A Quantitative Investigation of the Thresholds for Two Conventional Water Scarcity Indicators Using a State-of-the-Art Global Hydrological Model With Human Activities

Among the published global water resource assessments, regions where annual total water withdrawal exceeds 20% and 40% of the availability, or where per capita water availability falls below 1,700 and 500 m(3).person(-1).year(-1), are categorized as being moderately and highly water stressed, respectively, but the rationale for using these thresholds has not been fully explained. Here we show that these thresholds represent the sufficiency of local and renewable water resources, which are estimated daily by a state-of-the-art global hydrological model. We found that water abstraction from local and renewable sources is insufficient (i.e., more than 1% of the annual total water requirement becomes unmet) in a majority of grid cells where annual total water withdrawal exceeds 20% of the availability and/or per capita water availability falls below 1,700 person(-1).year(-1) (moderate stress). In grid cells where the corresponding values are 40% and/or below 500 m(3).person(-l).year(-1) (high water stress), more than 20% and 60% of the annual total water requirement must be supplied by nonlocal and nonrenewable sources, respectively. These results indicate that conventional indicators and thresholds can be used as a surrogate for measuring daily variations in the sufficiency of local and renewable water sources at the global scale. At the same time, the substitutability of the water withdrawal to availability (WTA) and the water availability per capita (APC) varies by region, which can be primarily attributed to the intensity of the seasonal variation in river flow and irrigated area per capita, respectively. Based on the analyses, new sets of thresholds for WTA and APC are proposed.