中文摘要：

以一座典型的供水防洪城市大型水库为研究对象,采用决策树方法,构建其水华发生预警及管理模型。结果显示,该水库水体叶绿素浓度水平受多重因素的影响,其中,总磷、氮磷比及调水量对水库水体叶绿素水平的影响最大,其次为水温和溶解氧。叶绿素浓度与氮磷比值之间存在显著的正相关关系（R~2=0.25）。水库水体的TN∶TP多年平均值为25∶1,该水库为磷限制。决策树分析结果表明,该水库水华暴发的重要因素为来水量、氮磷比和总磷浓度。防止研究区水华暴发的最有效措施是提高水库的来水量,当每月来水量高于7 776万m~3时,就能在一定程度上避免该水库暴发水华灾害。此外,该水库水华暴发时氮磷比一般高于30,通过控制水体总磷的浓度和氮磷比,可降低该水库水华暴发的风险。

英文摘要：

The decision tree method was used to construct an algal bloom early warning and management model based on the data of a typical urban large reservoir. The result shows that chlorophyll concentrations affected by multiple factors in the reservoir, and key factors included TP concentrations,ratio of TN ∶TP and water diversion. Water temperature and DO concentration also played important roles on the variation of chlorophyll. There is significant positive relationship（ R~2=0.25）between chlorophyll concentration（Chl） and the ratio of nitrogen and phosphorous（N∶P）. The reservoir was phosphorus limited and average value of the ratio of N∶P was 25∶1. The water quantity（Q）, ratio of N∶P and total phosphorous concentrations are the key factors of algae blooms in the reservoir based on the decision tree analysis. The outbreak of algal blooms can be avoided when Q 77.76 million m~3 monthly of the study reservoir. On the other hand, algal blooms can be significantly reduced by controlling the concentration of TP as the algal blooms generally appeared when ratio of N ∶P higher than 30.