1. Community Geodetic Model Jessica Murray, David Sandwell, and Rowena Lohman June 1, 2014

2. Motivation: Spatially and temporally dense time series of ongoing deformation utilizing the complimentary features of GPS and InSAR data are expected to be central to several SCEC4 core initiatives and science targets. The continued expansion of GPS coverage, launch of new SAR satellites, recent advances in InSAR time series analysis, and ongoing advances in noise assessment and mitigation can be leveraged in order to develop methodology for generating a combined GPS/InSAR time series product. The Community Geodetic Model

3. Some target applications for the CGM: Quantifying slip rates and strain rates and their spatial variations in the complexly-faulted southern California region, including off-fault strain Assessing non-tectonic time-varying signals without aliasing Detecting transient deformation and tracking its space/time evolution at sufficient precision to relate it to other processes such as seismicity Constraining lithospheric rheology and evaluating its role in earthquake cycle deformation Characterizing postseismic deformation and the underlying physical processes Providing input to develop or refine stress and stressing rate models and/or to validate candidate stressing rate models for the Community Stress Model Different applications require different spatial and temporal data coverage and resolution; GPS and InSAR can complement each other.

4. Available Data: GPS and EDM Figure: Duncan Agnew

5. Available Data: SAR Comprehensive time series ERS & Envisat 1992 - 2011 Sentinel start ~2014 Few targeted time series studies possible with COSMO-SkyMed & TerraSAR X (Currently sparse and costly) Ascending ALOS velocity field, but limited temporal resolution 1-to-8-day repeat 11-day repeat Scott Baker, Roland Burgmann 11,000 7000 4000 1400 800 2000 1700

6. InSAR to Support CGM - News Sentinel-1A (ESA) was successfully April 3, 2014, SAR collecting data! –C-band, 12-day repeat; provides continuity with ERS and ENVISAT –Mostly ScanSAR coverage of the SAF, ascending and descending –Completely open data access – finally!! –Sentinel-1B to be launched 2016 and will provide 6-day repeat interval ALOS-2 (JAXA) was successfully launched May 24, 2014, antenna deployed –L-band, 14-day repeat –Mostly ScanSAR coverage of the SAF, ascending and descending –PI proposal needed for data access –Limited quantities per PI –Perhaps SCEC should work with DPRI for improved access for SCEC research NISAR (NASA) was approved for phase A funding May 7, 2014 –L and S-band, 12-day repeat, launch 2020 –Mostly ScanSAR coverage of the SAF, ascending and descending –NASA-ISRO SAR Mission Applications Workshop, USGS, October 28-29, 2014

7. CGM Workshop: May 30 – 31, 2013, Menlo Park 29 participants including experts in GPS and InSAR as well as representatives of the CSM, Ductile Rheology, Transient Detection, and UCERF3 efforts Established what CGM will comprise: Will include campaign and continuous GPS, any available InSAR data Three basic categories: GPS-only, InSAR-only, combined product Raw time series: GPS 3D, InSAR LOS Derived quantities: secular rates, offsets, outliers, postseismic decay terms, seasonal models, noise parameters, common mode filter parameters, InSAR coseismic displacement fields Combined product will not be interpolated to a uniform grid; will be more dense near faults (in part driven by availability of GPS sites which are used to constraint InSAR) Two working groups formed: GPS and InSAR

8. Next Steps – 2014 activities Additional data gathering, comparisons, and method development are needed for both the GPS and InSAR components before we can begin merging them GPS action items: Gather all SGPS data and metadata Collect more campaign data if appropriate Compare different processed solutions; reconcile discrepancies Merge campaign and continuous processed results (reprocess if needed) Decide best strategy for modeling noise, postseismic, seasonal and apply InSAR action items: InSAR time series analysis test exercise to establish best practices Reach out to broader InSAR community, including international, as appropriate Both: Develop strategy and framework for combining GPS and InSAR time series to produce a joint data product including covariance information

9. Funded CGM-related GPS work (2014) Additional field campaign data collection Floyd & Funning: GPS sites where older data in archives can be leveraged to improve spatial coverage of velocity field Fialko: Coyote Creek fault and San Felipe shear zone cross-fault profile McGill: Resurvey new (2013) San Gorgonio Pass sites and part of the San Bernardino Mtn. network Agnew, Floyd, Herring, Shen: Compile campaign GPS data and metadata, especially to identify previously unused data and to verify metadata; leverage CMM infrastructure Floyd, Hammond, Herring, Owen, Shen: Combine processed solutions for multiple GPS networks; process/reprocess if necessary to produce time series; compare results from different solutions Floyd, Herring, Shen, Group: Identify and implement best approach for merging GPS time series results to form the GPS component of the CGM Segall: Assessment of GPS time series noise and its impact on velocities

10. Funning and Floyd M&B F Fialko McGill and Bennett

11. Big-picture InSAR goals (SCEC and WInSAR) Develop integrated GPS/InSAR deformation models (SCEC CGM) at 200 m spatial resolution and better than seasonal temporal resolution. Work with ESA to ensure SCEC science targets are included in Sentinel-1 acquisition plan. (Currently land surface monitoring is only 5 th highest priority.) SCEC and DPRI should approach JAXA on ALOS-2 data sharing agreement. ALOS-2 top priority is disaster monitoring. Promote a NASA L-band mission to provide data later in decade. Continue to develop open-source software tools for automated processing of the new ScanSAR data streams.

12. Average LOS velocity, InSAR time series Pomona, CA Ground water pumping Santa Fe Springs, CA Oil wells San Jacinto Fault, Tectonic signal Simons and Agram Variety of vertical deformation sources to consider

13. Lanari et al., 2004 Example: Earlier work highlighted Santa Ana aquifer signal

14. Jun 1992 – Sep 2010 105 ERS images + 60 Envisat images. 881 interferograms Acquisitions almost every month for 18 years. Monthly acquisitions allow study of seasonal signal in detail. From Piyush Agram Detailed study with more data and new methods is ongoing

15. InSAR illuminates deformation on short spatial scales Seasonal AmplitudeSeasonal Phase Simons and Agram Amplitude of signal appears “smooth” while phase has sharp boundaries

16. Time lag (days) of seasonal amplitude compared to location with maximum seasonal amplitude (*) Spatial variation related to differences in in lateral conductivity and may be distance-dependent Full scale Reduced color scale range InSAR time series also reveal temporal patterns Simons and Agram

17. First “real data” InSAR time series comparison Quantitatively Identify agreement among InSAR time series solutions and between InSAR and GPS; will consider – “flattening” of interferograms with/without GPS – unwrapping errors in rough terrain – tradeoffs between atmospheric signals and real temporal variations Dataset: 2 frames – Desert, mountains, urban – Low strain areas + regions with known anthropogenic signals – Dense GPS site coverage Targeted outputs – Gridded map of LOS secular velocities – Time series at selected GPS sites – Time series at unknown, non-GPS sites Agram Baker Funning Liu Lohman Shirzaei Tong (PI Sandwell) Tymofyeyeva (PI Fialko) Anticipated participants InSAR time series test exercise

18. Phase I (2013) – Inter-group comparison limited by differences in choice of interferograms, spatial scale – Comparison of InSAR results with velocities of individual GPS stations problematic – Intra-group comparisons focused on effect of atmospheric models Phase II (2014) – Participants will use the same set of interferograms and the same spatial scale. – Coordination telecons in June and August; comparison of results at September 2014 workshop Agram Baker Funning Liu Lohman Shirzaei Tong (PI Sandwell) Tymofyeyeva (PI Fialko) Anticipated participants InSAR time series test exercise

19. September 2014 CGM workshop GPS topics Review data compilation to identify remaining gaps in data or metadata Compare processed GPS time series to identify and resolve discrepancies Decide upon strategy for merging GPS solutions into single time series dataset; consider choice of reference frame Identify most appropriate approach for time series analysis and noise characterization to estimate derived quantities InSAR topics Inter-group comparison of InSAR time series analysis test exercise results Select best practices for producing InSAR time series for CGM Overall CGM topics Outline strategy for producing joint GPS-InSAR time series data product, drawing upon findings and discussions of each working group Set specific action items and schedule; consider group proposal for 2015

20. For more information Workshop webpage: http://www.scec.org/workshops/2013/cgm/index.html Workshop report: Murray, J., R. Lohman and D. Sandwell (2013), Combining GPS and Remotely Sensed Data to Characterize Time-Varying Crustal Motion, Eos Trans. AGU, 94(35), 309. Wiki: http://collaborate.scec.org/cgm Mailing lists: http://mailman.scec.org/mailman/listinfo/cgm (general postings) http://mailman.scec.org/mailman/listinfo/cgm-g (for GPS focus-group) http://mailman.scec.org/mailman/listinfo/cgm-i (for InSAR focus-group) Contact information: Jessica Murray (jrmurray@usgs.gov) Rowena Lohman (rolohman@gmail.com) David Sandwell (dsandwell@ucsd.edu)

22. Americas – EWS and IWS, ascending orbits, 12 days A

23. Americas – EWS and IWS, descending orbits, 12 days A

25. CGM InSAR Activity Objectives – develop best high spatial resolution crustal deformation time series using InSAR eventually augmented with GPS. InSAR group is refining test exercise begun in 2013 to compare InSAR time series methodologies starting with the same interferograms. InSAR group plans to meet virtually to refine test exercise, set resolutions and deadlines. Participants will assemble at the 2014 SCEC annual meeting to compare results. All of this is preparation for accurate and frequent acquisitions to be provided by Sentinel-, ALOS-2 and eventually NI-SAR.