Application of Radar Interferometry to the Detection of Surface Deformation in Southern California
Eneva, Mariana1; Adams, David1; Falorni, Giacomo2; Morgan, Jessica2; Novali, Fabrizio3; Allievi, Jacopo3
1Imageair Inc., UNITED STATES; 2TRE Canada, CANADA; 3TRE Italy, ITALY

InSAR (interferometric synthetic aperture radar) is applied in Imperial Valley of southern California to detect and characterize surface deformation in existing geothermal fields, possible future geothermal developments, and around faults. The study area is part of the Salton Trough, where ongoing regional extension is caused by the relative movements of the North American and Pacific plates. There is also localized tectonic deformation associated with fault networks, pull-apart basins, and rotational blocks. The area is characterized by a high heat flow, which is the reason for the existence of a number of current and prospective geothermal fields. The operation of some of the existing geothermal fields represents a man-made cause of surface displacements in the area. At the same time, Imperial Valley is covered by agricultural lands, which are very susceptible even to small surface deformation.

The data used are from the Envisat satellite, collected over the period 2003-2010. The specific InSAR technique applied is SqueeSARTM, the latest extension of an earlier PSInSARTM technique developed at TRE Italy. It identifies permanent and distributed scatterers (PS and DS), which play the role of numerous benchmarks throughout the study area. Deformation time series are obtained at thousands of individual scatterer locations. Their slopes are used to estimate the annual deformation rates. The SqueeSAR technique is particularly suitable for vegetated and rural areas, thus providing results from the agricultural lands of Imperial Valley, where conventional methods, such as differential InSAR (DInSAR) have not worked before. The SqueeSAR results are first obtained in the line-of-sight (LOS) to the satellite. Using measurements from two geometries, ascending and descending, makes it possible to decompose the two LOS movements into vertical and horizontal displacements.

Significant subsidence is observed at all geothermal fields operating during the period of satellite data (2003-2010). These include parts of the Salton Sea geothermal field, and the Heber and East Mesa geothermal fields. At Heber, uplift is also detected in an area adjacent to the subsidence. The results from SqueeSAR are in agreement with the ground-based measurements from the annual leveling surveys carried out at both Salton Sea and Heber. Regional GPS data and relocated earthquakes are also used to inform our analysis results. There are arguments that the surface deformation at the Salton Sea geothermal field may be mostly due to tectonics, while the displacements at Heber and East Mesa are mostly man-made.

Pre-production surface deformation is detected at sites, where production either started after, or at the end of the period covered by satellite data. These are a new development at the Salton Sea geothermal field (Hudson Ranch–I) and North Brawley. The baseline deformation at sites of geothermal potential not yet in development is also determined, such as areas of interest to the U.S. Navy Geothermal Program Office and Orita (formerly East Brawley). This will make it possible to distinguish between tectonic and man-made causes of surface deformation in the future, unlike the current fields, for which there is no pre-production information.

In addition, the SqueeSAR measurements reveal differential movements on both sides of faults, clearly marking the Superstition Hills fault, parts of the San Andreas fault, and the Imperial fault. Since the former two faults are in relatively dry areas, conventional DInSAR can detect those differences as well, although SqueeSAR provides more details. However, the Imperial fault transects agricultural fields, where DInSAR does not work, so our SqueeSAR observations are unique.

We also detect surface deformation effects on the strike-slip Superstition Hills fault, caused by an October 2006 aseismic event and by a M7.2 earthquake that occurred south of the U.S. - Mexico border in April 2010.

We thus conclude that InSAR provides unprecedented information on surface deformation in Imperial Valley, as long as a suitable technique, such as SqueeSARTM, is used to tease out signals from the agricultural areas. Such observations can be effectively used for pre-production geothermal reservoir assessment, feedback to mitigate any environmental impact that might occur in operating fields, exploration efforts to identify suitable geothermal drilling targets, and detecting the effects of regional earthquakes and local aseismic events.

Figure 1. Example of deformation along a profile and mean deformation rates within polygons from the Heber geothermal field. (a) Map of interpolated LOS ascending deformation rates (in mm/year) referenced to leveling benchmark A-33 (solid black triangle). Locations of other leveling benchmarks are shown with smaller empty triangles. Rates are color coded as shown by vertical bar to the right of the map. Small crosses show M>0 earthquake epicenters for the period January 1981-June 2011. Straight black line in the NW-SE direction shows a profile, along which LOS ascending deformation rates are illustrated in (b). Mean deformation time series (in mm) for the area outlined with black polygon in the uplift area (western part of map) are shown in (c), and for the polygon in the subsidence area (central part of map) in (d). Green lines and symbols show time series at leveling benchmarks within the polygons. Red and dark blue lines and symbols show mean LOS deformation series from the ascending and descending data, respectively. Pink and light blue symbols and lines show decomposed time series indicating mean vertical and east horizontal movements, respectively. Circles in (c) and (d) show times of earthquakes in the respective polygons.