中文摘要：

以中国东北地区典型工业基地沈阳浑河冲洪积扇为研究对象,采集区域内43个采样点水质样品,分析了16项地下水质关键指标,在水化学统计和水质分析确定水质等级及特征污染物的基础上,利用主成分分析（PCA）方法确定水质指标主要因子分类,运用Arc GIS软件刻画不同污染源的空间分布规律,利用绝对主成分得分多元线性回归受体模型（APCS-MLR）计算不同公因子对地下水质的贡献程度,并验证污染源解析结果的准确性.结果表明：研究区内地下水氮、磷和铁等元素超标严重,地下水水质演化受人类活动影响较大;4类主要污染因子分别是溶滤迁移作用因子（贡献率34.21%）、农业活动污染因子（贡献率20.13%）、地质环境背景因子（贡献率13.39%）和工业活动影响因子（贡献率8.97%）,方差累计贡献率为75.64%;工业生产、生活污废水排放和农业生产中农药化肥使用是地下水主要污染源且多分布于沈阳西部、西北部和南部等地区,淋溶、迁移-富集作用和农业污染对地下水质影响较为明显,PCA-APCS-MLR模型预测浓度与实测浓度一致,该方法对于地下水污染源的计算分配具有良好的针对性,适用于地下水污染源解析.

英文摘要：

Totally 43 groundwater samples were sampled in the Hunhe River alluvial fan, which was a typical industrial area in Shenyang at northeastern China and 16 key groundwater components were analyzed. The main factor of groundwater quality was determined by principal component analysis（PCA） based on water quality grade and characteristics pollutants using water chemistry statistics analysis, and the spatial distribution of different pollution sources was described by Arc GIS software. The contribution of different principal factor to groundwater quality was calculated by absolute principal component score multiple linear regression model（APCS-MLR）, and verified accuracyof pollution sources apportionment. Results showed that nitrogen, phosphorus and iron exceeded to groundwater quality standardsignificantly, the evolution of groundwater quality was mainly influenced by human activities. Four main pollution factors were： leaching migration with contribution of 34.21%, agricultural pollution with contribution of 20.13%, geological background with contribution of 13.39% and industrial activities with contribution of 8.97%. The contribution of cumulative variance was 75.64%. The industrial production, domestic sewage discharging and agriculture fertilizer pollution were main pollution sources of groundwater, which were distributed in northwestern and southwestern Shenyang. The pollution of leaching and migration and agriculture pollution affected on the groundwater quality significantly, and the predicted results was consistent with the measured concentration, which indicated that the PCA-APCS-MLR model wasof good pertinence for the distribution of pollution sources and it was suitable for source apportionment for groundwater.