Improved water storage estimates within the North China Plain by assimilating GRACE data into the CABLE model

JOURNAL OF HYDROLOGY

卷:590

Article

Yin, Wenjie^1,2  Han, Shin-Chan²  Zheng, Wei¹  Yeo, In-Young²  Hu, Litang³  Tangdamrongsub, Natthachet^4,5  Ghobadi-Far, Khosro² 

[1] China Acad Space Technol, Qian Xuesen Lab Space Technol, 104 Youyi Rd, Beijing 100094, Peoples R China

[2] Univ Newcastle, Sch Engn, Callaghan, NSW, Australia

[3] Beijing Normal Univ, Coll Water Sci, Beijing 100875, Peoples R China

[4] Univ Maryland, Earth Syst Sci Interdisciplinary Ctr, College Pk, MD 20742 USA

[5] NASA, Hydrol Sci Lab, Goddard Space Flight Ctr, Greenbelt, MD USA

关键词: GRACE;    CABLE model;    The North China Plain;    Data assimilation;    Ensemble Kalman Filter;   

DOI  :  10.1016/j.jhydrol.2020.125348

来源: Elsevier

【 摘 要 】

As one of the most important crop-producing bases, the North China Plain (NCP) has suffered serious groundwater depletion due to drying climate and intensive human activities. An accurate estimation of groundwater storage (GWS) is of great importance for the sustainable development of local society and region's economy. This study investigates a methodology to determine improved GWS variation throughout the NCP based on assimilation of terrestrial water storage (TWS) data from the Gravity Recovery and Climate Experiment (GRACE) data into the Community Atmosphere Biosphere Land Exchange (CABLE) model. We found that the CABLE model is not able to compute the prolonged decrease (and accelerated since 2013) in the GWS change around the southwestern region of the NCP where major cities are located. The model incorrectly characterizes the spatial patterns of interannual and annual changes of water storage variation, mainly caused by errors in the precipitation forcing data. The interannual TWS change is indicative of GWS, while the observed seasonal variation is primarily that of root-zone soil moisture. The GRACE data assimilation most effectively improves GWS computation. The GWS assimilation results were validated against a total of 156 in-situ groundwater level data in the NCP. Compared to the model computation, there was a significant improvement in terms of cross correlation, on average, from 0.12 (before assimilation) to 0.54 (after assimilation). This study demonstrates the effectiveness of GRACE data assimilation toward reliable estimation of ground water storage variation in the NCP, and its promise to quantify the potential implication of water supply from the South-to-North Water Transfer Project within the NCP.

【 授权许可】

Free