Subsidence-Derived Aquifer Volume Strain Models for the San Joaquin Valley, California and Central Arizona

Abstract:

Across the southwestern United States, groundwater overdraft causes large-scale subsidence. The San Joaquin Valley (SJV), in the southern Central Valley in California, has been subsiding since the early 1900’s, causing infrastructure damage costing more than a billion dollars. Interferometric synthetic aperture radar (InSAR), due to its high spatial (~200 m) and temporal resolution (< monthly), has been essential in estimating the magnitude of subsidence and groundwater loss in the region. In combination with groundwater monitoring wells, the deformation data can be used to solve for various poroelastic parameters, including the quasi-static bulk modulus. This parameter is measured under a condition in which pore pressure change drives volume strain at a slow deformation rate, which is the case for subsidence induced by groundwater loss. Here, we use InSAR-derived vertical land motion acquired from Dec. 24, 2006-Jan. 1, 2010 to model the volume strain rate with depth in the SJV. We find a maximum volume strain rate of -232 microstrain/yr at a depth of 0-200 m in Tulare and Kings County, California. We compare the InSAR-derived strain rate to aquifer pressure change in 300 observation well time series and estimate the quasi-static bulk modulus to be 0.19 GPa. In addition, we recognize that differential subsidence can lead to the development of extensional earth fissures, and so we use our volume strain model to generate an earth fissure hazard map. This map utilizes the R-ratio, which is the ratio of tensile stress to sediment tensile strength. Because the SJV contains few known surface fissures, we cannot verify whether our R-ratio map predicts fissure locations. Therefore, in order to verify this methodology, we use the same technique in central Arizona, a region that has also experienced groundwater extraction- induced subsidence and where up-to-date fissure maps are available through the Arizona Geological Survey. In the Arizona study region, we use InSAR LOS displacements from two different acquisition periods, Feb. 2004-Oct. 2010 and Dec. 2014-Aug. 2017, to generate volume strain models and R-ratio maps. In general, we find that mapped fissures correspond to areas with a predicted R-ratio that indicates a high risk of fissures, confirming the applicability of this method to help predict this hazard.