武汉南湖水华生消关键因子识别及临界值

Identification of Key Factors and Critical Values for Production and Elimination of Water Bloom in Wuhan South Lake

识别水华生消关键影响因子及其临界值对湖泊水华防控具有重要意义。以武汉市南湖为研究区,基于水文、水质和气象监测数据,采用皮尔逊相关分析、灰色关联分析、主成分分析方法,筛选并确定与叶绿素a浓度相关性较大的关键环境因子,利用受试者工作特征(ROC)曲线法分析关键环境影响因子的临界值。结果表明,COD_(Mn)、总磷、总氮、水温、气温与叶绿素a浓度具有较强的相关关系,是水华暴发的关键因子,其临界值分别为6.25 mg/L、0.146 mg/L、0.725 mg/L、27.05℃、22.35℃。研究发现,ROC曲线法对COD_(Mn)、总磷、总氮等日变幅较小环境因子临界值的求解精度优于气温、水温等日变幅较大环境因子,在湖泊水华暴发的营养盐浓度临界值确定方面具有较大优势。

The identification of key influencing factors and their critical values for the prevention and control of water bloom in lakes is of great significance.Taking Wuhan South Lake as the study area,based on hydrological,water quality and meteorological monitoring data,Pearson Correlation Analysis,Gray Correlation Analysis and Principal Component Analysis were used to screen and identify key environmental factors with high correlation with chlorophyll a tion,concentra-and the critical values of key environmental impact factors were analyzed by the Receiver Operating Characteristic(ROC)curve method.The results show that the CODMn,total phosphorus,total nitrogen,water temperature and air temperature had strong correlations with chlorophyll a concentration and were the key factors for water bloom outbreak,with the critical values of 6.25 mg/L,0.146 mg/L,0.725 mg/L,27.05℃and 22.35℃,respectively.It was found that the accuracy of the ROC curve method for solving the critical values of CODMn,total phosphorus,total nitrogen and other environmental factors with small daily variation was better than that of environmental factors with large daily tion varia-such as air temperature and water temperature,and the ROC curve method has a greater advantage in determining the critical values of nutrient concentrations for water blooms in lakes.