are widely used to extract the phase information of the displacement component and mitigate the negative effects of the errors sources in the interferogram stack. The PSI technique is based on the pixels with a dominate radar backscattering in comparison to the background (i.e. high coherence over time). The PSs dependency of the PSI technique and lack of sufficient PSs in the natural terrains have limited the abilities of PSI in vegetated areas. For the vegetated areas, which is considered nearly as Gaussian scatterers, high temporal decorrelation caused by vegetation and reliable phase unwrapping are the main challenges of the surface deformation estimation within a given period of time.

In this study, we present preliminary results of SBAS processing with Sentinel-1 IW data for monitoring the activity of two landslides located in the Alps: Corvara in Badia in the Autonomous Province of Bolzano-South Tyrol (Italy) and Schmirntal in Tyrol (Austria). Both study sites have displacement rates in the order of centimetres to meters per year. Two different approaches are tested using datasets acquired between 2015 and 2016. First, multi-looking data is processed to increase the Signal to Noise Ratio (SNR) and the reliability of the coherence estimation. Secondly, using the single look data to identify the isolated SBAS pixels (if surrounded by completely decorrelated pixels), which is limited when applying the multi-looking process due to the resulting coarser spatial resolution. For a more reliable phase unwrapping process particularly in areas with low coherent, in addition to the standard 2D unwrapping, a 3D phase unwrapping approach is tested. Remaining discontinuities caused by phase jumps in the unwrapped interferogram, are corrected by discarding the interferogram pairs showing low coherence while optimizing the adaptive filter strength. After the first estimation of the residual topography and displacement rate, the best fitting models corresponding to the landslide displacement behavior is selected for reprocessing the interferogram stack for improving and refining the final velocity rate. Finally, after applying low pass spatial and high pass temporal filters, the final displacement map is geocoded and is compared to the ground dGPS measurements. The results shows that the negative effects of the temporal decorrelation caused by the vegetation is still tangible and visible on the displacement map. We expect that in the further analysis, the higher frequency of the data due to the availability of additional Sentinel-1B data will reduce errors due to temporal decorrelation and allow describing the displacement rate.

Long-term monitoring of temporal gravity field and ground water level changes in Iran and its associated ground subsidence seen by geodetic methods are important for water source and hazard management.The high-rate (cm to dm/year) ground subsidence in Iran has been widely  investigated by using different geodetic techniques such as precise leveling, GPS and interferometric synthetic aperture radar (InSAR). The previous individual SAR sensors (e.g. ERS, ENVISAT and ALOS) or multi-sensors approach have successfully shown localized subsidence in different parts of Iran. Now, thanks to freely available new SAR sensor Sentinel-1A data, we aim at investigate further the subsidence problem in this region.

In this ongoing research, firstly, we use a series of Sentinel-1A SAR Images, acquired between 2014 to 2017 to generate subsidence-rate maps in different parts of the country. Then, we correlate the InSAR results with the monthly observations of the Gravity Recovery and Climate Experiment (GRACE) satellite mission in this region. The monthly GRACE data computed at CNES from 2002 to 2017 are used to compute the time series for total water storage changes. The Global Land Data Assimilation System( GLDAS)  hydrological model (i.e. soil moisture, snow water equivalent and surface water) is used to estimate Groundwater changes from total water storage changes obtiaend from GRACE data.

So far, we have generated a few interferograms, using Sentinel-1A data and SNAP software, which shows a few cm subsidence in western Tehran in last 2 years. We will try more Sentinel images for this area to better constrain the rate and extent of deformation and will continue InSAR processing for the rest of the country to localize the deformation zones and their rates. We will finally comapre the rates of subsidence obtained from InSAR and the rate of groundwater changes estimated from GRACE data. 

We produced Sentinel interferograms of Etna using three different methods, DIAPASON on the GEP-TEP, SNAP and GAMMA. The possibilities of tuning and filtering differ from one to the other software. The final wrapped interferograms are almost similar globally with differences that can be observed at the local scale. For interferograms over time periods of 6, 12 and 24 days, the coherence is variable spatially, and good in general, but it can be significantly variable from one couple to another. This is probably due primarily to the ground conditions. Indeed the expected DEM errors are not large thanks to the good orbit control and good quality DEM, and they are correlated from one interferogram to another, and the tropospheric differential heterogeneities below the 1-km wavelength cannot be very large. At the scale of 1-5km tropospheric effects are visible in many of the interferograms, with different patterns from one couple to another, and only part of them is correlated with the topography. The correlation between elevation and phase gradients is also variable spatially, and a correction estimated over central Sicily, away from the deforming zones of the volcano, does not allow necessarily to reducing the tropospheric effect over the volcano. Even at the scale of the volcano the regression function can be variable. During and after the crisis of December 2015, several faults are activated in the east flank of the volcano, and at different epochs. Sentinel permits to map with high accuracy the motion on those faults. The comparison with GNSS is very difficult due to the fact that the area cannot be covered densely enough with campaign points to fit with the small size of the discontinuities, and for the temporal aspect, because the GNSS data for the Sentinel period are not available to the community. The medium-long term analysis of those movements requires combining the interferograms in a time analysis and this requires a methodology to combine the couples. As unwrapping can be sometimes difficult over low coherent areas, with the risk of cycle slips, we investigate analysis methods based on the use of the original wrapped interferograms. Although popular, unwrapping is not a natural and not necessarily an easy way to handle interferograms in low coherence areas, especially when the phenomena to tackle can be time variable and take place over long periods of time as it is the in often the case over volcanoes and fault zones. We compare the quality of our results over Etna with that of the interferograms obtained with similar approach over the Corinth Rift Near Fault Observatory.

The urbanization of La Verbena, located in the municipality of Arcos de la Frontera in the south of Spain, has suffered the effects of being located on the top of the hill of an active slope. Ground movements apparently originated by an active landslide affect buildings on the area. Since the beginning of the decade geotechnical studies and reinforcement actions have been carried out by the local authorities such as armed injections and placement of drains in the area. All these activities have an important social and economic impact on the region.

In this study, we highlight the benefits of the detailed ground deformations measurements, which can be measured to Sentinel-1. In our case we have processed ascending and descending large datasets between January 2015 and July 2016 for mapping precisely the current state of the buildings and of the deformations on the slope. The very high density of the measurement point retrieved by the CPT processing, the detailed time series showing the accelerations in time and the horizontal and the vertical components of the motions retrieved by decomposing the two modes, helped in the interpretation and understanding of the dynamics of the geology of the area and to establish any potential implications regarding the activation of the slope (like for instance heavy rain episodes as occurred in 2009). This paper demonstrates that PSI measurements obtained by processing Sentinel-1 data can be a powerful forensic tool for the diagnosis of singular infrastructures. 

Problematic :

Water erosion is a natural and complex process: it concerns the detachment and the transport of soil particles by water, in a repository, which can be enhanced through anthropogenic activity by creating favorable conditions for the runoff and the cultivation of the soil. Erosion by gully is one of the most spectacular forms of water erosion, gullying is a mechanism of incision of rocks under the effect of concentrated runoff, and depending on the size of the incision , we speak of claws or erosion channels, in the case of discontinuous or continuous incision of a few decimetres in width and depth we speak of ravines or gullies. On a global scale, erosion by gullying ranges from 10% to 94% of total sediment production caused by water erosion (POESEN et al., 2003), in the same context another study carried out by POESEN et al, in 2002 on some Mediterranean watersheds, shows that the presence of active gullies in these sites seems to be a good indicator of the importance of sediment production through these basins. In other words, gullies are connectivity elements that facilitate the transfer of sediments into the landscape. The gullies can thus be responsible, with the wad for the detachment and transport of a large volume of sediments from the deep horizons to the reservoirs (POESEN et al., 2003). Similarly, the Maghreb region has also benefited from several studies on erosion processes and their damage (ROOSE ET AL, 2004, 2006a, 2006b, 2008, 2010, 2012). B. HEUSCH's work has shown that groundwater erosion is less important than gullying and ablation by rivers, the slope has less effect than the topographic position and the exceptional saturating showers play a major role in Solid transport (HEUSCH, 1970a, 1970b and 1986). Therefore, among the most spectacular consequences of erosion by gullying is the siltation of dams and that occurs as a result of the transport of alluvial deposits downstream, which accumulates in the reservoirs causing a decrease in their regularization capacities . This phenomenon has reached an annual rate of 50 million m3, or 0.5% of the total storage capacity of dams in Morocco (HCEFLCD-2003). This work concerns the silting of the Sidi Mohammed Ben Abdellah (SMBA) dam located at the Bouregreg watershed (Morocco).

Objectives:

The objective of this work is to determine the proportion of sediments from the banks of the lake to the silting of the dam, this work is divided into two sub-objectives. The first concerns the morpho-dynamic characterization of gullies and the quantification of eroded volume at the banks of the lake, the second aims to estimate the rate of sediment inputs arriving at the lake.

Methodological:

The methodological approach proposed to establish the sediment balance of the dam SMBA, is based on the determination of the volume eroded and the sedimentary inputs that arrive on the lake. The determination of the eroded volume can be done using photography with a resolution to the centimeter taken at low altitude by a drone, in order to create Digital models of diachronic terrain and measure the volume of sediments from the gullies. Then, we will compare the volume determined by the drone with the volume that will be determined by the radar interferometry.

- Radar: with the RADAR Images (Sentinel-1), we thought to use these data that are available and free to calculate the volume of the eroded soil by the system of gully erosion, and then compare it with the results obtained by the drone. Principle: Before putting in place our hypothesis we must mention that the source of the sediments in the gullies comes from the slopes. After the calculation of a few interferograms, we found on the slopes exposed to the direction of the satellite vision, that there is always a strong coherence even if slopes are eroded, we can explain this strong coherence by the presence of the slope; So the second objective of the thesis pushed the limits on using the method DInSAR and applying it for the calculation of soil displacement especially that there is always a strong coherence on the eroded zones.

Karst aquifers are one of the most complex form of groundwater systems and one of least understood and monitored water bodies. In Qatar, the slow process of karstification is associated to paleo-groundwater movement and heavy precipitations that occurred during the Middle Pleistocene in the fractured, calcareous, dolomitic and gypsiferous Dammam Formation (Eocene). The setting offers a unique opportunity to understand karst formation and associated surface subsidence associated to subsurface groundwater flow. The resulting formations of sinkholes and depressions are widespread along the NE-SW and NW-SE axis, with a mapped system of fractures. The above allow a rapid groundwater recharge following the scarce flash rain events in the peninsula of Qatar. Moreover, in hyper-arid areas, such as Qatar, evaporation and evapotranspiration substantially constrain the recharge of the aquifers, which occurs during the sparse and brief precipitations. Hence, Qatar has limited groundwater reserves of fresh water suitable for human use. Drinking water is exclusively produced from desalination. Although in the North of the country, some fresh groundwater is still used for the limited agriculture but its abstraction is localized in unsupervised private wells. Hence, groundwater quality is declining at a fast rate and aquifers become unsuitable for irrigation. Groundwater in Qatar shows a complex hydraulic conductivity between different heterogeneous karst aquifer systems in which the lack of reliable data limits the possibility to model its groundwater flow and estimate the water budget.

 

In the karst aquifers, ground subsidence is generally associated to the karstification of the limestone and in Qatar, to the dissolution of the evaporites. We use Synthetic Aperture Radar Differential Interferometry (DinSAR) to detect the ground subsidence over active hydromechanical and hydrochemical erosion, which could cause an increase of the hydraulic gradient. The karst-induced subsidences observed by DinSAR Time-Series are then validated by GPS Time-Series, field validations through all the Qatar peninsula including using low-frequency Ground Penetrating Radar mapping of the sub-surface structural anomalies associated to the ground deformations in specific sites with high deformation rates. The hyper-arid environment of Qatar and the very flat topography offers a unique study site for our approach with a minor topographic phase contribution and an exceptional interferometric coherence allowing for instance to produce three-years and one-year period interferograms respectively for the ALOS-Palsar-1 L-band and Sentinel-1A C-band satellites.

 

From ALOS-Palsar-1 and Sentinel-1A single interferograms, we suggest a ground subsidence of one and five centimeters per year (in Line Of Sight with the satellites) respectively for the 2007 to 2010 and 2015 and 2016 periods, for an extent of 50x10kilometers located in the South of Qatar, between two fractures, which trend NNE. At this exact location, situated near a large agriculture farm, the ground displacement is not associated to groundwater depletion but to a localized water table rise of 10 to 20 meters for the 1980 to 2009 period. From these preliminary results we suggest that the use of sewage-treated water for farming in this region of southern Qatar could potentially acts like an artificial recharge into the karst aquifer, hence increases the hydraulic gradient and likely increases the hydrochemical erosion of the limestone and evaporites. Consequently, subsidence is likely to occur. The two observations made from two different satellites are similar in size and amplitude, and spatially correlated to the rise of the groundwater levels. The results motivate the continuation of this work with DinSAR Time-Series extended for the whole country of Qatar. We believe that this DinSAR monitoring is crucial to improve the fidelity of Karst groundwater models, better understand the local surficial geology and potentially to forecast the occurrence of sinkholes.

Abstract: The phenomenon of land subsidence has made numerous challenges for agriculture, residential areas, roads and irrigation canals and other man-made imitation in recent decades. Excessive withdrawal of ground water, in recent years, due to climate change in many parts of the country for instance, Border of Uremia Lake that have been evaporated too much. In addition, extensive use of water in incorrect agricultural lands has had multipliers/ double effects on subsidence in these regions. The purpose of this study is to determine areas that have been affected, and finally, to estimate the subsidence of InSAR as a reliable method to measure accurately the changes on the earth's surface. This method is used to indicate the broad coverage and high spatial resolution with very high accuracy in this regard. Moreover, ENVISAT satellite radar images were used during 2009-2010. Accordingly, the results showed that the maximum subsidence is 6 cm and the minimum subsidence is 3 cm yearly in Tasuj plain. 

The Afar Depression, in the northeastern part of Ethiopia, offers unique opportunities to study the
transition from continental rifting to ocean floor spreading. This process, which is the outcome of
tectono-magmatic events, has been under investigation using different geophysical and geological
techniques. The current study mainly focused on GPS and InSAR methods to analyze the inter-dyking
deformation pattern along the Dabbahu-MnadaHararo (Afar) rift segment. The InSAR data was used to
identify the time of dyke events and the GPS data to study the inter-diking deformation pattern.
Asequence of 12 dyke events occurred from June 2006 to June 2009 and based on the GPS data two
major pattern of deformation have been identified. These are being categorized into before and after
June 2009 dyke events. Most of the GPS stations before June 2009 showed larger displacement rate
whereas after June 2009 intrusion, the displacement in most of the stations was relatively smaller. Even
though the deformation process is still active, sites such as DAFT, DA45, DAYR and DATR indicated that
the displacement rate is relatively stabilizing in the post seismic relaxation period. Moreover, sites such
as,DA25 and DA35, had large offsets in their time series right at the time of dyking events, which is an
indication of major deformation due to the rifting process. Except the distant sites DA60 and DASM all
the other stations were mostly affected by the dyke intrusion, such that there was an offset in the data
during the dike intrusion. Stations DA25, DA35, DA45, DA60, DAFT and DAYR that are located in the west
side of the rift, where the 2005 diking event took place, showed displacement as large as ~84mm/yr,
~53mm/yr, ~46mm/yr, ~17mm/yr, ~23mm/yr,~17mm/yr and ~6.36mm/yr towards the west direction
respectively. Sites DASM, GABH and DABB have a displacement towards the North East direction. The
site GABH, which is situated in Gabh’o volcano, shows rapid inflation from January to June (2006) and
continued with a slow uplifting till February 2007. Inflation began in June 2006 in the composite volcano
DABB site while subsidence that amount of about 16mm/yr is observed in Semera station called DASM.
Key words: Afar, Deformation, Inter-Dyking, InSAR, GPS, Rifting

The town of Vauvert, France, hosts since 1973 a mining activity of the Oligocene salt deposit located mainly between 2500 m to 3000 m deep (Valette, 1995). The salt is extracted by injection of water into a well. The pressure pushes the product of the salt dissolution (i.e. brine) to the surface through another well.

 

Since the beginning of the production, millions of tons of salt have been extracted from the ground (about 1 million tons per year at current annual production). The compaction of the salt cavities induces a surface subsidence of about 8 km in diameter with a maximum vertical deformation of about 2 cm/yr (Raucoules et al., 2003).

 

After a few years of non-production, old wells are over pressurized and must be purged. The first continuous purge began by the end of November 2016 and is expected to last about 30 years. This purging activity should modify the observed surface subsidence (in terms of magnitude and spatial extent), because:

-       The volume of salt produced from the purged wells is added to the daily produced volume, so an increase in total production volume is observed,

-       Unlike the conventional injection / extraction activity, the volume of brine is not replaced by water, since the purge consists of a depressurization of wells only.

 

Levelling has been historically used to estimate and quantify the effect of the mining activity on the long-term subsidence. Levelling benchmarks, including national levelling benchmarks and local network are measured once a year by IGN (Institut Géographique National).

 

We use several measurement techniques to improve the knowledge of the effect of the mining activity on the surface deformation. To do so we have:

-       Multi-temporal InSAR: PS-InSAR, SBAS-InSAR and combination, StaMPS software (Hooper et al., 2004) on Envisat and Sentinel-1 SAR images,

-       GNSS (4 permanent stations), daily solutions, time series analysis,

-       Micro-gravimetry, successive measurements with a Micro-G Lacoste FG5 on a well slab.

 

Separately, each technique has its own limitation. Atmospheric artefact for InSAR, multi-path for GNSS, lack of knowledge of all components (masses) measured by absolute gravimeter and single reference station for levelling are some well-known, biases that can lead to high level of uncertainty. The main objective is thereby to mix the techniques mentioned above to:

-       Densify the displacement fields and reduce their uncertainties,

-       Improve knowledge of the long-term subsidence,

-       Detect transient signals possibly due to (1) spatial and temporal displacement of salt extraction and (2) well purges.

 

The use of both Envisat and Sentinel-1 SAR images allows us to study the subsidence induced by the mining activity from 2003 to date, in spite of data gaps. Envisat dataset spans about 10 years thus allowing for a reliable long-term analysis. On the other hand, the high temporal resolution of Sentinel dataset gives a unique access to the determination of transient ground deformation.

 

Utilization of InSAR, GNSS, levelling and micro-gravimetry datasets allows discussing their global consistency and the advantages of each of these different datasets for reservoir monitoring. Levelling analysis is compared to Envisat results for an improved long-term analysis. Otherwise GNSS and Sentinel time series are compared for a more detailed analysis of transient ground deformation. Finally, micro-gravimetry measurements bring valuable information about local effect of deep mass displacements, and consequently help us to link production to ground deformation.

 

Thus, this study is divided into  (1) a long-term spatial and temporal deformation analysis and (2) a focus on transient aspects of the deformation. These results are eventually compared with the time evolution of the salt production.

 

 

Volcán Descabezado Grande is part of the Descabezado Grande-Quizapú Volcanic Complex, located in the Maule region - Chile at 35°35'11”S, 70°44'38” W. It is a late Pleistocene to Holocene andesitic-to-rhyodacitic stratovolcano with an ice-filled summit crater of 1.5 km diameter. The volcano reaches a height of 3830m with a basal diameter of 11km (Benavente Zolezzi, 2010).

 

Descabezado Grande has been active for the last 300k years, being its volcanic products andesitic to rhyodacitic lavas, tefras and ash-flows reaching up to 7km in length. Last eruption occurred in June 1932, shortly after the great eruption of the neighboring volcano Quizapú. During this eruption, a crater of ~900m diameter, known as “Respiradero”, opened in the northern flank, generating an ash plume several kilometers high and 8km length pyroclastic fall deposits. Recent activity is weak, consisting in a few fumaroles detected at Respiradero crater during 2009, as reported by OVDAS-SERNAGEOMIN (http://www.sernageomin.cl).

 

In this contribution, we present the first results obtained by DInSAR time-series processing of the described area. We analyzed 17 L-Band ALOS/PALSAR stripmap scenes, acquired between January 2007 and December 2010. We computed 39 differential interferograms, and combined them by using the well-known Small Baseline Subsets (SBAS) technique (Berardino et al., 2002). Topographic phase removal was performed with the SRTM arc1 Digital Elevation Model (Jarvis et al., 2006).

 

Our results suggest that, between 2007 and 2011, the Descabezado Grande presented a deflation process of ~3cm/yr. The observed pattern is oval shaped and extends to the north of the volcano edifice. Maximum deformation is located between the summit and the Respiradero craters.

 

 

References

 

Benavente Zolezzi, O.M., 2010. Actividad Hidrotermal Asociada a los Complejos Volcánicos Planchón-Peteroa y Descabezado Grande-Quizapu-Cerro Azul, 35° sy 36^os, Zona Volcánica Sur, Chile. Universidad de Chile, Santiago de Chile.

Berardino, P., Fornaro, G., Lanari, R., Sansosti, E., 2002. A New Algorithm for Surface Deformation Monitoring Based on Small Baseline Differential SAR Interferograms. IEEE Trans. Geosci. Remote Sens. 40, 2375–2383.

Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2006. Hole-filled SRTM for the globe Version 3. CGIAR-CSI SRTM 90m Database: http://srtm.csi.cgiar.org.

Mitidja is a crucial region of 1400 square kilometres situated in the central north of Algeria. With its fertile lands, moderate climate and strategic geographic placement, the region across history attracted many populations. Nowadays it is considered among the most economical, populated and vital areas.  Many devastating earthquakes stroked this region through history; the last one was in May 2003, known as Boumerdes earthquake which generated 2278 killed persons, 10147 injured, 15000 homeless and much structural degradation (buildings, bridges and roads). This permanent seismic activity is caused by the African – Eurasian fault which traverses a major part of the region.

The paper focuses on two aspects. The first one gives a brief description of a project dealing with Mitidja region. The second one deals with Boumerdes earthquake (21 May 2003) in which we compare faults observed on ground with those obtained from InSAR images.

By the mean of SAR interferometry, we plan to locate areas prone to landslides and monitor the ground surface deformation in order to provide early warning of potential disasters. Also, a linked objective is the close observation of critical public infrastructures which are vital for the post disaster management rescue operations. Through this permanent monitoring of the Mitidja region, we intend to characterise particular landslides which could compose new undiscovered seismic faults. In this project, we plan to use Permanent Scatterers Interferometric Synthetic Aperture Radar (PS-InSAR) technique to detect and measure ground movement. With this tool and high-resolution of TerraSAR-X/Tandem-X data, we aim to track land subsidence precisely around strategic buildings (vital constructions, dense populated areas, bridges, highways, dams, etc.). For a good appreciation of the InSAR results, a validation procedure is envisaged to compare these outputs to some available infrastructure’s measurements related to classical land monitoring system such as levelling or GPS.

The second aspect of the paper deals with the attempt to find good coherence between ground deformations generated by Boumerdes earthquake (the strongest seismic event felt in the region since 1716) and those observed and measured on site. Many difficulties are exposed here relating with this issue. The mapping of the  coseismic surface displacement field caused by the earthquake was obtained using ENVISAT ASAR (IS2) and RADARSAT standard beam (ST5) data. The result allows obtaining coseismic interferograms from both the ascending and descending orbits of ENVISAT satellite.

The ESA Sentinel-1 mission, providing SAR images with a small revisit time, smaller spatial baseline and high coherence gives an enhanced capacity to investigate deformation phenomena using Differential interferometry (DInSAR) technique. In this study, a set of 12 C-band Sentinel-1 Stripmap mode images of the Barren Island, India are used to investigate the surface deformation. Barren Island, a part of the Andaman Islands is chosen as the study area. The Island (12.278°N, 93.858°E) situated in the Andaman Sea, has a circular shape with a diameter of about 3 kmand is mostly uninhabited. The Island is 135 km North-East of Port Blair. The area is considered to be affected by the only active volcano in South-East Asia whose latest eruption occurred in September 2010 and which continued up to January 2011. Apart from this, there were extended periods of eruptions from December 2013 to February 2014. The area has encountered about 10 volcanic eruptions since the first eruption in 1787.

This research work shows an approach for successfully integrating the SNAP toolbox based DInSAR and StaMPS algorithm, providing an opening for the advanced time series deformation analysis of the new plentiful Sentinel-1A interferometric data sets. The deformation is estimated using PS-InSAR processing in StaMPS environment, thereby taking advantage of better look angle error estimator and 3D phase unwrapping. During validation, the DInSAR technique is found to be inefficient in estimating look angle error and in performing 2D phase unwrapping. On the contrary, the integration of SNAP with StaMPS shows better estimation of look angle error and more accurate 3D phase unwrapping, proving the superiority of the integration of SNAP with StaMPS over SNAP alone. The results obtained by the PS processing shows subsidence in the region encompassing lava delta and its path.

In the past few years, SAR Interferometry specially InSAR and D-InSAR were extensively used for deformation monitoring related applications. Due to temporal and spatial decorrelation in dense vegetated areas, effectiveness of InSAR and D-InSAR observations were always under scrutiny. Multi-temporal InSAR methods are developed in recent times to retrieve the deformation signal from pixels with different scattering characteristics.Presently, two classes of multi-temporal InSAR algorithms are available- Persistent Scatterer (PS) and Small Baseline (SB) methods. This presentation discusses the Stanford Method for Persistent Scatterer (StaMPS) based PS-InSAR and the Small Baselines Subset (SBAS) techniques to estimate the surface deformation in Tehri and Nainital township of Uttarkhand state, India. Tehri town is inhabited near Tehri Dam (260.5 m high), one of the largest in India, at the union of Bhagirathi and Bhilangana rivers. Previously, several field investigations have found the localization of landslide along the reservoir region. However, Nainital town, in lesser Himalaya, a popular tourist destination, is spread around the periphery of a 1.4 km long kidney shaped lake, surrounded by high hill ranges. The township has a history of major landslide due to inadequate drainage management system, seismically active Main Boundary Thrust (MBT) passing from the area, unwanted construction on highly unstable slopes and low insitu strength of rocks. For both study areas, both PS-InSAR and SBAS approaches used multi-temporal Envisat ASAR C-Band images for generating single master and multiple master interferograms stack respectively and their StaMPS processing resulted in time series 1D-Line of Sight (LOS) mean velocity maps which are indicative of deformation in terms of movement towards and away from the satellites. From 1D LOS velocity maps, localization of landslide is evident along the Tehri dam reservoir rim area which was also observed in the previous studies. However, in Nainital, generated deformation patterns closely follow the landslide hazard zonation map prepared by state disaster agency. Both PS-InSAR and SBAS effectively extract measurement pixels in the study region, and the general results provided by both approaches show a similar deformation pattern. Further, we conclude that StaMPS based PS-InSAR method performs better in terms of extracting more number of measurement pixels and in the estimation of mean Line of Sight (LOS) velocity as compared to SBAS method.

Differential synthetic aperture radar interferometry (D-InSAR) has become a useful technique for monitoring ground movement. The technique enables the analysis of very small ground movements in continuous, large areas and has the advantages of high accuracy, high resolution, all-weather adaptability, low cost and inaccessible area coverage. Thus, D-InSAR has been widely used in the investigation of geologic hazards, such as subsidence, landslide, earthquake, and volcanic activity. In this research, D-InSAR technique was applied to detection of unstable slopes and determining moving displacement rate. For this mean, eight SAR images of PALSAR sensor of ALOS satellite were selected for processing based on D-InSAR approach. Obtained results were validated by field observations prates. This paper is only represented results related to image pair processing of 5^th July to 5^th October 2007 with 92 days interval. Garm Chay basin with 940km² area is located in North Eastern and 40km far from Meyaneh city in East Azarbijan province, Iran. This region with 380 landslides is considered as one of the unstable landslide proven area in East Azarbijan. Occurrence of these slides caused heavy damages to rural properties and arable lands. As a result, Sovin village in north western part of this basin was moved to other safesite. The results clearly show that some old stabilized landslides are still active. Because of their proximity to local stream networks (check the possibility for stream displacement), they can be considered as important source of sediment yield. Before mentioned period, the maximum displacement was calculated 5.8cm in landslide surface near to Avin, Atajan and Benavaran villages.

Persistent Scatterer Interferometry (PSI) represents a powerful tool to detect and measure surface displacements with millimetre accuracy and also to reconstruct the deformations history through displacement time series analysis. In many cases, PSI data are combined with the geological, geotechnical and hydrogeological conditions of the investigated areas providing substantial information for the interpretation of the land subsidence phenomenon.

The site under investigation is located at the widernorthern suburban Athens area (Oinofyta - Schimatari - Oropos). The area is used for main and secondary residence and at the same time is subjected to industrial development and intensification of agricultural activity. Available land motion mapping data, produced by PSI (Persistent Scatterer Interferometry) analysis, revealed substantial vertical displacements, corresponding to land subsidence caused by intense groundwater extraction. The deformations rates, based on previous studies show that, during the period from May 1992 to December 2000 have reached the maximum values of -15 to -20mm/yr. Differential ground deformations can trigger damages on structures as well as on linear and point infrastructures (pipeline and road network deformations, well-casing failures and protrusion etc.). The detection of the phenomena at an initial stage is extremely important, as further extension of the affected area and damages on settlements and infrastructure can be prevented.

At the current research study interferometric results from 2002-2010 by ENVISAT satellites has been processed and analyzed to investigate spatial distribution and patterns of the land motion, attempting to validate them according to the local geological conditions.

 

Land subsidence simply defined as a gradual differential settling of the Earth surface to a lower level. Some of the specific causes includes: underground mining of solid minerals and the collapse of such mines roofs, excessive withdrawal of groundwater, oil and gas exploration, dewatering or drainage of organic soils, sink holes, wetting of dry low density soil, and natural sediments compaction.

The coastal region of Nigeria is subsiding not only because it was formed in a tectonic setting but because of the continuing dewatering and compaction of its sediments which were deposited rapidly. The Rates of landsubsidence in this area are gradually increasing as a consequences of the excessive withdrawal of fluids, including oil and gas, particularly in Warri, Yenagoa and Port Harcourt cities from underground aquifers and reservoir strate. This uncontrolled exploitation of the groundwater, oil and gas has led to progressive decline of the aquifer level and a continuous need for opening deeper drillings to exploit deeper aquifers.

In this study, we focus on Lagos state where previous InSAR studies have shown that the land subsidence rates in the surrounding cities like Lekki, Badagry, Ikorodu, Ketu, Akoka, Banana and Victoria highlands towards the Atlantic Ocean and lagoons are much higher than that of Ikeja and other cities in the interland, where Epe town is located on a much higher elevation. Preliminary investigations revealed heavy structures, particularly buildings, constructed mostly on sand filled areas where the sediments compaction rates are very high. Multi-Temporal InSAR techniques using COSMO-SkyMed data in the period between 2014 and 2015 are used to monitor this land subsidence behavior in this coastal area.

 

Keywords: Coastal Areas, Subsidence, Nigeria, Lagos, MT-InSAR.

 

 

 

The slope instability in alpine areas is a phenomenon related to superficial landslides, deep-seated gravitational slope deformations and permafrost creep. The systematic monitoring of changes over time caused by the slope movements is of high importance for a proactive management of natural hazards related to these phenomena. Moreover, since permafrost is sensitive to the changing in climate conditions, observing its dynamics is a key issue in alpine environments. Rock glaciers, which are the most common geomorphological evidence of permafrost in alpine regions, are characterized by creeping processes that generate a downstream displacement with speed rates ranging from a few cm to more than 1 meter per year. This displacement varies from year to year and seasonally.

The three-years project ALPSMOTION (ALPine Slow slope Movement moniTorIng and detectiON with remote and proximal sensing) project started in August 2016 aims at combining and assimilating different approaches to detect slow movements in alpine regions using in situ and remote sensing data. To this end, the area of Lazaun located in Schnalstal/Val Senales (South Tyrol, Italy) has been selected as test area. Lazaun includes an active rock glacier and a complete set of glacial and periglacial landforms. The kinematic behaviour of Lazaun rock glacier is still poorly investigated, and the various techniques comprised by the project will be tested on it in order to fully understand its dynamics in relation to climate. The project activities will consider the following data sets: Sentinel-1, COSMO SkyMed and TerraSAR-XSAR (Synthetic Aperture Radar), GB-SAR (Ground Based SAR), GPS (Global Positioning System), UAV (Unmanned Aerial Vehicles) and TLS (Terrestrial Laser Scanner). These data will be exploited in order to analyse the permafrost deformation at different temporal and spatial scales.

One of the core techniques used in this project is the InSAR. InSAR can measure the surface displacement of vegetation- and snow-free areas with a millimetric accuracy. In addition, the space-born InSAR is capable to acquire data over very large areas with high repetition frequency (up to a few days). These characteristics, will allow both the estimation of the Lazaun rock glacier displacement rate and the detection of the activity status of the rock glaciers over the entire South Tyrol updating the regional rock glacier inventory. In particular, the second activity will enable to distinguish active and inactive landforms based on a kinematic criterion using an automatic and cost-effective methodology. It is worth noting that, the accuracy of satellite-based derived products will be assessed over the Lazaun test area exploiting the ground based data collected over the area (e.g., GPS and GB-SAR data).

Additionally, space-born InSAR is potentially capable to detect short-term displacement variations due to their high repetition frequency e.g. less than 5 days for COSMO SkyMed and 6 days for Sentinel-1. These are among the most promising results since the short-term rock glacier dynamics, such as the seasonal and infra-seasonal rhythms, is still poorly known and difficult to be systematically monitored using ground-based techniques.

In this work we present and discuss the first results obtained with the project and the approaches used to combine data coming from different sources over different time and space scales.

High-resolution SAR data from different satellites as presently available, has proved to be a strong source to monitor the land subsidence. In this paper, high-resolution Persistent Scatterers Interferometry (PSI), an established method has been used to estimate the urban subsidence due to depletion in ground water in the Delhi NCR-region. The urban area, which has the large development project generally show the subsidence phenomenon. In the study 68 TerraSAR-X images for the period of 26 months during September 2009 to November 2013 have been used for the purpose of estimating urban subsidence over the Delhi NCR area. The study area has been covered with the dense urban area, large infrastructure projects such as, railway lines, bridges, highways and metro bridges and tunnels. The result has been compared with the ground water table data, which has indicated a general trend of subsidence induced by the over-exploitation of the ground water in the area. In some of the area, rapid construction activities has also contributed to the subsidence phenomenon.

Sarproz software has been used to process the data and to generate the time-series result of the area. Most of the interfergrom was coherent due to very good spatial and temporal baseline of the InSAR pairs. In the most affected area, rate of subsidence was around 25-30 millimeter per year, while in another larger area, rate of subsidence was around 10 millimeters per year. The study suggests that there is an urgent need to address the issue of groundwater exploitation in the affected area to check the current phenomenon of subsidence. 

Polarimetric Persistent Scatterer InSAR (PSI) is a well-known technique to increase number and quality of selected PS pixels. Existing polarimetric PSI methods which are not quite optimal optimise amplitude-based criteria and spatial coherence. An optimised channel which is selected for each pixel based on amplitude is successful only for high amplitude scatterers such as man-made structures. Moreover, optimisation based on spatial coherence estimation assumes that pixels in a surrounding window all have the same scattering mechanism and leads to non-optimal solution.

In this study, we present a new polarimetric PSI method in which we use a phase-based criterion to select the optimal channel for each pixel. Our new polarimetric PSI approach is applicable in areas lacking man-made structures and retains the full spatial resolution of the input images.  This algorithm is based on polarimetric optimisation of temporal coherence, as defined in the Stanford Method for Persistent Scatterers (StaMPS), to identify scatterers with stable phase characteristics.

 We apply our algorithm to an area in the Tehran basin that is covered primarily by vegetation. Our results confirm that the algorithm substantially improves on StaMPS performance, increasing the number of PS pixels by 50%, 56% and 61% with respect to HH+VV, HH and VV channel, respectively, and increasing the signal-to-noise ratio of selected pixels.

Landslides are one of the major effects occurring after natural disasters such as short-term and intensive rains or after show melting. According to the national authority responsible for landslides monitoring and mitigation in the last two years their number almost doubled. This is reason why elaboration of fast and accurate method for observing them is needed. One possible solution is to use DInSAR derived information regarding Earth’s crust deformations such as subsistence and horizontal movement. In the last years this kind of information proved to provide reliable results for the said tasks and together with the dependable source of operational data from SAR missions such as Sentinel-1 form a solid basis for establishing a procedure for creation of maps the vulnerable regions prone to landslides in Bulgaria. One of those regions is located in the North-East of Bulgaria and is well known for several large active and potential landslides. The tectonic structure of the Black Sea basin is a complex zone of collision between the African and Eurasian plates and a movement around the various microplates has created an area that has the potential for occurrence of landslides. The Northern part of the Black Sea coast used in our research falls within the eastern part of Moesian platform. However, the majority of tectonic activity in the area of study dates from before Quaternary and this the reason it is commonly adopted that the majority of the faults present in the region of study are not highly active. In previous researches considering the geography of this region it was concluded that geomorphological conditions are highly favorable for landslides formation. One particularity of the area investigated is the abrasion process which is an additional factor in triggering landslide activities. Other reasons in this area for landslides induction are the increased construction activities of houses and roads which introduce additional instability on the slopes.
In this research we used SAR data in order to obtain information for the ongoing Earth deformation processes in the abovementioned zone. The achieved final results are in the form of interferograms witnessing the Earth’s crust motions and could be used for determination of areas needing more detailed surveys. Thus the method used provides cost effective manner for regular monitoring of the sites investigated and can provide complementary data in updating the landslide thematic maps. This study was focused on relatively narrow strip along the coast where most of the slopes are highly susceptible to landslides thus exhibiting some disagreement with the studies at national or EU level.

In 2014 Polish Geological Institute - National Research Institute (PGI-NRI) initiated the demonstration project that aimed to measure and monitor any potential terrain surface deformations that might be associated with ongoing fracturing activities in Poland. The project was aimed to answer to a social debate related to impact on environment of shale gas exploitation and to check whether any surface deformations could be identified.

For the project purposes 3 test areas surrounding the hydraulic fracturing sites has been selected. For each of the sites monitoring infrastructure has been designed and installed. In order to detect a sub-millimeter displacement the monitoring system consisting of geodetic (leveling and GNSS) benchmarks, and specially designed the corner reflectors (CR) for InSAR were deployed. This geodetic monitoring network was also strengthened by additional leveling and GNSS benchmarks that has been installed outside of expected potential deformation zone. In terms of SAR, data from TerraSAR-X and Sentinel-1 were acquired from April 2014 to December 2016. Also the historical data analysis (ERS-1/2 and Envisat ASAR) was to performed to determine whether any long-term, or previous terrain surface deformations occurred within the area of interest. The historical data analysis greatly improve a model of seasonal changes.

The terrestrial measurements in combination with CR-interferometry were capable to reveal and precisely locate any potential deformations and allow their validation. The results more than two years of analysis indicate that the study areas can be considered stable. However, extremely dry summer of 2015 and high temperatures influenced the results in a form of measurable seasonal effects. We will report on the current status of the project, and elaborate on the initial analysis based on collected data record.

Landslides are catastrophic phenomena, especially for those who live in landslide-prone areas. In general landslides occur on relatively steep, unstable slopes. The triggering of catastrophic landslide movement is usually associated with intense and extreme rainfalls. However, before catastrophe there is almost always an evidence that particular landslide remains or became active. This kind of movement may help the local authorities to make decision and may have an important impact to the general landslide susceptibility degree of the area.

This contribution will focus on study the evidence of landslide movement by using multiband SAR and LIDAR technique. SAR interferometry is an effective remote sensing technique for monitoring of small magnitude surface movement, and it has been successfully used in many areas especially for subsidence monitoring. SAR could penetrate the cloud and fog, and it is very useful for imaging landslides during raining season. LIDAR is an optical remote sensing technology that can measure the distance from the target to the sensor by illuminating the target with laser; it is commonly used for high resolution and high accuracy topographic mapping. By a repeat-visit LIDAR acquisitions of the landslide, the movement could be obtained as well, however,  the terrestrial LIDAR is capable to detect much bigger magnitude of deformation and it also highly depends on the weather conditions. Thus  the integration of these two methods is considered in this study for a better performance in landslides monitoring.

Since landslides happened on rugged mountain area, these areas are often difficult to monitor with InSAR technique due to layover, foreshortening and shadow. In addition, rugged topography cause a big challenge in compensating residual topographic phase even with accurate DEM. However, many standard InSAR processing algorithms are not performing well on such conditions. For the project purpose a new attempt including external DEM assisted and local optimum criteria based SAR image coregistration methods will be evaluated.  Moreover, in order to improve the time series analysis technique and make it applicable to the mountain area many additional improvements of SAR interferometric technique will be tested: a multi criteria stable point targets selection method, and DEM-based network construction method, offset tracking method, orbit re-estimation method, precise CR positing method and CR peak extracting method. The ground GPS data will be used for validation. And for LIDAR, it will be used to obtain the high resolution and high accuracy DEM for selected areas and used to measure surface displacements for validation.

In this study we are going to use of Sentinel-1, ENVISAT ASAR, ALOS PALSAR and TerraSAR-X sensors. They will be used for independent cross validation and for combined derivation of landslides movement as well. The study will be performed on selected landslides in Three Gorges area (Shuping, Fanjiaping) and Longnan region in China and in  Carpathians in Poland (Roznow area)

This research work is supported by the DRAGON-4 project (id. 32365) and is funded by the 5 years National Natural Science Foundation of China project. It has also support of Polish Ministry of Science and High Education within Poland-China bilateral cooperation and and PGI-NRI Landslide Counteracting System project (SOPO) funded by National Fund for Environmental Protection and Water Management of Polish Ministry of Environment.

A landslide occurred inside the loess plateau in west of China on January 18, 2016. And resultantly a tunnel running through the landslide mass for a high-speed rail way from Lanzhou to Xining was cracked by the abrupt sliding. 3 days later, a Ms6.4 earthquake happening at 160km northwest from the landslide aggravated the crack damages of the tunnel. In order to investigate the sliding reason and monitor the landslide development, the displacements of the landslide are detected with InSAR from ALOS-1/2, RadarSAT-2 and Sentinel-1A satellite data.

Firstly, for the investigation of the sliding reason, the pre-sliding interferometric displacements were derived from 21 SAR images acquired by Alos-1/2 and RadarSAT-2 during February 7, 2009 and December 13, 2015. The detected time serial displacements at the site of landslide are almost zero from February 7, 2009 to July 26, 2015. However, the accumulative deformations abruptly amount to 210mm during July 26, 2015 and December 13, 2015.

Secondly, in order to detect the deformation resulted from landslide and monitor the sliding proceeding, 6 Sentinel-1A images acquired during January 13, 2016 and March 25, 2016 are collected to deduce the displacements. Especially, because of approaching the landslide’s occurrence, two images acquired on January 13, 2016 and February 6, 2016 respectively are selected to investigate the simultaneous sliding deformation. The monitored maximum simultaneous sliding deformation is about 30mm. The displacements derived from other 4 images acquired from February 6, 2016 to March 25, 2016 trend to zero. This measure is called post-sliding deformation because it is derived from the post-sliding observations.

Finally, the seismic deformation and its covered area are detected from Sentinel-1A images acquired through earthquake. All the mapped displacements are essential for geologists to research the mechanics of the landslide. Furthermore, the differences of the deformation measurements detected from different SAR data will be discussed in this paper.

The main aim of this paper is to improve the performance of StaMPS PSI techniques in high gradient deformation area using TSX data. Eleven TerraSAR-X images, acquired from 2011.02.07 to 2012.16.05, have been used in this project. Tehran, the capital city of Iran, is affected by subsidence due to the over exploitation of under ground water, which led to the high rate deformation in some seasons, during last decades. So, this populous city has been selected as our study area in this project. The SBAS method has been applied on this stack of data to have a comprehensive cross-validation between the results obtained from two methods. Second phase of our project has been dedicated to compare the ability of TerraSAR-X and ENVISAT data for high rate deformation monitoring. Therefore, the available ENVISAT data in the area from 2003 to 2008 were processed using StaMPS method to have a comprehensive comparison between the results of two stacks of dataset. Finally, the obtained results were validated by pre-existing leveling data. Using TSX data had a significant role in improving the performance of StaMPS method when it comes to monitoring high gradient deformation compared to ENVISAT dataset. High spatial and temporal sampling in TSX data have been discussed as the main reasons of improvement in this research. The 2900 PS/km2 density of the identified PSs and 12 cm/yr maximum estimated rate has shown a remarkable improvement in comparison with ENVISAT dataset.

Kuchinoerabujima is an active volcanic island located on the volcanic front of the Ryukyu island arc. After 34 years of dormancy, the volcano erupted on 3 August 2014 and on 29 May 2015. Ground displacements near the summit area of Shindake were detected by InSAR analysis using the ALOS-2/PALSAR-2 data [e.g., http://www.eorc.jaxa.jp/ALOS-2/en/img_up/dis_pal2_kuchinoerabu_20140821.htm].  

Before these eruptions, seismic swarms were accompanied by a radially outward deformation from the summit crater repeated almost every two years since 1999 [Iguchi, 2007] and eruptive activities of Kuchinoerabujima volcano occurred at two active craters of Shindake and Furudake. InSAR analysis using the ALOS/PALSAR data [e.g., Yamamoto, 2009; Tanaka and Lundgren, 2013] also detected ground displacements near the summit area of Shindake. 

We applied the InSAR time-series analysis using the software package StaMPS/MTI [Hooper, 2010] to the ALOS/PALSAR data acquired on both ascending and descending orbits from May 2006 to March 2011. The line-of-sight (LOS) displacements show a rather complicated pattern compared with previous results obtained using GPS measurements and InSAR analysis. The mean velocity maps show two focused areas of LOS shortening located beneath Shindake and Furudake at a rate of 20 mm/year, confirming the inflation trend. The observed deformation near the summit area of Shindake was consistent with previous results. Also, it suggests another deformation source beneath Furudake, which was not clearly accounted for previously. We model the Kuchinoerabujima volcano sources that produced clear and distinct fringe patterns using a Markov Chain Monte Carlo optimization. 

A destructive landslide, covering about 40 hectares, re-activated on 16th July 2015 in Devrek District (Zonguldak- Northwest Turkey) damaging 2 houses, 1 high-school building, 1 mosque, and 1 bridge. Landslide movements continued for a couple of days after event occurred and eviction of 27 houses was also decided by Devrek District Governorship depending on field investigations made by officials from Republic of Turkey Prime Ministry & Emergency Management Authority (AFAD). According to records of Mineral Research and Exploration Institute (MTA), land sliding problem in the region exists since 1950s. The area is located in Black Sea Region of Turkey, which landslide events are common due to geological and meteorological conditions. In addition, there is not dense vegetation cover over the landslide area. In the present study, Persistent Scatterers Interferometry (PSI) was applied to measure surface displacement of landslide for a long term (1992-2010) by using ERS-1/2 and Envisat ASAR data. All data used in the study was obtained from European Space Agency (ESA) archives. While 82 Envisat ASAR data were used, 15 ERS-1 data and 55 ERS-2 data were used. All PSI analysis were carried out by using SARscape version 5.2.1. Preliminary results of PSI for determining historical surface displacement of landslide were evaluated.

Key words: ERS, Envisat ASAR, Historical Displacement, Landslide, PSI

Persistent Scatterers Interferometry (PSI) is advanced class of Differential Interferometric Synthetic Aperture Radar (DInSAR). PSI is time series analysis of InSAR data, emerged as an important tool for monitoring and measuring the displacement of the Earth’s surface. In the present study, PSI technique was applied to two C-band SAR dataset (ERS-2 and Envisat ASAR) to measure and monitor temporal evaluation of surface deformation caused by land sliding. The study area, called as Gschliefgraben landslide, is located in Upper Austria (municipality Gmunden). Gschliefgraben landslide, a big landslide system, reactivated recently and threated buildings and infrastructures. Following its reactivation in 2007, detailed studies describing the landslide evolution and mitigation measures in details were made. However, interferometric SAR application was applied for first time in Gschliefgraben landslide. While 55 Envisat ASAR data were obtained, 23 ERS-2 data were obtained between 2005 and 2011 from Europen Space Agency archives. All PSI analysis were carried out by using SARscape version 5.2.1. Moreover, 1 m LIDAR data between 2007 and 2012 were obtained. Preliminary results of PSI and LIDAR DEM based analysis were evaluated to measure surface displacement due to land sliding.

Key words: ERS, Envisat ASAR, DInSAR, Landslide Deformation, PSI

Subsidence is the downward movement of the Earth's surface relative to a datum. Many factors including underground mining, drainage of organic soils, natural compaction of soft soil, sinkholes, permafrost thawing and aquifer-system compaction occasion land subsidence and elevation changes of the ground surface in various regions. These variations can detect and estimate by different techniques such as GPS survey, leveling, and observations from disparate satellite sensors. Compared to conventional approaches, InSAR technique causes a revolutionary change in assessing displacement fields derived from seismic faults, landslides, subsidence, volcanoes, mining activities and other land deformation phenomena, since it can monitor broad areas with low cost, the short period of time, high precision and extraordinary spatial density of measurement points. In this research, 34 differential interferograms acquired from ENVISAT ASAR sensor over Mashhad plain in northeast Iran (from Sep. 2003 to Oct. 2008, in descending orbit and with normal baseline less than 300 m) have been processed. After DInSAR processing, to estimate land subsidence in the studied region, time series analysis using small baseline subset (SBAS) algorithm on the interferograms has been implemented. The results of the time series analysis are greatly compliant with continuous GPS observations in stations of the region and indicate that the maximum value of the cumulative subsidence equals ~98 cm during the studied period.

The presentation will provide an overview of the main results of Sentinel-1A data interferometric processing for the territory of mining activity in the Kola Peninsula.
In the Western sector of the Russian Arctic the majority of large mining companies are concentrated. This mining activity along with high social and economic significance has a great influence on natural environment. In the Kola Peninsula and in shelf area of Barents Sea and White Sea geomechanical hazards and natural and technogenic earthquakes with magnitude more than 4 take place and lead to catastrophic damages of both underground mining and ground infrastructure. The modern Earth crust tectonic movements are being activated and, in consequence, complicate mining operations and decrease industrial and ecological safety. The main goal of the research work was to estimate the potential of interferometric SAR data for monitoring of open-cast mines and production infrastructure of mineral deposits.
Sentinel-1A was successfully launched on 3rd of April, 2014. The wavelength is 5.6cm (C-band), the repeat pass period is 12 days. The Sentinel-1A data are available from October, 2014. The satellite has new modern sensing technology TOPSAR - Terrain Observation with Progressive ScanSAR. The territory of the Kola Peninsula is covered by Sentinel-1A images acquired with Interferometric Wide mode and data are available in Single Look Complex (SLC) format.
In this research work the interferometric processing of Sentinel-1A multi-temporal radar acquisitions is carried out. The regions of interest include mineral resources extraction area in the Kola Peninsula. The report will describe next results:
1. The displacements maps of ground surface on extraction sites of Kovdor’s Mining Plant and Kola mining Company were made using Persistent Scatterers Interferometry method.
2. The ground surface movements due to open-cast mine slope failure was analyzed using 2-pass differential interferometry method.
3. Horizontal displacements of slopes of Zhelezny open-cast mine were estimated using interferometric processing both ascending and descending Sentinel-1A acquisitions.
The given work was supported by Russian Foundation for Basic Research grant 15-29-06037.

The ability to perform time series analysis of distributed scatterers in SAR stacks opens up possibilities for the development or improvement of novel applications of InSAR. Recently, interest for InSAR applications to near-surface hydrological processes related to soil moisture is growing. Here we investigate clay soil swelling and shrinkage induced by soil moisture variations for hydrological and geotechnical purposes. Two main mechanisms of how soil moisture affects interferometric phase have been described: (i) clay swelling and shrinkage, causing actual deformations as a result of soil moisture content variations, and (ii) changes in the soil dielectric constant, influencing the propagation of the electromagnetic wave related to soil moisture content.

Disentangling the contributions of each of the aforementioned mechanisms in interferometric phases is challenging, as both mechanisms, combined with other phase contributions, may occur simultaneously over clay soils. To fully explore the potential of radar interferometry to measure clay swelling and shrinkage, improved understanding of all soil moisture related phase governing mechanisms is needed. Mechanisms that are not fully understood, and are therefore not incorporated in InSAR models, will affect the phase estimation, potentially limiting new applications or giving rise to misinterpretation.

Here, we study the interferometric phase over an agricultural area with clay soils in the Netherlands. The goal is to estimate vertical deformation as a result of clay shrinkage, and thereby apply corrections (i.e. phase reduction) of unwanted signals and develop methodology to improve phase unwrapping. We show that the phase contribution from clay shrinkage can be much more significant than the soil moisture induced dielectric phase, and that time series can be corrected for using a simple soil moisture phase model. Subsequently, a simple clay shrinkage model, based on widely available contextual data, can be used to improve phase unwrapping. Methodology and models are developed and validated with in-situ measurements of deformation. The results show that clay shrinkage can have a big impact on deformation estimates from distributed scatterers and can be exploited for hydrological studies or corrected for in other deformation studies. This study is relevant for hydrological monitoring as elevation change observations from clay soils might serve as a proxy for soil water storage change, especially on larger scales.

The salt diapirs, are subjected to the underground mining exploitation in many countries in Europe and also considered as potential structures for storage of hydrocarbons. The later application require detailed and accurate information regarding their geological stability. Polish Geological Institute – National Research Institute (PGI-NRI) initiated a series of studies to determine the magnitude and extent of the Quaternary and recent dynamics of the terrain surface of the area of salt tectonics in Poland. The study utilize SAR interferometry based on ERS, Envisat ASAR data, and Sentinel-1 and for the specific area of interest, that exhibiting deformation and containing sensitive infrastructure a detailed monitoring study including corner reflector, high-resolution TerraSAR-X data acquisitions and geodetic measurements is performed. The problematic areas are also investigated with shallow geophysical methods like Electrical Resistivity Tomography (ERT) and microgravity.

In this contribution we will elaborate on a design of the monitoring system and present results of one of the validation studies – the area of Wapno city. In the wider area of Wapno catastrophic deformation occurred after salt mine collapse that took place in 1977. Due to still unstable hydrogeological balance the area was subjected to subsidence up to 7.5 mm/yr. Since 2007 the subsidence is also associated with sinkholes that significantly increase the hazard to urban areas. To monitor the terrain deformations in Wapno area the existing network for in-situ measurements was supplemented by 10 levelling benchmark and 7 corner reflectors located in key areas. From July 2015 to september 2016 a TerraSAR-X data of Wapno area were systematically acquired including both ascending and descending passes. Independently to TerraSAR-X, and Sentinel-1 data were also routinely acquired over the area since 2014, and are also incorporated into InSAR analysis. We will report on results of historical data processing, present monitoring results of the Wapno area, and put them into a wider geophysical context.

Finally, some general conclusions and recommendations will be derived. Also we will address some of the identified weaknesses and limitations of our concept, and frame an open questions and challenges that grow from our work, and pose them to the community.

The presentation will provide an overview of the research work with the main goal developing of distributed system for SAR data interferometric processing using computational powers of workstation included in a local area network of a single organization.
Satellite radar measurements give a unique information about ground surface reflectivity. The satellite radar interferometry technique uses the effect of the interference of electromagnetic waves and is based on mathematical processing of several coherent amplitude-and-phase measurements of the same site of ground surface with shift in space of the radar receiving antenna. Researchers can use commercial (SARScape, GAMMA) and open-source (DORIS, StaMPS/MTI) software for radar data interferometric processing and monitoring of ground surface deformations and technogenic objects displacements. Recently InSAR analysts prefer to use modern approaches for multi-temporal radar acquisitions processing: Persistent Scatterers Interferometry (PSI), Small BAselines Subset, SqueeSAR. If enough amount of data is available, these methods make it possible to assess ground surface deformations and technogenic objects displacements with accuracy of few millimeters. As a consequence InSAR products have a wide field of application for monitoring of mineral resources deposits, industrial and urban infrastructure.
The interferometric processing of large amount of SAR using modern techniques like PSI, SBAS and SqueeSAR requires significant computational resources or lasts long time. The use of high-performance computer cluster can resolve the problem but such computer facilities are not available for any researcher. Hence the development of distributed system for SAR data interferometric processing using personal workstation as nodes is an actual task.
The report will describe the next achieved results:
1. The cross-platform software for a single computational node was developed. The interferometric processing chain requires matrix operations, linear algebra procedures and signal processing functions. Several available software libraries for serial (BLAS, LAPACK, FFTW) and parallel (ATLAS, OpenBLAS, MKL) computations were analyzed. As a results developed node software can operate under Linux and Windows, uses MKL for common computational procedures and OpenMP standard for more complex processing.
2. The distributed processing system including server, client and computational nodes was realized.
3. The sub-system for estimation of computational resources of each node was developed and realized. Each workstation can share processor, memory and part of disk space with the system. The user itself sets the number of threads and RAM memory space. The computational node software estimates run time of simple benchmark (2 matrices multiplication) and transfer this information to the server. Thus, the system rates nodes by availability of their computational resources.
4. The algorithms of differential interferometric processing including data import, coregistration, interferogram computation and flattening were realized as a part of the system software.
5. Two experiments including processing time estimation were carried out. 30 TerraSAR-X/TanDEM-X radar frames were used. The first experiment was a part of a typical PSI processing chain and included 29 differential interferograms computation with single master frame and 29 slave frames. The second experiment consisted in computing all possible interferograms (N*N-N)/2, where N is the number of frames. These 435 interferograms were used to analyze how interferometric coherence depends on date of acquisition and temporal baseline. The full set of interferograms will allow to detected localized short-term deformations of ground surface. In two next years it is planning to develop new method based on processing all possible interferograms. The conducted experiments showed that the developed distributed system make it possible to decrease time of SAR data interferometric processing without expensive computer cluster.
The project was supported by Russian Foundation for Basic Research grant 16-37-00224.

Deep-seated landslide, also known as deep-seated gravitational slope deformation (DSGSD), is generally linked to high relief mountain environments. Deep-seated landslides are normally slow and continuous movements of a large volume of soil and rock, and are sometimes the cause of catastrophic failures. Due to geodynamic context caused by active mountain building and sub-tropical climate setting dominated by high precipitations, deep-seated landslides are commonly observed in mountainous region in Taiwan. In Taiwan, over thousand landslides, including some fast catastrophic failure of deep-seated landslides, occur every year. Therefore, how to locate deep-seated landslides and monitor their activity has become an urgent task for the island to mitigate landslide hazards. This study illustrates the superiority of using the temporarily coherent point InSAR (TCP-InSAR) technique to monitoring the activity of deep-seated landslides in mountainous densely vegetated areas of Taiwan. In order to overcome the great topographic relief and heavy vegetation of mountain environment, the LiDAR derived 1 m resolution DEM is used to recognize deep-seated landslides with their landslide morphologic features. After that, L-band ALOS/PALSAR satellite radar images are selected for TCP-InSAR analysis due to their longer wavelength, which can better penetrate vegetation to determine the displacement of deformed slope. The study area is located in the Central Mountain Range of Central Taiwan. The selected study area is extremely susceptible to landslides during heavy rainfall because of its great topographic relief, high-slope gradients, and presence of highly foliated and weathered slate and thick poorly consolidated soils. The elevation of the study area decreases from 2,200 m in the northern mountain ridge to 855 m. The distributions of slope gradient in the study area, which calculated from the LiDAR derived 1 m resolution DEM via spatial analyst tool of ArcGIS, fall mainly in ranges of from 20 degrees to 40 degrees. The area with a slope gradient between 30 degrees and 40 degrees is 32% of the study area, while the area with a gradient between 20 degrees and 30 degrees accounts for 34. 22% of the study area has a slope of less than 20 degrees and only 1 % has a slope greater than 50 degrees. In addition, in order to reduce long wavelength error and localize atmospheric error, only 100 km2 is selected as the study area for TCP-InSAR analysis. Within the study area, an averaged density over 1300 TCPs/km2, which is significantly higher than the result of traditional PS InSAR analysis, is obtained. Within the study area, over ten DSGSDs with an area over 10 ha have been recognized in LiDAR derived 1 resolution DEM. Three specific deep-seated landslides are selected to illustrate the results of TCP InSAR analysis that one located at Chingin area (site D057) and two located at Lushan area (Site D066 and D067) are selected for detailed discussions. The average annual down moving displacement rate within the period of 2006-2011 is -15.1 mm/yr, -14.4 mm/yr, and -12.6 mm/yr. The standard deviation of TCP measurement in Lushan area is estimated about 7-8 mm by comparison the TCP data with GPS data. In addition, significant seasonal variations, which indicate higher and lower moving rates during rainfall and the dry season respectively, are also observed.

The beneficiary of natural phenomenon is based on their effect on the society or environment. Destructive events such as land subsidence, landslide, and earthquake, which might be predictable in some situations, cause many problems for people, infrastructures, and transmission lines. Therefore, monitoring these events and determination of their time, location, and agents are very important and also essential to decrease the potential damages. In the past, this monitoring was performed using field observations, collected by physical instruments in the field. However, the offline monitoring and prevention of the phenomena are the main disadvantages of this method.

Traditionally, these events were monitored by GPS observations and remote sensing satellite images. Meantime, the appearance of radar satellite images with wide spatial and temporal resolutions increases the potential of remote sensing technology in these researches, in comparison to GPS based techniques.

Interferometric synthetic aperture radar (InSAR) is a RS-based powerful technology to explore the artificial and natural geological events. In fact, development of radar satellites, radar images with wide spatial and temporal resolutions, and algorithms take more attention to use this technology for tectonic and ground researches. Obviously, InSAR technology has several error sources caused in accuracy decrement e.g., spatial and temporal de-correlation, digital elevation model (DEM), atmospheric and orbit errors, and thermal noise. In this study, all of these uncertainties are considered to access the excellent accuracy.

Landslide phenomenon is a form of mass wasting that includes a wide range of land movements, such as rock falls, deep failure of slopes, and shallow debris flows. Landslide can occur in underwater, called a submarine landslide, coastal and onshore environments. Although the action of gravity is the primary driving force for landslide to occur, there are other contributing factors affecting the original slope stability. Typically, pre-conditional factors build up specific sub-surface conditions that make the area/slope prone to failure, whereas the actual landslide often requires a trigger before being released. Landslides should not be confused with mud flows, a form of mass wasting involving very to extremely rapid flow of debris that has become partially or fully liquefied by the addition of significant amounts of water to the source material.

Landslides occur when the slope changes from a stable to an unstable condition. A change in the stability of a slope can be caused by a number of factors, acting together or alone. Natural causes of landslides include groundwater (pore water) pressure acting to destabilize the slope, loss or absence of vertical vegetative structure, soil nutrients, and soil structure, erosion of the slope toe by rivers or ocean waves, weakening of slope through saturation by snow melt, glaciers melting, or heavy rains, earthquakes adding loads to barely stable slope, earthquake caused liquefaction destabilizing slopes, and volcanic eruptions. Also, landslides are aggravated by human activities, such as deforestation, cultivation and construction, which destabilize the already fragile slopes, vibrations from machinery or traffic, blasting, earthwork which alters the shape of a slope, or which imposes new loads on an existing slope, in shallow soils, the removal of deep-rooted vegetation that binds colluvium to bedrock, and construction, agricultural or forestry activities (logging) which change the amount of water infiltrating the soil.

Based on Geological Survey of Iran (GSI) and Forest, Range and Watershed Management Organization of Iran (FRWO) report, Symareh landslide is the oldest and biggest in the world, which has eleven thousand years old. It is located on Southeast of Ilam province in northern slope of kabirkuh mountain and on the limestone of ASMARI formations in southwestern of Iran. This landslide is kind of the deep-seated one, which involve deep regolith, weathered rock, or bedrock and include large slope failure associated with translational, rotational, or complex movement, and has stumbled the mass weighing twenty-seven billion tons. The dimension of this landslide are 15 kilometer length, 2500 meter width, and 300 meter thickness of ASMARI formations. The different factors involved to creation of Symareh landslide includes high gradient of ground layers, existence of Pabde-Gurpi formations underground, disemboguing the support of ASMARI formations by Karkhe river, existence of Karast phenomenon in support of ASMARI formations, and high rainfall in this region. According to reports of geologists, Symareh landslide can activate the available faults and lead to the earthquake in the region.

Investigation of geological activities and ground displacement in this study area is the main purpose of this research. Due to cover the whole of study are in an appropriate period by Envisat and Sentinel-1 radar satellite images, the InSAR technique is used to monitor this landslide and estimate the value of the displacement with the high accuracy. Also, due to vegetated coverage of the area, the different techniques such as Persistence Scatterer (PS), Small Base-Line Subset (SBAS), Quasi-PSI, and Multi-Temporal InSAR (MTI) techniques are used to measure the displacement values with the lowest uncertainties. Preliminary results demonstrated that this landslide has been reactivated to has a considerable amount of displacement.

Built in 2008, Beijing-Tianjin intercity high-speed railway plays an important role in these two mega-cities of China. As the political, cultural and economic centre of China, Beijing is one of the most water-scarce cities in the world, and groundwater is the major water source, accounting for about two-thirds of water use. Tianjin also suffers water shortage because of its semiarid climate, natural geographic condition and large population. Due to over-exploitation of groundwater, Beijing and Tianjin mega-cities have been suffering from land subsidence for decades. Land subsidence is a severe hazard threating the safety of urban man-made linear infrastructure. It is well known that a tiny displacement of the rail could result in serious consequences. Therefore continuous monitoring of land subsidence along the Beijing-Tianjin intercity high-speed railway is critical to maintain its safety operation.

Advanced InSAR techniques such as Persistent Scatterer Interferometry (PSI) and Small Baseline Subset (SBAS) have been used to measure land subsidence along railways. Ge et al. [2009] empolyed PSI with two tracks of Envisat ASAR images (acquired from 2007 to 2008) to investigate the land deformation rate maps along Beijing-Tianjin high-speed railway showing that the maximum velocity is near -60 to -70 mm/year. Ge et al. [2010] used InSAR to map the land subsidence along the Beijing-Tianjin high-speed railway and studied the impact of the subsidence on the railway. Perissin et al. [2012] used the SARPROZ software to study land subsidence along newly excavated subway tunnels in Shanghai with 33 Cosmo-SkyMed SAR data. Yu et al. [2013] adopted PSI to investigate land deformation along road network in Tianjin with high resolution TerraSAR-X images. The results showed that the subsidence rates range from -68.7 to -1.3 mm/yr, indicating the uneven subsidence pattern along the road network. Chang et al. [2016] integrated 213 acquisitions over three independent satellite tracks of Radarsat-2 images between 2010 and 2015 for monitoring railway infrastructure over a national wide scale in the Netherlands.

Terrain Observation by Progressive Scan (TOPS) mode from the Sentinel-1 satellites can provide high-quality SAR data with a wide ground coverage, attracting great attention from researchers. In this study, ascending and descending Sentinel-1 TOPS data between 2015 and 2016 and the ascending TerraSAR-X stripmap images between 2012 and 2016 are used to investigate the land subsidence along the Beijing-Tianjin intercity high-speed railway, particularly in the Beijing region. Small baseline InSAR is applied to obtain land deformation information. Comparisons of InSAR-derived subsidence rates from multi-platform SAR datasets show a high correlation coefficient with a small RMS difference, indicating the reliability of InSAR retrievals of land subsidence rates. GPS derived subsidence rates were also used to compare with InSAR retrievals, and a high consistency was also obtained, which provides additional supporting evidence for the robustness of InSAR retrievals. Three-dimensional displacements are then constructed using data from multi-platform InSAR measurements. Our preliminary results show that the intercity high-speed railway passes through the edge of the deformation centre of Dongbalizhuang-Dajiaoting with a subsidence rate greater than 50 mm/yr. The spatio-temporal evolution characteristics of land subsidence and the uneven deformation along the Beijing-Tianjin intercity high-speed railway are analyzed together with hydrogeological data. The combined analysis of multi-platform data enables us to better understand the subsidence funnel and the causes of different subsidence rates, which is important for predicting potential hazards and designing compensation strategies.

The SAR Differential Interferometry (DInSAR) is nowadays one of the methods with the greatest potential of development for identifying movements on the Earth surface. Landslide monitoring with DInSAR techniques will soon be available to many research groups and also to the Administration. This will be possible due to the development of on-demand web tools such as the Grid Processing on Demand (G-POD) environment that is a part of the ESA’s Geohazards Exploitation Platform (GEP). Here, we present the results of an analysis carried out with the SBAS InSAR service, a tool included in the G-POD platform. The results have served to identify many areas affected by subsidence and active landslides en the southern coast of Spain. The activity of slope movements on several of these areas was not known at the time of the analyses. Subsequent field surveys found clear evidences of these movements such as cracks, fissures and fresh scarps in the ground surface and damages on buildings. Additionally, the comparison of the results provided by the SBAS InSAR service with previous results achieved by using different DInSAR techniques allowed validating the data obtained by the ESA service. Regarding the detected active landslides, two examples worth noting are the movements monitored in the urban resorts of “Marina del Este” and “Carmenes del Mar”. They are two cases studies already described on two precedent scientific papers. Considering the ground subsidence, the known settlement observed on various coastal municipalities due to water withdrawal is also detected by the analysis carried out in the GEP. Although in the South of Spain the analysis had successful results, the service has still limitations because currently is a prototype. However, all the observed results indicate that the SBAS InSAR service of the GEP is a reliable and powerful tool for performing a preliminary recognition of active slope movements and zones affected by ground subsidence in arid areas without vegetation. 

Low-lying coastal communities face increasing challenges from rise in sea level, more extreme storm surge levels and floods. In addition, changing groundwater levels and precipitation patterns may further exacerbate the water-related impacts of climate change on society. Approximately 40,000 km² of Europe’s North Sea region is already flood prone. Storm surges pose a real and substantial risk to this area, especially the densely populated areas. Climate and sea level research seek to provide robust regional projections of change and to address uncertainties and errors inherent in climate models. It is a challenge for coastal communities to transform this information in order to provide for local impact assessments and to implement adaptive measures. To this end, information about potential subsidence, its magnitudes and causes is important: subsidence may adversely affect the probability, extent and depths of future floods, and knowledge about subsidence will serve to reduce the total uncertainty about the anticipated climate impacts. If included in an ‘impact integration system’, reliable subsidence mapping may serve to deal with possible future outcomes in local management and planning.

The paper presents subsidence mapping using Sentinel-1 (S-1) data over a case study area on the Danish North Sea coast, and it addresses challenges to validate and reference results to the national datum levelling network. For this, repeated precision levelling (2006-2015) and ERS2 (1995-2001) data are used. In addition, the Sentinel-1 time series for selected scatter points are compared to groundwater level data from 10 wells and sea level data from two tide gauges to analyse their effect in the S-1 data. Likewise, the variations in the ocean water level (from tidal excursion and positive/negative surges etc.) and in the groundwater table (from ocean level and gradient, wave run-up, precipitation etc.) may in an initial evaluation suggest time-dependent and water-related mechanisms for the inferred subsidence encountered. These variations may thus serve to detail our understanding of S-1 results, and they may be indicative of system responses to subsidence under climate change scenarios. Results are put into perspective in relation to additional S-1 studies carried out by the authors as well as to literature to outline perspectives of further work to relate and apply S-1 data to improve local coastal climate impact assessments and adaptation.

With more than two years of Synthetic Aperture Radar (SAR) observations since the launch of the first Sentinel satellite Sentinel-1A in April 2014, the Sentinel-1 tandem provides a reliable, constantly growing dataset for many applications. For interferometric uses - after the precise co-registration that is required due to the new scanning method Terrain Observation with Progressive Scanning (TOPS) - we can not only create interferograms but perform persistent scatterer (PS) analysis for long-term stability and deformation analysis as well.

In this study, we examine our early Sentinel PS results in terms of quality and errors, e.g. the accuracy and precision of PS velocities using statistical methods. It is necessary to thoroughly investigate these as Sentinel satellites serve as successors of earlier ESA SAR missions ERS and Envisat, which have been already providing high-quality results due to the extensive temporal coverage with more than 20 years of measurements.

Budapest, the capital of Hungary with a population of more than 1.5 million is situated in the selected pilot area. It was chosen because it is expected to provide hundreds of thousands persistent scatterers with high amplitude, good phase stability and with many different scattering and movement characteristics. The different behaviour of scatterers can help us understand the composition of phase components more thoroughly, to find better estimates for reducing atmospheric contribution, and to accurately determine the real underlying deformation both quantitatively and qualitatively.

The processed data consist of 3 bursts of the same sub-swath. Multilooking is done by factors of 10 and 2 in azimuth and range directions, respectively. We chose the area and set data processing parameters by keeping the storage and processing power limitations of a single workstation in mind. The persistent scatterer processing is entirely done by the Gamma software from Single Look Complex (SLC) data to displacement time series, using a single master approach. The dataset covers more than 2 years and consists of 50 descending images - the last one has already been acquired by the Sentinel-1B satellite.

High coherence is expected between subsequent acquisitions due to the small temporal baseline of 12 days. This results in high PS density in urban areas and sparse coverage of rural areas where no agricultural activity is present. The main source of error is the atmosphere, as it is the largest unknown addition to the measured phase. Removal of this phase contribution is essential to reach the maximum potential of Sentinel deformation measurements. We are testing filtering methods in both temporal and spatial domain.

In summary, we present our preliminary results of Sentinel PSInSAR applied to the mainly urban setting of Budapest area and inspect it in terms of quality and accuracy.

Land subsidence caused by groundwater pumping is a common geohazard in many countries of the world. In Iran, this is a serious challenge for many regions, particularly in the plains with arid and semi-arid climate. This necessitates the study of this phenomenon. A large area in Qazvin plain, located in north-center of Iran, is subject to the land subsidence induced by overexploitation of groundwater.

Initially, the Qazvin plain subsidence area is monitored by “ASAR ENVISAT InSAR”. This SAR imagery has become known as an important and unique technique in order to study the land subsidence. This data set consists of 20 and 18 images of descending tracks D192 and D421 covering 2003 to 2010 Then, the time series analysis of permanent scatterer (PS) and small baseline subset (SBAS) algorithms are used to estimate the deformation rate. Mean line of sight deformation velocity maps obtained from time series analysis (PS and SBAS) demonstrated a considerable land subsidence in the studied area. Results show a good agreement between the PS and SBAS time series, both approaches identify peak amplitude of ~ 30-35 mm/yr for South-East part of the Qazvin plain during 2003-2010.

In the next step, the Qazvin plain subsidence is monitored by Sentinel-1A images acquired by the new satellite Sentinel-1A of the European Space Agency (ESA).  In the time period 2014-2016, a descending track by 22 images of 6, 12, 24 average revisit times (in days) is selected. Afterwards, the GMTSAR software package is applied. The preliminary results reveal that Qazvin plain is still suffering from subsidence.  The consistency of the Sentinel-1A and ASAR ENVISAT indicate the efficiency of Sentinel-1A in monitoring land surface subsidence.  

 

Popocatepetl is located only 55 km away from Mexico City urban area (~20 million inhabitants) and is one of the most active volcanoes in Mexico. The eruptive activity in the volcano restarted in 1997 after 70 years of inactivity and has continued since with cyclic dome growth and destruction, ash emissions, tremors, and moderate explosive eruptions. The high risk it represents requires permanent monitoring.

The current generation of X-band SAR satellites provides a unique opportunity to monitor deformation in the volcano by using Interferometric SAR (InSAR). Here we present deformation results from 2012 to 2016 obtained using the Radarsat-2 satellite of the Canadian Space Agency, and the Terra-SAR-X satellite of the German Aerospace Center, using the Small Baselines (SBAS) method. We complement the deformation analysis with seismic and GPS information.