摘要
长期摄入高碘地下水(碘浓度>100μg/L)会造成人体甲状腺机能损伤,掌握区域高碘地下水的空间分布规律对于有效规避劣质地下水,保障地下水资源的可持续安全供给至关重要。但大规模地下水水质调查耗费大量的人力、财力、物力。基于江汉平原177组常规的浅层地下水水质调查数据,选取DOC、HCO^(-)_(3)、Mg^(2+)、Fe^(2+)、NH^(+)_(4)-N、SO_(4)^(2-)等水质参数作为预测变量,建立江汉平原高碘地下水风险极端梯度提升机器学习预测模型,用于预测研究区高碘地下水的空间分布。结果表明:该模型通过测试数据集检验,预测的准确率达到86.4%;模型预测结果显示,江汉平原高碘地下水主要分布在长江河曲沿岸,零星分布在平原腹地河湖区,并识别出江汉平原西北部丘陵前缘的汉江沿岸也是高碘地下水分布的潜在区域。该研究结果将有助于圈划高碘地下水的空间分布范围,可为确定未来地下水水质监测的优先区域提供科学指导。
Long-term intake of geogenic iodine-contaminated groundwater(iodine concentration >100 μg/L) can cause damage to human thyroid function,and it is crucial to identify the spatial distribution of regional high iodine groundwater to effectively avoid geogenic-contaminated groundwater and ensure sustainable and safe supply of groundwater resources.However,the large-scale groundwater quality surveys are time consuming and laborious.In this study,an extreme gradient boost(XGBoost) model is developed to predict the spatial distribution of high iodine groundwater based on conventional groundwater quality survey data(n=177) in Jianghan Plain,with DOC,HCO^(-)_(3),Mg^(2+),Fe^(2+),NH^(+)_(4)-N and SO_(4)^(2-) as predictor variables.The model correctly predicted 86.4% of the available groundwater iodine observations in the study area by test-dataset.According to the prediction of XGBoost,high iodine groundwater in Jianghan Plain is mainly distributed in the oxbow lake of the Yangtze River and scattered in the hinterland of rivers and lakes,and potential areas of high iodine groundwater in the northwestern hilly foreland of Jianghan Plain are identified.The study is helpful to delineate the spatial distribution of high iodine groundwater and to identify priority areas for groundwater quality monitoring.
作者
范瑞宇
邓娅敏
薛江凯
FAN Ruiyu;DENG Yamin;XUE Jiangkai(School of Environmental Studies,China University of Geosciences(Wuhan),Wuhan 430078,China;Institute of Geological Survey,China University of Geosciences(Wuhan),Wuhan 430074,China)
出处
《安全与环境工程》
CAS
CSCD
北大核心
2022年第5期70-77,共8页
Safety and Environmental Engineering
基金
国家自然科学基金面上项目(41977174)。
关键词
高碘地下水
极端梯度提升模型
机器学习
江汉平原
geogenic iodine-contaminated groundwater
extreme gradient boost(XGBoost)model
machine learning
Jianghan Plain
