在研究水华风险过程中，综合考虑内源和外源两个层面构建水华风险评价指标体系，针对水质评价及影响 因子具有模糊不确定性，利用可变模糊集理论构建一种湖库型水源地水华风险可变模糊评价模型。该模型借助 相对差异函数刻画水华风险的模糊性，通过变化参数组合，克服单一模型的局限性。将该模型应用于乔店水库水 华风险评价。结果表明，乔店水库水华风险级别特征值为 3.19，呈Ⅲ级风险；其中由外源因素和内源因素引起的 风险等级分别为Ⅱ级风险和Ⅳ级风险，表明引发水华的内源风险不可忽视。该评价结果与实地调研结论相吻合， 进一步验证该评价模型的可靠性，弥补仅考虑外源风险的不足。该指标体系和模型可推广应用于其他类似湖库， 对精准防治水华提供理论支撑。

In recent years, water bloom occurrence frequency in water sources is increasing gradually. The excessive proliferation of algae not only intensifies the accumulation of toxins in the water body but also leads to a series of chain reactions such as water pollution, which seriously affects human life and health through the food chain. In the long run, the ecosystem will be seriously damaged, attracting the attention of scholars at home and abroad. Therefore, the study of the degree of bloom risk has an important impact on the management and the improvement of water quality.The risk of water bloom in water source areas is affected by many factors, such as internal and external factors. A relatively comprehensive evaluation index system of water bloom risk in water source areas had been constructed from the internal and external levels, which could overcome the limitations caused by only considering external risks. Given the fuzzy characteristics and uncertainties of water quality assessment and impact factors, a variable fuzzy assessment model of water bloom risk in water source areas was constructed by the variable fuzzy set theory. The model could describe the fuzziness of water bloom risk with the help of a relative difference function. Through the transformation of different parameter combinations, the limitation of only a single model was overcome, and the risk of water bloom in lake reservoir-type water source areas was difficult to be quantitatively described because the coupling of multiple factors was solved.The model was applied to the risk assessment of water bloom in Qiaodian Reservoir. The results showed that the characteristic value of the water bloom risk level was 3.19 (grade III risk). The characteristic value of risk level caused by external factors was 2.41 (grade II risk). Among them, the amount of fertilizer per unit area, the impact of upstream water on water bloom and the duration of light were the key indicators that affected the level of exogenous risk. The characteristic value of risk level caused by endogenous factors was 3.54 (grade IV risk), which was closely related to the sedimentation, water temperature and hydraulic retention time of Qiaodian Reservoir.The constructed variable fuzzy evaluation model formed 4 different evaluation models by changing the combination of parameters, which further reflected the nonlinearity of the water bloom risk evaluation system. The multi-dimensional evaluation index system of water bloom risk in the water source area was comprehensively evaluated, which improved the reliability and reliability of the evaluation results. The application results of the example showed that it was consistent with the research conclusion, indicating that the endogenous risk of causing water bloom could not be ignored. The result of the single factor risk grade further showed that the risk fluctuation of water bloom was comprehensively affected by internal and external risks. Therefore, the index system and model can be applied to other similar lake reservoir-type water source areas to provide scientific theoretical support for precise prevention and control of water bloom. In the future, the water source managers should adjust the control direction by giving priority to the treatment of internal pollution, strengthening the control of non-point source pollution, then adjusting the hydraulic cycle, and improving the self-purification capacity of water bodies.