Predictive modeling of soil water and salt dynamics for ecological restoration of mines in arid zones

Ecological restoration of mines in arid zones is an important initiative to promote the harmonious coexistence of man and nature and realize green and low-carbon development, in which the prevention and control of soil secondary salinization is crucial for ecological restoration. This paper takes the soil and water environment (soil secondary salinization) of the Shendong mining area as the research object, and after systematically analyzing the ecological restoration of the study area, collects relevant soil water and salt data and processes them, and analyzes the characteristics of soil water and salt changes in the study area respectively. On this basis, combined with the basic mathematical model of soil water-salt transport, a 3-layer feed-forward BP network modeling method was invoked to predict the soil water-salt dynamics under the study’s comprehensive conditions, and then the sensitivity of soil salinity and soil alkalinity to each factor in the input layer was analyzed by using the default factor test method, and predictions were made for long-term ecological restoration. The sensitivity indices of soil water content to soil salinity as well as soil alkalinity were large, with R of 4.46 and 3.89, respectively, and the effects of each factor on soil water-salt dynamics should be considered comprehensively in the actual ecological restoration of mines in arid zones. In addition, through the prediction of soil water salinity dynamics, the crops in the study area were irrigated with 3.2g/L brackish water four times during the reproductive period, with an irrigation quota of 360mm, and with the measure of spring irrigation water volume of 148mm, which could achieve the purpose of reasonable utilization of salty water and water-saving irrigation, and effectively realize the ecological restoration of mines in arid zones.