河套灌区土壤砷含量的季节变化特征模拟

Simulation of seasonal change characteristics of soil arsenic content in Hetao irrigation district

土壤砷(As)是河套灌区粮食安全生产的潜在风险,开展土壤As含量的时空分异机制研究具有重要意义.基于随机森林(RF)和SHAP模型,利用气候、遥感和地形等特征变量对河套灌区秋季采样点土壤As含量进行模拟和解释,并根据气候因子和遥感因子等特征变量的季节变化,对夏季土壤As含量进行预测.结果表明:河套灌区土壤As含量主要受潜在蒸发量、植被指数和高程影响;在中部和东部地势较平缓的地区,与夏季相比,秋季大部分样点土壤As含量增加;而在地势较高的西部,与夏季相比,秋季土壤As含量降低.本研究结果能够为大尺度土壤As的季节变化预测提供科学依据.

Soil arsenic(As)is a potential risk to food safety production in the Hetao irrigation district,and it is of great significance to investigate spatiotemporal differentiation mechanism of soil As content.Based on Random Forest(RF)and SHAP models,the climatic,remote sensing,and topographic variables were used to simulate and explain the soil As content of sampling points in autumn in the Hetao irrigation district.Based on the seasonal variation of the predictive variables such as climate and remote sensing factors,the soil As content in summer was predicted.The results showed that the soil As content in the Hetao irrigation district was mainly influenced by potential evapotranspiration,vegetation index,and elevation.In the central and eastern regions with relatively flat terrain,the soil As content increased in most areas in autumn compared to summer;while in the western regions with higher terrain,the soil As content decreased in autumn compared to summer.The results of this study can provide scientific basis for seasonal variation prediction of soil As content on a large scale.