| The flash flood disaster is one of the most severe terrestrial natural disasters in the world. It often happens occasionally but quickly, so it is difficult to make accurate predictions. A lot of lives were lost and properties were damaged in flash flood disasters in the populated mountain area year after year. With the changes of global climate and the intensifying of human action, flash flood disaster has affected the development of economy and society greatly and become severe threaten to human beings and therefore, it’s urgent to find effective approaches to forecast the flash flood disaster. The traditional way to judge the occurrence probability of flash flood is based on the warning indexes like critical rainfall value in the pre-classified risk area. However the approach is lack of comprehensive scientific basis. Recently, to use the distributed hydrological model based on the technologies of geographical information system (GIS) and remote sensing becomes a main warning way to predict flash flood disaster. Considering the factors of data availability in the flash flood disaster-prone area and easy promotion of warning technology, simple but reliable distributed models based on physical principles need to be advocated. In addition, it has increasingly been focused that how to make full use of the information technology, especially the spatial information technology, to spread the warning information to related personnel quickly. Therefore, the research proposed the warning method of flash flood disaster based on scenarios modeling on the basis of summary of the existed warning ways of flash flood disaster. The hydrological model-SCS-CN (Soil Conservation Service-Curve Number) and travel time method were employed in surface run-off modeling and the inundated area was computed using the source submerged algorithm in the different rainfall scenarios. Taking Maolinxi watershed in Tongan, Xiamen, Fujian Province as the study area, in one rainfall event of on August14,2005, the different rainfall scenarios were designed as the input data to analyze the flash flood disaster risk of different scenarios. At last, the warning system of flash flood disaster based on the method was developed using oriented-object programming language and visual programming technology to achieve the transformation from the scenarios of meteorological prediction to the specific scenarios of flash flood disaster risk in the study area. The approach takes advantages from satellite remote sensing and GIS. It requires very little hydrological data and its performance is fast and efficient. Therefore, the approach presents great potential to early warning and forecasting of flash flood disaster for medium-or small-size watersheds in China at present. |
PDF Full Text Request