蓝藻水华显著影响因子识别模型

Identification Model of Significant Impact Factors for Cyanobacterial Bloom

根据蓝藻水华反演图判断蓝藻水华发生与否,进而构建了直接以蓝藻水华发生与否的指标为因变量,同时以水质、水文和气象3类共19个指标为自变量的蓝藻水华环境影响因子识别模型(Probit模型),并将该模型应用到太湖.经模型分析,识别出太湖蓝藻水华暴发的显著环境影响因子:气温和硝酸盐浓度与蓝藻水华暴发的概率正相关;风速、湿度和电导率则与水华暴发的概率负相关.将离散因变量模型(二元选择模型)引入蓝藻水华研究领域,对湖泊(水库)蓝藻水华暴发的预测和预防具有参考价值.

A probit model was constructed for identifying significant environmental abiotic impact factors for cyanobacterial bloom.In the model,the variable of occurrence-or-not of cyanobacterial bloom,which is the result of judgement on inverting pictures for cyanobacterial bloom,was used as the dependent variable,and the indices of water quality,hydrology and weather,altogether,19 indices in 3 categories,as independent variables.The model was applied to the inspection of Lake Tai in China.Through the analysis of probit model,the significant environmental impact factors for bloom outbreak in Lake Tai were systematically identified on the basis of historical data.Air temperature and nitrate concentration have positive correlation with the probability of cyanobacterial bloom outbreak,while wind speed,humidity and conductivity have negative correlation with the probability.They are the significant impact factors for bloom outbreak.The model with discrete dependent variables（binary choice economic analysis model） was introduced successfully in the research field of cyanobacterial bloom.The results are of referencial value for prediction and prevention of cyanobacterial bloom outbreak in lakes and reservoirs.