California’s Central Valley is one of the most productive agricultural regions in the world. It accounts for 9% of the value of crops grown in the United States on just 1% of its agricultural land area. However, recent droughts (e.g. 2012 - 2016) and climate change jeopardize the future food - water security of the region.
Accurate estimates of changes in groundwater storage are critical to the sustainable management of worldwide aquifers, but few methods exist that can provide timely estimates of aquifer dynamics, especially in regions where groundwater monitoring wells are sparse. Groundwater shortages are increasingly impacting arid regions like California, South Asia and Africa as climate change alters the timing and duration of rainfall, snowmelt, and other hydrologic processes. Excessive pumping of groundwater, which is common during droughts, causes a number of negative externalities including: (1) dry domestic and agricultural wells, (2) damage to groundwater-dependent ecosystems, (3) permanent land subsidence, and (4) aquifer contamination by arsenic and other chemicals.
The proliferation of satellite remote sensing instruments and data allow us to measure hydrologic processes with unprecedented spatial resolution and temporal sampling frequency. By combining remotely-sensed data from a variety of instruments with on-the-ground measurements, we can rapidly estimate changes in regional groundwater storage. Groundwater storage changes are calculated as the residual of inflows (precipitation, streamflow, runoff) minus outflows (evapotranspiration, streamflow), and corrected for fluctuations in reservoir levels, snowpack, and soil moisture. This mass balance approach facilitates efficient computation of changes in groundwater levels integrated over an area, and may be extensible to data-scarce regions of the world, since relies heavily on freely available satellite data.
Figure 1: Map and graph of evapotranspiration data over time in California.
We developed a novel method that synthesizes satellite and ground-based measurements of hydrologic properties (e.g., rainfall and soil moisture) to calculate changes in groundwater storage, and demonstrated this method in California’s Central Valley. The Hybrid Data Remote Sensing Assimilation (HyDRA) system integrates data from more than 15 sources to calculate changes in groundwater storage, and accompanying uncertainties, at multiple spatial scales.
Figure 2: Maps and graphs of changes in Groundwater Storage over time at two spatial scales.
Results agree with gravity measurements (NASA GRACE), physics-based flow models (DWR C2VSIM), and groundwater storage changes calculated from water levels in wells (DWR CASGEM). The correspondence in trend, magnitude, and timing of storage changes confirmed by independent datasets improves our certainty in the amount of groundwater depletion in California occurring over the last ~20 years. Our approach relies heavily on remotely-sensed data; thus it may be readily extensible to other semiarid regions worldwide, and areas with limited access to data.