Spatial Heterogeneity of Groundwater Hydrochemistry, Quality, and Degradation Mechanisms in Hyperarid Endorheic Watersheds: Insights From the Qaidam Basin on Tibetan Plateau

Abstract

Phreatic groundwater sustains ecosystems and human livelihoods in hyperarid endorheic watersheds but faces severe threats from widespread nitrogen pollution and extreme salinity. This study systematically investigates water quality suitability, nitrogen health risks, and geochemical degradation mechanisms in the Golmud River watershed on the Tibetan Plateau. The results indicate that the regional groundwater is generally alkaline and features relatively high mineralization. Water quality suitability is relatively poor, and only 37.78% (EWQI < 100) of the groundwater is suitable for daily human use. Water quality exhibits a continuous and sharp deterioration trend from pristine upstream mountainous areas to severely polluted downstream salt marsh plains. This poses extensive health risks, which mainly arise from nitrogen contamination. These risks exhibit significant population heterogeneity and spatial consistency among different groups. NO2− is the primary nitrogen species causing health threats. Infants face the most severe health risks among all the evaluated populations. High nitrogen risks areas in the watershed are concentrated in the downstream salt marsh plain and the agricultural zones of the loess plain. Strong evaporative concentration under extreme drought and anthropogenic pollution from agricultural expansion are the core mechanisms degrading phreatic groundwater quality and increasing nitrogen risks. Intensive agricultural activities in the loess plain directly trigger NO2− nonpoint source pollution. Extreme evaporation and weak hydrodynamics in the downstream salt marsh plain highly enrich hydrochemical components and cause long‐term accumulation of industrial nitrogen from salt lake mining. These findings provide a reference for the pollution prevention and sustainable development of hypersaline groundwater globally.