Impacts of thickening unsaturated zone on groundwater recharge in the North China Plain

•Coupling soil water balance and unsaturated zone flow model to estimate recharge.•Dynamic bottom boundary is necessary for unsaturated flow modeling.•Thickening unsaturated zone can result in significant reduction in recharge.•Significant part of percolation can be converted to unsaturated zone storage. Unsustainable groundwater development shown by rapid groundwater depletion in the North China Plain (NCP) underscores the need to quantify spatiotemporal variability in groundwater recharge for improved management of the resource. The objective of this study was to assess spatiotemporal variability in recharge in response to thickening of the unsaturated zone in the NCP. Recharge was estimated by linking a soil water balance (SWB) model, on the basis of monthly meteorological data, irrigation applications, and soil moisture monitoring data (1993–2008), to the water table using a deep unsaturated zone flow model. The dynamic bottom boundary (water table) position was provided by the saturated zone flow component, which simulates regional pumping. The model results clearly indicate the effects of unsaturated zone thickening on both temporal distribution and magnitude of recharge: smoothing temporal variability in recharge, and increasing unsaturated storage and lag time between percolation and recharge. The thickening unsaturated zone can result in average recharge reduction of up to ∼70% in loam soils with water table declines ⩾30m. Declining groundwater levels with irrigation sourced by groundwater converts percolation to unsaturated zone storage, averaging 14mm equivalent water depth per year in mostly loam soil over the study period, accounting for ∼30% of the saturated groundwater storage depletion. This study demonstrates that, in thickening unsaturated zones, modeling approaches that directly equate deep drainage with recharge will overestimate the amount and underestimate the time lag between percolation and recharge, emphasizing the importance of more realistic simulation of the continuity of unsaturated and saturated storage to provide more reliable estimates of spatiotemporal variability in recharge.