Changes of groundwater arsenic risk in different seasons in Hetao Basin based on machine learning model

Abstract

Arsenic pollution of shallow groundwater is serious in Hetao Basin. At present, there are few studies on the seasonal variation and mechanism of high As groundwater. In order to master the risk difference and influence mechanism of high As groundwater in different seasons, we collected 506 shallow groundwater samples in the Hetao Basin, and used climatic factors, topographic factors, and others (influence of irrigation channels, vegetation index) that are closely distributed with As in groundwater to establish a high-precision random forest model of high As groundwater in the Hetao Basin in summer. We used climate factors as dynamic predictors to predict the distribution of high As risks in winter and established human health risk zones in the Hetao Basin. The results show that from winter to summer, the probability of high As in high risk areas further increases with the influence of factors such as temperature increase, rainfall increase, and enhanced evapotranspiration, while the probability of high As in low risk areas is the opposite and shows a downward trend. The areas with increased probability of high human health risks and stable areas are mainly distributed along the drainage canals and concentrated in the middle of the basin. From winter to summer, as the local residents' demand for groundwater increases, the probability of high As has increased and stabilized in high risk areas. The number of threatened populations reached 246,000 and 108,000, respectively. Therefore, we need to focus on them. The results of this research explored the changing trend and mechanism of high As groundwater risks under the influence of climate, further enriching the regional high As groundwater research system, and can also be provided as a reference for similar research in other regions.

Keywords: Arsenic enrichment; Groundwater; Hetao Basin; Random forest; Risk change.