年度洪涝灾害风险评估模型及其应用——以湖南为例

Annual Flood Disaster Risk Assessment Model and its Application

依据灾害概率风险评估理论,考虑汛期降水的可预报性,基于历史洪涝灾害事件,尝试性地构建了年度洪涝灾害风险评估模型,以湖南为例,对评估模型进行了检验与应用。结果表明:模型建立了前期海温和环流指数等因子与汛期区域降水时空分布的回归预报方程,应用概率风险分析得到年度不同降水下直接经济损失分布,结合蒙特卡洛仿真模拟求解损失的超越概率曲线,评估年度洪涝灾害单次最大可能损失、年度总损失以及年期望损失;模型集"未来年度汛期降水预测"、"降水与损失分布关系拟合"、"损失超越概率评估"于一体,是对全过程年度洪涝灾害风险评估方法的新探索;湖南年度洪涝灾害评估案例表明该模型可操作,结果与实际情况相符。研究可为完善灾害风险评估内容与技术方法提供新视角,亦可为开展业务实践提供方法借鉴。

Based on the theory of disaster probabilistic risk assessment, considering the predictability of precipitation in flood season and the historical flood events databases, the annual flood disaster risk assessment (AFDRA) model is constructed. Taking Hunan as an example, the AFDRA model is tested and applied. The results show that:(1) the model establishes the regression forecasting equation from the temporal and spatial distribution of regional precipitation and sea surface temperature (SST), atmospheric circulation index factors; to delineate the direct economic loss distribution under different annual precipitation scenarios by the support of probabilistic risk analysis; the biggest annual flood disaster loss, the total loss of the year and the expected annual loss are evaluated based on Monte Carlo simulation.(2) The AFDRA model is a new exploration of the all-process of the annual flood disaster risk assessment method.(3) The results from the Hunan case show that the AFDRA model is operable and are consistent with the actual situation. This study can provide a new perspective for improving the content and technical methodology of disaster risk assessment, and can also provide a reference for the practice of business practice for annual flood disaster risk assessment.