Climate change has significantly increased extreme precipitation and altered regional hydrological cycle, aggravating flood in the watershed. The effective measurement of the risk brought by climate change is an effective way to cope with flood hazard in the future. At the same time, the quality of the simulation of climate change scenarios will also affect the accuracy of flood risk assessment. Therefore, a comprehensive method is needed to measure the long-term disaster risk. However, the current method of subjectively assigning indicator weights is still subjective and difficult to be promoted and applied. So a new model for assessing watershed risk is constructed in this study. Based on the game cross-efficiency data envelopment analysis method and the combination of simulations of climate scenarios, the model can determine the input factors of the assessment and the influencing level of the input factors by using the Principal Component Analysis and Tobit model. The model comprehensively evaluates the flood risk level in the watershed with the results of the simulation of hazard in different climate scenarios, hazard exposure and social vulnerability as input factors, and the degree of disaster loss as the output factor. Results: (1) the hazard, exposure, and social vulnerability are spatially mismatched; (2) the overall risk in the watershed presents such a pattern: upstream (0.751) > downstream (0.418) > midstream (0.362); (3) the long-term flood hazard may be reduced under the influence of climate change. The research is helpful to formulate long-term flood mitigation strategies in the future.

气候变化大大增加了极端降水，改变了区域水文循环，加剧了流域的洪灾。有效测量气候变化带来的风险是将来应对洪水灾害的有效方法。同时，气候变化情景模拟的质量也将影响洪水风险评估的准确性。因此，需要一种综合的方法来衡量长期灾难风险。但是，当前主观分配指标权重的方法仍然是主观的，难以推广和应用。因此，本研究构建了一种评估流域风险的新模型。基于游戏交叉效率数据包络分析方法和气候情景模拟的结合，该模型可以使用主成分分析和Tobit模型确定评估的输入因子和输入因子的影响程度。该模型以不同气候情景下的灾害模拟结果，灾害暴露和社会脆弱性作为输入因子以及灾害损失程度作为输出因子的结果，对流域的洪水风险水平进行了综合评估。结果：（1）危害，暴露和社会脆弱性在空间上不匹配；（2）流域的总体风险呈现出这样一种模式：上游（0.751）>下游（0.418）>中游（0.362）；（3）在气候变化的影响下，可以减少长期的洪水灾害。该研究有助于将来制定长期的防洪策略。