Groundwater storage variations in the North China Plain from GRACE with spatial constraints
Feng Wei,Wang Chang-Qing,Mu Da-Peng,Zhong Min,Zhong Yu-Long,Xu Hou-Ze
DOI: https://doi.org/10.6038/cjg20170502
2017-01-01
Abstract:As the largest agricultural production area and an industrial, political and economic center of China, the North China Plain (NCP) is facing severe water shortages. Here we present a joint analysis of groundwater storage (GWS) variations in the NCP using the Gravity Recovery and Climate Experiment (GRACE) satellite data, hydrological models, phreatic well observations, and groundwater models. Spatial constraints with the Tikhonov regularization are applied to original temporal gravity field products from GRACE, to minimize the leakage-out effect of mass variations in the NCP. From 2002 to 2014, GRACE detects significant GWS depletion in the NCP, at a rate of -7.40 +/- 0.9 km(3) . a(-1). The GWS depletion rate in shallow unconfined aquifers of NCP estimated from phreatic well observations is -1.2 km(3) . a(-1), far smaller than the GRACE estimate. The significant difference between them indicates the large GWS depletion in the confined aquifers of NCP. From 2002 to 2008, the GWS rate from GRACE is 5.3 +/- 2.2km(3) . a(-1), which agrees well with the mean rate from two regional groundwater models (i.e., -5.4 km(3) . a(-1)). Good agreement between them confirms the potential of GRACE to estimate GWS variations in the NCP. Besides the long-term depletion of GWS in the NCP, significant interannual GWS variations are detected by GRACE and well observations, which are related to precipitation anomalies in this region. During the drought years, i.e., 2002, 2005-2009, and 2014, GWS in the NCP decreases correspondingly. In spatial domain, large GWS depletion in the piedmont region of the Taihang Mountains and central plains of NCP is detected by GRACE, which agrees with well observations and regional groundwater models. Comparing with GRACE and regional groundwater models, global hydrological models, which simulate groundwater changes, fail to estimate GWS variations in the NCP with an acceptable accuracy or a reasonable spatial pattern. For example, WGHM overestimates the GWS rate in the NCP, while PCR-GLOBWB fails to simulate the spatial pattern of GWS in the NCP. This study indicates that GRACE provides an important observation tool to assess GWS variations in the NCP.