中文摘要：

山区降雨预报的不确定性强,实际降雨信息难以实时测报,使得准确及时的洪水预警难以实现。从群测群防角度看,山区居民可粗略测量各自位置上的实时降雨量,但不了解上游流域内的降雨情况,也不掌握对洪水进行可靠估计的方法。基于智能手机、云计算等信息技术,提出并初步实现了一个用于资料缺乏山区洪水预警系统的框架。该系统综合了群测群防思想和水文模拟技术,在降雨发生后通过居民的智能手机终端收集地理坐标和雨量估计值,驱动云端的数字流域模型自动运行,根据径流模拟结果推定不同地理位置的预警级别,并将预警信息推送到雨量信息提供者和潜在受灾者,使用户在洪水到达前接收到预警信息。该系统的应用将提高群测群防的技术水平。

英文摘要：

The rainfall forecast in mountainous region generally has high uncertainty, and the measurement and real-time transfer of actu- 8.1 rainfall data is difficult and expensive. Based on information technologies, including smartphones and cloud computing, the framework of an interactive flood warning system was proposed and tested for ungauged mountainous regions according to the mass prediction and prevention principle. Firstly, geographic coordinates and rainfall estimate inputs from local residents were collected via smartphone terminal. Then the parallel hydrological model on the cloud server was automatically executed to simulate the rainfall-runoff process. Finally, according to the allowed flood discharge at different locations, warning levels were derived from runoff simulation results and pushed to the rainfall information provider and potential flood sufferers via smartphones. Because of the high efficiency of data collection and hydrological simulation, residents can receive warning information before the flood arrives, and save time to escape. The application of this system would essentially improve the technical level of mass prediction and prevention.