Sensitivity of Groundwater Depletion Rates Estimated from Different GRACE Products and Water Balance Models of Different Spatial Scale

Abstract:

Observations of water storage anomalies monitored by the Gravity Recovery and Climate Experiments (GRACE) satellite mission have recently been used to estimate regional-scale groundwater depletion. Different GRACE products are available such as the JPL 1°x1° GRACE Tellus dataset and the University of Colorado (CU) dataset of regional-averages. In addition, in order to estimate groundwater depletion rates, GRACE data must be used in conjunction with auxiliary hydrologic and land-use information such as soil moisture, vegetation types, snow storage, etc. These auxiliary datasets are usually estimated using coarse scale (usually 1°x1°) hydrological/land surface observations and models. Therefore, it is important to quantify the sensitivity of estimates of groundwater depletion to different GRACE products, types of auxiliary datasets, and different spatial scales. In this study we show results of a sensitivity analysis of groundwater depletion trends for the Central Valley aquifer of California using three different GRACE products: i) GRACE Tellus, ii) CU GRACE data and, iii) Scaled CU data (filtered CU data optimally scaled to capture peaks in observed water balance). Using soil moisture and snow water storage estimated from 1/8^th degree Variable Infiltration Capacity (VIC) model and observed surface water storages (e.g., reservoirs etc.), groundwater depletion from April 2006-March 2010 for three GRACE products is estimated as 15.8, 13.9 and 36.3 km³ respectively. For the time period April 2006-September 2009 these estimates are 7.95, 5.5 and 29.2 km³, respectively. In addition, a sensitivity analysis with respect to spatial scale of VIC model will be presented, which involves using estimates of soil moisture and snow storage at the following three different spatial scales: i) 1 degree, ii) 1/8^th degree and, iii) 1/16^th degree.