Earthquakes and tectonics II

We present inter-, co- and post-seismic displacements observed in the 2015 Mw8.3 Illapel earthquake area by using Synthetic Aperture Radar Interferometry (InSAR) and Global Positioning System (GPS) techniques. RADARSAT-2, ALOS-2 and Sentinel-1A interferograms captured the inter-, co- and post-seismic displacements in the deformation area of the 2015 Mw8.3 Illapel (Chile) earthquake. As in a sparely vegetated area, the RADARSAT-2 (RS2) interferograms with over 3-year long time interval still keep excellent interferometric coherence. Significant interseismic deformation potentially related to the viscoelastic rebounds have been clearly revealed with these two RS2 interferograms. Based on a layered Earth structure, we modeled both co- and post-seismic faulting behaviour on the subduction interface of central Chile based on Sentinel-1A interferograms and GPS observations. The best-fit coseismic slip model shows that the earthquake ruptured a 200 km×200 km area with a maximum slip of 10 m at a depth of 20 km. Two distinct slip centers, likely controlled by the local ramp-flat structure, are revealed in this model. The total coseismic geodetic moment is 2.76×10²¹ Nm, equivalent to a moment magnitude of 8.3. The accumulated afterslip in the first two months after the mainshock is observed on both sides of the coseismic rupture zone with both ascending and descending Sentinel-1A interferograms. A limited overlapping zone between the co- and post-seismic slip models can be observed, suggesting partitioning of the frictional properties within the Illapel earthquake rupture zone. The total afterslip releases ~5.0×10²⁰ Nm geodetic moment, which is equivalent to an earthquake of Mw 7.7. The 2010 Mw 8.8 Maule earthquake that occurred ~400 km away from the Illapel earthquake epicenter could have exerted certain effects on the seismic cycle of the Illapel earthquake area. The local seismicity records from 2000 to 2015 show that in the Illapel earthquake area the rate of annual seismic moment release dropped from 0.4 to 0.2 ×10¹⁹ Nm/yr after the Maule earthquake. Based on the forward modeling with the best-fit slip models determined in this study, we reproduce the local surface displacements before, during and after the 2015 Mw8.3 Illapel earthquake. A rough deformation cycle, 105±29 years, derived by using the coseismic displacements and interseismic rate is consistent with the revisit interval of M8 events in the adjacent areas of the Illapel earthquake, suggesting that elastic rebound theory is applicable for the long-term strong earthquake prediction in this region. 

The Cerro Prieto Geothermal Field (CPFG) lies at the step-over between the Imperial and the Cerro Prieto Faults in northern Baja California, Mexico. While tectonically this is the most active section of the southern San Andreas Fault system, the spatial and temporal deformation in the area is poorly resolved by the sparse Global Positioning System (GPS) data coverage. Moreover, interferograms from satellite observations spanning more than a few months are de-correlated due to the extensive agricultural activity in this region.  Here we investigate the use of frequent, short temporal baseline interferograms offered by the new Sentinel-1A satellite to recover two components of deformation time series across these faults.  The Sentinel-1A satellite uses a new TOPS mode to acquire complete spatial coverage on a 12-day or 24-day cadence from two look directions [Meta et al., 2010].  While this new technique enables frequent acquisitions, it brings new challenges to interferometric synthetic aperture radar (InSAR) data processing in alignment and resampling. Following previous studies [Gonzalez et al. 2015, Prats-Iraola et al., 2012, Sansosti et al. 2006], we have developed a purely geometric approach for image alignment that achieves better than 1/200 pixel alignment needed for accurate phase recovery.  This is implemented in GMT5SAR.  There are two significant advantages to the geometric alignment approach. First the alignment does not rely on phase coherence between the master and slave images so one can accurately align images with large time separation that may be completely de-correlated. Second, the high accuracy of the geometric approach eliminates phase errors associated with slight misalignment so long time span deformation can be accurately constructed from a sum over short time span interferograms because atmospheric and orbital errors of the intervening SAR images will cancel. To demonstrate these advantages we combine a two-year time series of ascending (34) and descending (42) Sentinel-1A images to map the details of the vertical and fault-parallel deformation of the CPFG region. In practice, we construct InSAR time series using a coherence-based SBAS method [Tong and Schmidt, 2016] with atmospheric corrections by means of common-point stacking [Tymofyeyeva and Fialko 2015]. With these algorithms, the subsidence at CPGF is clearly resolved. The maximum subsidence rate of 160 mm/yr, due to extraction of geothermal fluids and heat, dominates the ~40 mm/yr deformation across the proximal ends of the Imperial and the Cerro Prieto Faults. 

Since 2009, Oklahoma has experienced a sore in induced seismicity, a side effect of extensive saltwater injection into subsurface sedimentary rocks. The seismic hazard entailed by regional-scale injection operations is however difficult to assess. The September 3, 2016, Mw5.7 Pawnee earthquake is the largest since the increase of seismic activity. Using Sentinel-1A and Sentinel-1B spaceborne interferometric synthetic aperture radar, we unambiguously show that the earthquake produced vertical displacement of 2-3 cm at the surface. Kinematic inversion of geodetic and seismological data shows that the main seismic rupture occurred between 4 and 9km depth, over a length of 8km, with slip reaching at least 40cm. The causative fault is entirely buried within the Precambrian basement, i.e. well beneath the Paleozoic sedimentary pile where injection is taking place. Potentially seismogenic faults in the basement of Oklahoma being poorly known, the risk of Mw≥6 events triggered by fluid injection remains an open question.

The Sentinel-1 constellation represents a major advance in our ability to monitor our planet’s hazardous tectonic and volcanic zones operationally. Sentinel-1 uniquely offers routine acquisitions with short revisits, a commitment to a long-duration mission, systematic wide-area coverage, good orbital control, and a free and open data policy. Together, these give us the unprecedented ability to respond to most earthquakes and eruptions and to build the long deformation time series that are required to resolve the slow deformation that occurs between events. Here we present the latest progress from COMET(*), where we are now providing processed products and derived results to the community for volcanoes and the tectonic belts (**).

COMET’s work on earthquakes and volcanoes can be split into response and preparedness. We now respond routinely to most significant earthquakes that occur in the continents, providing interferograms and interpretations to the community rapidly – Sentinel-1 allows us to do this within a few days for most earthquakes. For example, after the M7.8 Kaikoura (New Zealand) earthquake, on 14 November 2016, and with assistance from ESA, we supplied a processed interferogram to the community at 1 pm on 15 November, just 5 hours and 37 minutes after the Sentinel-1 acquisition. This data set was used extensively by colleagues at GNS in New Zealand to help them identify faults that had failed in the earthquake – vital in this case as it was one of the most complex earthquakes ever to have occurred. The fault maps models that resulted from the InSAR data (from Sentinel-1 and ALOS) completely changed the local and USGS estimates of ground shaking, and are likely to lead to modifications to seismic hazard codes worldwide. We are currently automating our response systems to take advantage of the guaranteed acquisitions that Sentinel-1 offers. By the end of 2017, we expect to be producing interferogram products systematically for all earthquakes larger than M~6.0.

Preparing for earthquake and volcanic hazard first requires identification and characterisation of the hazard. Deformation data are now becoming a key piece of information in that process. For example, Biggs et al (Nature Communications 2014) showed that there is a strong diagnostic link between volcanoes that deform and volcanoes that erupt. Of equal importance, they showed that volcanoes that do not deform only rarely erupt. At fault zones, strain energy accumulates over long periods of time around faults that eventually fail in earthquakes. By mapping the accumulation of strain, we can place constraints on how often earthquakes can occur in a given region. To make an impact for volcano and fault zone monitoring, we need to be able to measure deformation rates on the order of 1 mm/yr or less. This requires mass processing of long time series of radar acquisitions. In COMET, we are currently (December 2016) processing interferograms systematically for the entire Alpine-Himalayan belt, which stretches over 9000 km from Italy through to China, and is up to 2000 km wide, and making interferograms and coherence products available to the community. By June 2017, we expect to be processing a wider tectonic area and all ~1500 volcanoes that have erupted in the Holocene. We plan to provide average deformation rates and time series for all these areas. Results will be made available through our dedicated portal as part of the COMET-LiCS project (**), and are being linked to the G-TEP portal and EPOS during 2017.

We will show the latest wide area results for tectonics and volcanism, and discuss how these can be used to build value-added products, including (i) maps of tectonic strain (ii) maps of seismic hazard (iii) volcano deformation alerts. The accuracy of these products will improve as the number of data products acquired by Sentinel-1 increases, and as the time series lengthen.

Finally, we end the presentation by discussing what we hope the future, from Sentinel-1 and a range of other SAR sensors that will be launched in the next decade, and look beyond Sentinel-1 to what we might hope for from a SAR system, or system of systems, in the 2030s and beyond.

*COMET is the UK Natural Environment Research Council’s Centre for the Observation and Modelling of Earthquakes, Volcanoes and Tectonics: http://comet.nerc.ac.uk

**Data are available for download at http://comet.nerc.ac.uk/COMET-LiCS-portal/

Earthquakes and tectonics II