宁夏地下水原生氟污染空间分布预测

Spatial Distribution Prediction of Primary Fluoride Pollution in Groundwater in Ningxia Hui Autonomous Region

地方性氟中毒在中国几乎所有省份都存在,长期摄入含高氟的地下水是氟中毒的主要原因之一。宁夏地表水资源匮乏,地下水是其重要的供水来源,而地下水氟污染对供水安全造成了潜在危害。为明确宁夏地下水氟污染的分布状况,收集整理了宁夏333个地下水中氟化物浓度样本数据以及地质、气候、土壤等相关空间变量数据,利用人工神经网络方法对宁夏地下水中氟化物浓度与对应的预测变量进行建模,并利用建立的高氟地下水人工神经网络预测模型生成了分辨率为0.5弧分的宁夏高氟地下水风险空间分布预测图,并对预测变量重要性进行了分析。结果表明:建立的高氟地下水人工神经网络预测模型在测试集上的准确率、敏感性和特异性分别为86.8%、85.9%和88.1%;模型预测得到的高氟地下水高风险区主要分布于宁夏中部和东部地区;气候变量是高氟地下水最重要的预测变量。该研究结果有助于宁夏改水降氟工程的实施,对促进宁夏生态环境可持续发展具有重要的指导作用。

Endemic fluorosis exists in almost all provinces of China.The long-term ingestion of high fluoride groundwater(fluoride>1.5 mg/L)is one of the main causes of fluorosis.Groundwater is an important source of water supply in Ningxia Hui Autonomous Region due to the lack of surface water resources.However,groundwater fluoride contamination may cause potential harm to water supply security.In order to clarify the distribution of groundwater fluoride contamination in Ningxia Hui Autonomous Region,333 groundwater fluoride concentration samples and available spatial continuous variables such as geological,climatic and soil parameters as proxy predictors are collected,and artificial neural network is then used to model the relationship between groundwater fluoride concentrations and proxy predictors.The results are as follows:the accuracy,sensitivity,and specificity of the developed artificial neural network prediction(ANN)model for high fluoride groundwater are 86.8%,85.9%,and 88.1%,respectively;climate variables are the most important predictors for high fluoride groundwater.The developed ANN model for high fluo-ride groundwater is used to generate a prediction map of high fluoride groundwater in Ningxia Hui Autonomous Region with a resolution of 0.5 arc minutes,and the importance analysis of the predictor variables is performed.The predicted high risk areas of high fluoride groundwater are mainly located in the central and eastern Ningxia Hui Autonomous Region.The results of this study are useful for guiding the implementation of water improvement and defluoridation projects and promoting the sustainable development of Ningxia Hui Autonomous Region.