Global Water Stress Assessment Using a Coupled Hydrological-Socioeconomic Modeling Framework

Abstract

Water is essential for human activities, yet most global water stress assessments treat water supply and demand separately, limiting representation of dynamic environment–society interactions. Water is essential for human activities; we developed a coupled IAM-LSM framework linking the global change analysis model (GCAM) and the Integrated Land Simulator (ILS) through land-use change. We conducted simulations under the SSP1-2.6 scenario for 2020–2100. We evaluated the impact of coupling by comparing simulated monthly river discharge climatology from ILS with observations from the Global Runoff Data Centre (GRDC). The coupled system reduced discharge biases and significantly improved the representation of river flow seasonality (global mean correlation increased from 0.30 to 0.31; p = 0.017), demonstrating a statistically significant though modest improvement in hydrological performance. Using a flexible method to quantify inter-sub-basin water transfers, we assessed water stress and population exposure at the geopolitical sub-basin scale defined by basins and political boundaries at both annual and seasonal scales. While annual and seasonal estimates of spatial distribution and population exposure broadly agreed with previous global studies, additional water-stress hotspots were identified in regions such as South Africa, Argentina, and southeastern Brazil. Our seasonal-scale assessment further revealed water stress and population exposure that are obscured by annual aggregation, reaffirming the importance of seasonal-scale water stress evaluation. Annual exposure to severe water stress reached 2.15 billion people in 2020, whereas seasonal assessment indicated that up to 67% of the global population experienced severe stress in at least one season. Moreover, despite a simplified representation, our estimation of non-surface water use showed strong agreement in global magnitude and spatial patterns with AQUASTAT data (R² = 0.91), indicating that aggregate groundwater and non-conventional water dependence can be approximated from surface water deficits without explicitly modeling individual components. Overall, the coupled IAM–LSM framework provides an internally consistent and scalable basis for assessing global water stress under interacting climatic and socioeconomic changes.