Assimilating GRACE Into a Land Surface Model in the Presence of an Irrigation‐Induced Groundwater Trend. Issue 12 (28th December 2019)

• Record Type: Journal Article
• Authors: Nie, Wanshu; Zaitchik, Benjamin F.; Rodell, Matthew; Kumar, Sujay V.; Arsenault, Kristi R.; Li, Bailing; Getirana, Augusto
• Is Part Of: Water resources research. Volume 55:Issue 12(2019)
• Page Start: 11274
• Page End: 11294
• Publication Date: 2019-12-28
• Subjects: GRACE TWS -- data assimilation -- irrigation -- groundwater withdrawal -- water and energy fluxes -- High Plains Aquifer; Hydrology -- Periodicals; 333.91
• DOI: 10.1029/2019WR025363
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: Assimilating terrestrial water storage observations from the Gravity Recovery and Climate Experiment (GRACE) mission into land surface models (LSMs) provides an opportunity to disaggregate and downscale GRACE information to finer scales and improve water component estimates in LSMs. However, the performance of GRACE data assimilation (GRACE‐DA) is limited by the lack of representation of human activities in most LSMs. To simultaneously improve GRACE‐DA and reduce the uncertainties in the modeled anthropogenic processes, we assimilate mascon‐based GRACE terrestrial water storage into the Noah‐Multiparameterization LSM that includes groundwater extraction for irrigation. Simulations with and without GRACE‐DA and with and without groundwater pumping for irrigation are performed to study the isolated and combined effects of groundwater irrigation and GRACE‐DA on water and energy fluxes over the High Plains Aquifer (HPA). The results reveal that the DA‐only simulation may erroneously distribute increments across water storage components and affect the related fluxes through biased feedbacks, while the irrigation‐only simulation may overestimate groundwater decline due to shortcomings in the irrigation and groundwater parameterizations. Assimilating GRACE when irrigation is simulated produces the best overall performance for water storage trends over the northern HPA. For the southern HPA, GRACE assimilation with irrigation performs similarly to irrigation‐only simulation for water storage components and evapotranspiration. GRACE assimilation also improves results in nonirrigated regions and can potentially alleviate the overestimation of groundwater trends in regions with greater irrigation uncertainties. This study highlights the potential to advance hydrological data assimilation in the context of anthropogenic water consumption and land‐atmosphere interactions. Key Points: GRACE TWS data are assimilated into the Noah‐MP land surface model that includes groundwater extraction for irrigation Assimilating GRACE without irrigation produces biased increments, while simulating irrigation without GRACE may exacerbate model error Combining GRACE‐DA and irrigation shows the overall best performance for water storage components and evapotranspiration