银川平原地下水水化学特征及基于随机森林模型的砷健康风险预测  

Hydrochemical characteristics of groundwater in Yinchuan Plain and health risk prediction of arsenic using random forest model

作　　者：张娟娟[1,2] 李祥志 付世骞 孙玉芳 赵莉花 任宇 ZHANG Juanjuan;LI Xiangzhi;FU Shiqian;SUN Yufang;ZHAO Lihua;REN Yu(Institute of Hydrogeology and Environmental Geology,Chinese Academy of Geological Sciences,Shijiazhuang 050061;Sixth Geological Brigade of Hebei Geological and Mineral Exploration and Development Bureau(Aviation Measurement Application Center of Hebei Geological and Mineral Exploration and Development Bureau),Shijiazhuang 050000;Key Laboratory of Groundwater Pollution Mechanisms and Remediation,Hebei Province/China Geological Survey,Shijiazhuang 05006l;Ningxia Hui Autonomous Region Hydrogeological and Environmental Geological Survey Institute,Yinchuan 750026;Hebei Provincial Academy of Water Resources,Shijiazhuang 050057,China)

机构地区：[1]中国地质科学院水文地质环境地质研究所,石家庄050061 [2]河北省地质矿产勘查开发局第六地质大队(河北省地质矿产勘查开发局航空测量应用中心),石家庄050000 [3]河北省/中国地质调查局地下水污染机理与修复重点实验室,石家庄050061 [4]宁夏回族自治区水文环境地质调查院,银川750026 [5]河北省水利科学研究院,石家庄050057

出　　处：《干旱区资源与环境》2025年第1期117-127,共11页Journal of Arid Land Resources and Environment

基　　金：河北省中央引导地方科技发展资金项目(246Z3601G);宁夏地下水与生态环境效应研究人才小高地项目资助。

摘　　要：银川平原是我国饮水型砷中毒分布区之一。受限于样本获取难度,从区域尺度对地下水中砷的分布和风险做出较为全面的评价是当前面临的重要挑战。文中旨在利用有限的地下水样本,查明高砷地下水分布区的水化学特征,结合区域地理空间要素,构建随机森林模型对有限的数据进行充分挖掘,预测区域高砷地下水风险概率,探寻高砷地下水分布的主控地理要素。研究结果表明,银川平原含水层呈现弱碱性的还原环境,阳离子以Na+为主,阴离子以HCO-3为主,承压水含水层水质状况受外部环境的影响较弱,与潜水层相比,各水化学指标和常量离子的浓度值较低,水质类型简单,但二者具有类似的水化学分布规律。从全区来看,各采样点不同砷浓度对应地下水的水化学类型有所差异,高砷地下水(As≥0.01mg/L)对应的水化学类型主要以HCO_(3)-Na·Ca型和HCO_(3)·SO_(4)-Ca·Mg·Na型为主。随机森林模型预测结果表明,地下水砷风险整体偏高,高砷地下水主要集中分布于平原北部,呈现由北向南逐渐减少趋势,平原南部高砷地下水呈零散分布状,潜在的高砷危害区面积达到1236.69km^(2),占研究区总面积的15.33%。高砷地下水分布主要受地下水埋深、蒸散发、植被、坡度、地面高程、降水、温度等因素的影响。The Yinchuan Plain is one of the areas in China where drinking-water type of endemic arsenism is prevalent.Limmited by the difficulty in obtaining samples,conducting a comprehensive assessment of the distribution and risk of arsenic in groundwater at the regional scale remains a significant challenge currently.This study aims to identify the hydrochemical characteristics of areas with high arsenic groundwater by using a limited number of groundwater sample,combining regional geographical spatial factors,construct a Random Forest model to fully explore the limited data,predict the risk probability of high arsenic groundwater in the region,and explore the main controlling geographical factors for the distribution of high arsenic groundwater.The results indicate that the aquifers in the Yinchuan Plain exhibit a weakly alkaline reducing environment,with cations mainly dominated by Na~+and anions by HCO~-_3.The quality of water in the confined aquifers is less affected by the external environment,and compared with the unconfined aquifers,the contents of ions and chemical indicators are lower,the water quality type is simpler,but both have similar hydrochemical distribution patterns.Looking at the entire region,the hydrochemical types corresponding to different arsenic concentrations at various sampling points vary.The hydrochemical types corresponding to high arsenic groundwater(As≥0.01mg/L)are mainly HCO_(3)-Na·Ca type and HCO_(3)·SO_(4)-Ca·Mg·Na type.The prediction results of the random forest model indicate that the risk of groundwater arsenic is generally high,with high arsenic groundwater mainly concentrated in the northern part of the plain,showing a trend of gradually decreasing from north to south.In the southern part of the plain,high arsenic groundwater is scattered,and the potential high arsenic hazard areas cover an area of 1236.69km^(2),accounting for 15.33%of the total area of the study region.The distribution of high arsenic groundwater is mainly influenced by factors such as groundwater depth,eva

关 键 词：银川平原 高砷地下水 随机森林 地理空间要素 

分 类 号：P641[天文地球—地质矿产勘探]