Identifying the facts and driving factors of deceleration of groundwater table decline in Beijing during 1999–2018

Abstract

Domestic water supply and agricultural irrigation are highly dependent on groundwater in Beijing. Groundwater table (GWT) decline and ecological issues caused by groundwater overexploitation have been of growing concern. However, some positive measures such as groundwater extraction control, improvement of water use efficiency, and South-to-North Water Diversion Middle Route Project (SNWDMRP) have alleviated severe groundwater shortage situation in recent years. Therefore, quantifying the contributions of different factors to the deceleration of groundwater table decline contributes to guide groundwater resources. In this study, in-situ groundwater table data, climate data, and water use data were used to identify the deceleration and driving factors of GWT decline by comprehensively using water table fluctuation (WTF) method, Mann-Kendall (MK) trend test, wavelet analysis methods, and Random Forest Regression (RFR). Firstly, the MK trend test evaluated the trend of the defined groundwater stress index (GSI) and groundwater table decline rate (GTDR) changes, indicating the fact that GWT decline was decelerated during 1999–2018. Then, the WTF method combined with in-situ groundwater level data and specific yield were used to derive the monthly groundwater storage anomalies (GWSA), which revealed a slight GWS recovery in the past decade, especially a significant increase of GWS with a rate of 5.76 mm/month since the operation of South-to-North Water Diversion Middle Route Project (SNWDMRP). Wavelet analysis identified a correlation and coherence between GWSA and precipitation in the time–frequency domain. Furthermore, correlation analysis indicated that the deceleration of GWT decline may be attributed to climatic factors (e.g., precipitation, temperature), the SNWDMRP, and water supply and use structure. The RFR quantitatively identified the greatest contribution of water use (e.g., agricultural and domestic water use) to GWT dynamics. Moreover, the effect of SNWDMRP is greater than climatic factors (e.g., precipitation, temperature). These findings have important implications for groundwater resource management in Beijing and other areas suffering from severe groundwater depletion.