Remote sensing of wetland evolution in predicting shallow groundwater arsenic distribution in two typical inland basins.

A large number of studies have shown that the existence of wetlands may influence arsenic concentrations in adjacent shallow groundwater. However, little is known about the linkage between wetland evolution and arsenic enrichment in shallow groundwater. This study investigated wetland evolutions from 1973 to 2015 in two arid-semiarid inland basins along the Yellow River catchment (i.e., the Yinchuan Basin and the Hetao Basin) based on remote sensing data, and their association with arsenic distributions based on arsenic concentrations of 244 and 570 shallow groundwater samples, respectively. The long-term Landsat images reveal that the covering area of wetlands exhibited increasing trends in both the Yinchuan Basin and the Hetao Basin. Wetlands in the Yinchuan Basin and the Yellow River water-irrigation area in the Hetao Basin varied with precipitation before 2000, but exhibited increasing trends because of wetland restoration policies since 2000. Wetlands in groundwater-irrigation area in the Hetao Basin decreased due to increasing exploitation of shallow groundwater. Wetlands with long existence time were mainly distributed along the Yellow River and drainage channels and in large lakes in the northern Yinchuan Basin and the Hetao Basin, where high‑arsenic (>10 μg/L) groundwater occurred. The probability of high‑arsenic groundwater distribution increased with the proportion of wetland existence time to the entire studied period (42 years), which can be best explained by a BiDoseResp growth curve. Longer existence of wetlands may cause greater probability of high‑arsenic groundwater. This was likely related to long-term introduction of biodegradable organic matter into shallow aquifers and thereafter enhancement of arsenic mobility and/or arsenic being released beneath wetlands and transported into shallow aquifers under continuing wetland water recharge. We therefore suggest that mapping wetland evolutions could probably serve as a good indicator for predicting high arsenic groundwater distributions in shallow aquifers.