近１０ａ来，用统计方法分析风险取得了较大的进步。风险管理理论可以定义为事故识别和事故造成的可能破坏以及应采取的减轻和防护措施。最值得关注的是这样一个事实，即洪水风险概率的量化研究本质上是个实践问题，已成为深奥的统计分析研究课题，并脱离了工程设计和水库调度的需要。皮尔格林早在１９８６年提出：现行的研究倾向值得质疑，研究学者和工程界需要更广泛的联系。 Nearly 10 years, with statistical methods to analyze the risk has made great progress. Risk management theory can be defined as accident identification and possible damage caused by an accident and mitigation and protection measures to be taken. What deserves the most attention is the fact that the quantitative study of the probability of flood risk is a matter of practice in nature and has become an esoteric subject of statistical analysis and research, and has deviated from the needs of engineering design and reservoir dispatching. Pilgrim put forward as early as 1986: the current research tends to question the need for more extensive connections between researchers and engineering.