蓝藻水华强度的显著相关环境因素识别模型

Model for Identifying Environmental Factors Significantly Associated With Cyanobacterial Bloom Intensity

为识别蓝藻水华强度的显著相关环境因素,克服现有研究中因变量选择不合理、时间与空间精度较低等问题,构建了以蓝藻水华强度等级为因变量,以水质、水文和气象3类监测指标为自变量的多元线性回归模型,并将该模型应用于太湖蓝藻水华研究。基于水华面积和集聚强度数据,用7级量表生成水华强度等级值,使因变量具有宏观性,避免了仅使用叶绿素a浓度等类似指标表示水华强度所体现出的微观性不足。该数据集的时间精度达到每天采样2次,空间精度则达到太湖湖湾内的某个水域空间范围。因此因变量具有适度宏观性,而自变量的值则与因变量的值在较高的时间和空间精度基础上严格对应。模型的分析结果显示,太湖大贡山水域蓝藻水华强度与气温和硝酸盐浓度呈显著正相关,与风速、湿度和电导率呈显著负相关。上述结论与该研究领域的主流结论一致,验证了该模型的有效性。

To identify environmental factors significantly associated with cyanobacterial bloom intensity and to overcome the problems,such as unreasonable selection of dependent variables,low spatio-temporal precision in the present researches,a multi-factor linear regression model was developed for the study on cyanobacterial bloom in the Taihu Lake,using cyanobacterial bloom intensity scale as dependent variable,and monitoring indices of water quality,hydrology,and weather as independent variables.Based on the data of covering areas and aggregation intensities of the blooms,a 7-level scale was designed to indicate bloom intensity,thus making the dependent variable macroscopic and avoiding the problem of lacking microscopicity in using only analogous indices,like chlorophyll a concentration,to represent bloom intensity.The temporal precision of this data set reached as high as twice a day for sampling,and the spatial precision as high as the spatial scale of a certain patch of water in a bay of the Taihu Lake.Thus,the dependent variable becomes somewhat macroscopic,while the values of independent variables correspond strictly to the value of the dependent variable on the basis of a high spatial-temporal precision.This model was applied to the study on cyanobacterial bloom in the Taihu Lake of China.Analysis using the model reveals that the intensity grade of cyanobacterial bloom in the Dagongshan water area of the Taihu Lake was significantly and positively related to air temperature and nitrate concentration in the water and negatively to wind speed,humidity,and electrical conductivity.The above conclusions conform to the mainstream conclusions in this research field,which manifests the effectiveness of this identifications model.