A model comparison assessing the importance of lateral groundwater flows at the global scale

Current global-scale models of water resources do not generally represent groundwater lateral flows and groundwater-surface water interactions. But, models that do represent groundwater in more detail are becoming available and this raises the question of how estimates of water flow, availability, and impacts might change compared to previous global estimates. In this study, we provide the first global quantification of cell-to-cell groundwater flow (GWF) using a high-resolution global-scale GWF model and compare estimated impacts of groundwater pumping using two model setups: (a) with and (b) without including cell-to-cell GWFs and realistic simulation of groundwater-surface water interactions at the global scale (simulated over 1960-2010). Results show that 40% of the land-surface cell-to-cell flows are a notable part of the cell's water budget and that globally large differences in the impact of groundwater pumping are estimatd between the two runs. Globally, simulated groundwater discharge to rivers and streams increased by a factor of 1.2-2.2 when GWFs and interactions between groundwater and surface water were included. For eight heavily pumped aquifers, estimates of groundwater depletion decrease by a factor of 1.7-22. Furthermore, our results show that GWFs and interactions between groundwater and surface water contribute to the volume of groundwater that can be pumped without causing notable changes in storage. However, in approximately 40% of the world's watersheds where groundwater is used, groundwater is being pumped notably at the expense of river flow, and in 15% of the area globally depletion is increased as a result of nearby groundwater pumping. Evaluation of the model results showed that when groundwater lateral flows and groundwater - surface water interactions were taken into account, the indirect observations of groundwater depletion and groundwater discharge were mimicked much better than when these fluxes were not included. Based on these findings, we suggest that including GWFs in large-scale water resources assessments will benefit a realistic assessment of groundwater availability worldwide, the estimation of impacts associated with groundwater pumping, especially when one is interested in the feedback between groundwater use and groundwater and surface water availability, and the impacts of current and future groundwater uses on the hydrological system.