SCEC Award Number 17046 View PDF
Proposal Category Individual Proposal (Data Gathering and Products)
Proposal Title Complementing CGM with Sentinel-1 InSAR data
Name Organization
Yuri Fialko University of California, San Diego
Other Participants Katia Tymofyeyeva
SCEC Priorities 1a, 1e, 2a SCEC Groups Geodesy
Report Due Date 06/15/2018 Date Report Submitted 08/21/2018
Project Abstract
The SCEC Community Geodetic Model (CGM) aims to describe surface deformation in Southern California at highest possible spatio-temporal resolution and accuracy. This requires an optimal integration of GPS and InSAR data. Over the time span of SCEC5, there will be a dramatic increase in the amount of InSAR data thanks to the European Space Agency (ESA) mission Sentinel-1. Sentinel-1 mission will provide several key improvements over the existing InSAR data sets, including: i) frequent and regular acquisitions. The nominal revisit time for the currently operational Sentinel-1A and 1B satellites is 6 days. This can be compared to the minimum revisit time of 35 days for the previous ESA missions such as ERS-1/2 and ENVISAT. ii) A smaller revisit time not only improves temporal resolution, but also significantly reduces problems with decorrelation of the radar phase, and helps mitigate atmospheric artifacts by virtue of averaging. iii) Wide-swath capability. 300-km-wide swathes of Sentinel-1 ensure a complete coverage of Southern California with just a few tracks. iv) Uniform coverage from both ascending and descending satellite orbits. Data from two different look directions allow us to separate horizontal and vertical components of surface displacements. Incorporation of Sentinel-1 data is therefore expected to result in a significant improvement of CGM. Over the last year we have set up a system for routine systematic processing of all Sentinel-1 data from Southern California. We also started generating higher-level products for integration into CGM.
Intellectual Merit Interferometric Synthetic Aperture Radar (InSAR) data are increasingly used to image deformation due to active faults. Frequent InSAR acquisitions are expected to provide an improved signal to noise ratio for low-amplitude deformation signals. However, we find that increasing number of radar interferograms used in the analysis of surface deformation results in the accumulation of high-frequency spatial noise, which is introduced in the time series analysis due to filtering of the radar phase prior to unwrapping. We propose a new method for "unfiltering" the filtered unwrapped radar phase. We demonstrate the feasibility of the proposed method using Sentinel-1 InSAR and Global Positioning System (GPS) data. We combine data collected by Sentinel-1 between 2014-2018 with continuous GPS measurements to calculate the three components of the surface velocity field over the southern San Andreas and San Jacinto fault zones at the resolution of InSAR data (~100 m). We obtain the 3 orthogonal components of surface motion using overlapping InSAR tracks with different look geometries, together with an additional constraint provided by GPS measurements of the local azimuth of the horizontal velocity vector. We estimate both secular velocities and displacement time series. The latter are calculated by combining InSAR time series from different lines of sight with time-dependent azimuths computed using continuous GPS time series at every InSAR epoch. We use CANDIS method, a technique based on iterative common point stacking, to correct the InSAR data for tropospheric and ionospheric artifacts when calculating secular velocities and time series, and to isolate low-amplitude deformation signals in our study region. This three-component, time-dependent description of surface deformation from a combination of geodetic data sets can be used as part of the SCEC Community Geodetic Model.
Broader Impacts Evaluation of seismic hazard is based primarily on historic seismicity and long-term fault slip rates inferred from paleoseismic data. Geodetic observations provide an important additional source of information about contemporaneous accumulation of strain in the seismogenic layer. UCERF3 model now incorporates estimates of fault slip rates based on geodetic data. A major outstanding question is whether geodetic observations can help identify areas of seismic hazard that haven't been recognized based on available seismic and geologic data. While mature faults such as the San Andreas fault by and large have clear expression in geomorphology, young developing faults and fault zones may be more difficult to recognize. Sentinel-1 InSAR data will help us better understand a potential contribution of geodetic observations to estimates of seismic hazard such as UCERF. The proposed collection and analysis of space geodetic data will improve our understanding of the associated seismic hazard to populated areas in Southern California. This project has provided training and support for one graduate student.
Exemplary Figure Figure 2:
Secular velocity in the plate-parallel direction, calculated from the combination of ascending and descending Sentinel-1 acquisitions, and horizontal azimuths obtained from secular GPS velocities (CMM4). GPS velocities (colored circles) are plotted for comparison. Dashed blue lines indicate profiles across the plate boundary shown in Figure 3.