苏州古城区水体污染时空分异特征及污染源解析

利用2012年苏州古城区30个监测断面的11个水质指标数据,综合运用水质指数模型(WQI)、层次聚类(HCA)、后退式判别(DA)、因子分析(FA)、绝对主成分多元线性回归(APCS-MLR)及GIS平台系统分析了旅游城市苏州古城区河网水体污染物时空分异特征及污染源解析.结果表明:①苏州古城区内城河及外围河道水体的CCME WQI值介于40~74之间,其中,66.67%的监测点水环境遭到严重破坏,主要集中在内城河河道;②系统聚类分析将采样时间分为1—3月、11月及4—6月、7—10月3个时段;将采样点分为2类,从空间上反映了古城区内城河与外围河道的污染程度;③采样时间和采样点聚类分析结果的判别分析交叉验证,正确率分别达到88.1%和78.5%;表征时间差异性用了7个指标,分别为总氮(TN)、总磷(TP)、溶解氧(DO)、水温(T)、高锰酸盐指数(COD Mn)、藻密度(Algae density)、叶绿素(Chl),表征空间差异性用了5个指标,分别为总氮(TN)、氨氮(NH3-N)、溶解氧(DO)、浊度(Turb)、水温(T);显著性指标的时空差异性比较明显;④古城区内城河河道在1—3月、11月及4—6月、7—10月3个时段内的因子分析分别提取4、3和4个因子,累积解释方差分别为83.64%、72.67%和77.98%;古城区外围河道在全年内的因子分析提取了3个因子,累积解释了62.28%的总方差;因子分析表明,古城区内城河及外围河道主要为氮、磷等营养物质污染及夏季藻类爆发的问题,与古城区外围河道相比内城河河道的污染更为严重,应优先开展治理;内城河氮、磷污染除了有来自生活、餐饮旅游等第三产业的污染外,还受到降雨地表径流及河道底泥释放的非点源污染影响;⑤绝对主成分多元线性回归(APCS-MLR)表明,总氮(TN)、氨氮(NH3-N)、总磷(TP)和高锰酸盐指数(COD Mn)主要来自城市生活及餐饮旅游等第三产业污水.研究结果可为苏州古城区河道水环境改善治理提供参考.

Spatio-temporal characteristics and source identification of water pollutants in ancient town of Suzhou

Characterization of the spatio-temporal patterns and apportionment of the pollution sources of water bodies are important for the management and protection of water resources. In this work, water-quality index modeling, Hierarchical cluster Analysis (HCA), backward Discriminant Analysis (DA), Factor Analysis (FA) and APCS-MLR were used to analyze data sets of river network water quality for 11 parameters monitored at 30 different sites in ancient town of Suzhou during 2012. With the help of GIS platform, the spatiotemporal characteristics and source identification of water pollutants were studied. The results showed that: 1 The value of WQI is between 40~74 around the river network in ancient town of Suzhou, with water quality of approximately 70% of monitoring site has been seriously impaired; 2The sampling periods were categorized into three clusters (January to March and November, April to June, and July to October). Sampling sites were classified into two groups which represented different water quality levels and pollution degrees; 3Discriminant analysis gave the best results for both spatial and temporal analysis. It provided an important data reduction as it uses only seven parameters (TN, TP, DO, T, COD_Mn, Algae density, Chl), affording more than 88.1% correct assignations in temporal analysis, and five parameters(TN, NH₃-N, DO, Turb,T), affording more than 78.5% correct assignations in spatial analysis; 4Factor analysis applied to the data sets of inner river (Cluster A) in the three periods resulted in four, three and four latent factors explaining 83.64%,72.67% and 77.98% of the total variance, respectively. Factor analysis applied to the data sets of peripheral channel (Cluster B) in the whole year resulted in three latent factors explaining 62.28% of the total variance. The varifactors obtained from factor analysis indicate that the main contaminations of inner river and peripheral channel are nitrogen and phosphorus pollutants. Compared with the peripheral channel, the inner river's pollution is more serious, therefore should be controlled with priority; Meanwhile, the sources of the main contaminations is not only from domestic wastewater but also from rainfall runoff and river sediment release; 5The results of APCS-MLR show TN, ammonia (NH₃-N), TP and COD_Mn mainly from domestic sewage and tertiary industry wastewater. These results provide indications in developing optimal strategies and determining priorities for river pollution control and effective water resources management.