Achieving sustainable agricultural groundwater use remains a significant challenge worldwide in part because data on groundwater withdrawals are limited. A new generation of interferometric synthetic aperture radar (InSAR) data provide high spatial- and temporal- resolution measurements of subsidence and uplift of the earth’s surface, which are known to reflect change in groundwater storage below. Here, we establish for the first time a quantitative link between local irrigation water demand and InSAR-derived vertical land surface displacements in California’s San Joaquin Valley, with the goal of increasing the utility of remotely sensed displacements for understanding and managing groundwater at fine scales. We relate 100 m, sub-monthly displacements to estimates of irrigation water demand generated from land cover and weather data by performing a suite of physical process-motivated statistical analyses that leverage (i) temporal variations in surface water supplies created by climate, and (ii) spatial variations in water demand created by different land uses. Between 2015 and 2017, cultivated land experienced up to seven times the total subsidence, and up to nine times the dry year subsidence rate, of uncultivated land. In uncultivated areas, subsidence rates differed minimally across dry and wet years. In contrast, within cultivated areas, dry year subsidence rates were more than double wet year rates, indicating increased agricultural groundwater pumping under diminished precipitation and surface water supplies. Mean subsidence, and the marginal response of subsidence to water demand, were greatest for cultivated lands and for field and pasture crops in particular, and were also proportional to distance from surface water supply infrastructure. These findings demonstrate that land surface observations have the potential for use in the quantification of connected surface water and groundwater processes at policy-relevant scales.
