Implications of Assimilating GRACE into a Land Surface Model with human-water interface: Improving Water Component Estimates in High Plains Aquifer

Assimilating observations of terrestrial water storage (TWS) from the Gravity Recovery and Climate Experiment (GRACE) satellite mission into Land Surface Models (LSMs) provides an opportunity to disaggregate and downscale GRACE information into finer scales and improve the water component estimates in LSMs. However, previous studies show that GRACE data assimilation (GRACE-DA) may degrade simulations of soil moisture and evapotranspiration for groundwater depletion areas due to unmodeled human water management processes. This limits the value of GRACE-DA for water resource analysis and poses a problem for GRACE-DA application to drought monitoring and forecasting anywhere that extraction has affected the long-term water storage trend. To address these limitations, we assimilate the JPL GRACE mass concentration product into Noah-MP LSM that has been modified to include groundwater extraction for irrigation. A set of four experiments were performed to study the impact of groundwater irrigation and GRACE-DA on the simulation of water and energy fluxes and the representation of a recent drought in High Plains Aquifer (HPA) region. The results show that a simulation with groundwater extraction but without GRACE-DA improved groundwater estimates for Kansas and surface soil moisture over the HPA but overestimated the TWS decline for Nebraska during the post-drought period. The GRACE-DA only simulation showed less skill at groundwater and no improvement for soil moisture. The combined system lead to overall better estimates with improved TWS, surface soil moisture and groundwater. This study highlights the potential to advance hydrological data assimilation within a context of anthropogenic water regulation.