Sensitivity analysis and parameter estimation of anthropogenic water uses for quantifying relation between groundwater overuse and water stress in Mongolia

Abstract

In Mongolia, overuse and degradation of groundwater is a serious issue, mainly in the urban and economic hub in Ulaanbaatar, and the Southern Gobi mining hub. The authors have recently applied a process-based eco-hydrology model, NICE (National Integrated Catchment-based Eco-hydrology) to these two contrasting river basins in order to explicitly quantify spatio-temporal variations in water availability. In this study, they further extended the model to couple with inverse method for sensitivity analysis and parameter estimation of anthropogenic water uses (NICE-INVERSE). The model quantified the spatio-temporal variations of livestock water use in these river basins in comparison with the estimated values by aggregating statistical data, GIS data, observation data, and field surveys in their previous studies. The amount of water use by animals were generally small for each soum (district), and could also be heavily returned back to the ecosystems. The result also showed a temporal decreasing trend of unit water use in some typical livestock (cattle, sheep, and goats), suggesting a substantial increase in water stress due to local-regional eco-hydrological degradation by urbanization and mining. These results clarified the substantial increase in water stress of groundwater degradation was mainly caused by urbanization and mining locally in addition to climatic change. Sensitivity analysis and inverse estimation of model parameters helped to improve the accuracy of hydrologic budgets in basins. This methodology is powerful for evaluating spatio-temporal variations of water availability in regions with fewer inventory data on urbanization, mining, and livestock.