The Use of InSAR Data to Map Hydraulic Head Levels in the San Luis Valley, Colorado

Abstract:

A measurement of hydraulic head in the aquifer is a critical metric needed for effective water management. Here we report on the use of InSAR data to inform groundwater management in the San Luis Valley (SLV), Colorado. We selected the SLV for study, as it is an important agricultural area struggling with groundwater management after a period of extreme drought. The techniques developed here can also be used to assist groundwater management in other agricultural areas, such as California’s Central Valley.

Reeves at al. (2013) analyzed 30 C-band ERS scenes over the SLV between 1992 and 2000 using the small baseline subset method (SBAS). They concluded that InSAR data can be used to calibrate the subsidence-to-head relationship and temporally extend hydraulic head measurements. To study the extent to which InSAR deformation time series can be used to interpolate spatial gaps in the well-based head measurements, we need to derive an InSAR ground deformation map due to groundwater extraction over the entire SLV. A major error source of InSAR data is highly variable temporal decorrelation over agricultural areas such as the SLV, which limits the number of InSAR pixels that can be selected for the SBAS analysis. Because L-band spaceborne radar systems such as ALOS PALSAR suffer less from temporal decorrelation than C-band and X-band spaceborne radar systems, we processed 92 small baseline interferograms using 17 ALOS scenes acquired between January 2007 and March 2011. We further employed an SBAS ground deformation model to represent surface subsidence due to groundwater extraction, which consists of a seasonal deformation superimposed on a linear trend. We observed that the long-term subsidence rate is smaller than 5 mm/year over the entire SLV and the magnitude of the seasonal subsidence can be up to 2 cm. Our results indicate that there is no severe long-term water storage loss in the confined aquifer at the SLV, consistent with previous C-band InSAR study (Reeves at al., 2013) and well-based hydraulic head data. The next step in our research is to estimate Sk through calibration using co-located InSAR subsidence map and hydraulic head data measured at 36 confined aquifer wells. We can then interpolate the estimates of Sk spatially and reconstruct a hydraulic head map over the entire SLV jointly using InSAR and well data.