耦合多源信息的山洪灾害预报预警方法研究

The torrential flood disaster has the characteristics of the complex mechanism, numerous influencing factors, strong abruptness and huge destructive effect. Carrying out the forecasting and warning of torrential flood disaster is an important safeguard to public safety and economic and social development. The comprehensive study based on coupling multi-source information is an innovative research field of disaster prevention and mitigation. This study mainly research the method of forecasting and warning of torrential flood disaster by coupling multi-source information based on the formation, causes and mechanism of torrential flood in Liaoning Province. The research contents includes: . (1) The frequency and scale of torrential-flood disaster are analyzed combining with the characteristics of torrential flood disaster, in order to reveal the space-time coupling effect and establish a composite Indication System of inducing factor and susceptibility. . (2) Based on the complete information in the study area, formation and mechanism of torrential flood are analyzed combining with comprehensive Indication System. A model of early warning of torrential flood disaster is established to conduct simulation and evaluatefeasibility combining with the history data and actual situation.. (3) The database of the spatial and characteristic about the torrential flood disaster is established. And the DEA based on GIS is used to evaluate the risk of torrential flood disaster. Combining with the evaluation results, the risk of historical disaster data is regionalized, and then the risk zoning is plotted and validated based on historic disaster data.. (4) Some feasible suggestions of prevention of torrential flood disaster are proposed based on the engineering and non-engineering measures.

山洪灾害具有形成机理复杂、影响因素多、突发性强和破坏性大等特点，开展山洪灾害预报预警是公共安全和经济社会发展的重要保障，而基于多源信息耦合的山洪预报预警方法研究是防灾减灾的创新性研究领域。本基金以辽宁省山洪形成过程、成因和机理为基础，耦合多源信息研究山洪预报预警方法。研究内容包括:.（1）结合山洪灾害特性统计山洪灾害发生的频率和规模，揭示山洪灾害发生的时空耦合效应，建立描述诱发因子及易发性的综合指标体系。.（2）依据研究区的资料完整情况，应用综合指标体系，分析山洪形成过程及作用机理；建立区域性山洪灾害预报预警模型，结合历史资料和山洪实况进行模拟和可行性评价。.（3）根据已建立的山洪灾害空间和属性数据库，借助GIS工具建立DEA模型进行山洪灾害风险性评价，结合历史灾情数据进行风险区划，绘制风险区划图并加以验证。.（4）结合工程措施和非工程措施，对山洪灾害防治提出可行性的对策建议。

我国作为世界上受山洪灾害影响最为严重的国家之一，近年来受人类活动和气候变化的影响，突发性、局地性山洪灾害频发，致使大量人员伤亡，财产损失严重。本基金以辽宁省山洪形成过程、成因和机理为基础，耦合多源信息研究山洪灾害预报预警方法。利用近30年辽宁省山洪灾害统计数据与灾情信息，通过统计分析与现场调研综合方法，分析辽宁省及本溪县历史山洪灾害特征情况，同时采用ArcGIS软件、周期分析等方法对辽宁省泥石流、滑坡灾害发生的时空特征进行研究；根据多年洪灾统计数据分析辽宁省及本溪县山洪灾害损失特征，探究辽宁省及本溪县山洪灾害频率与规模的相关性；通过聚类分析、专家打分等方法，对山洪灾害发生的孕灾环境进行分析，结果表明山洪灾害的各项参数均呈现出正态分布，同时地形和暴雨对山洪灾害的影响更大；结合山洪暴发可能性、山洪活动属性、综合考虑各致灾因子的影响程度，对山洪灾害易发区进行等级划分，同时得到山洪灾害时空耦合效应与临界条件，为做好山洪致灾临界条件和预警做好技术铺垫；叠加溪河洪水、滑坡和泥石流三种灾害历史发生地点，绘制出辽宁省及本溪县山洪灾害风险区划图，以期为山洪监测和治理提供参考依据；采用HEC-HMS水文模型对本溪县进行了水文预报模拟应用，绘制出HEC-HMS水文模型洪水模拟过程变化图。研究成果可为区域山洪灾害预报预警技术发展提供重要依据。