Effect Of Spatial-temporal Scale On Flood Simulation

A concept distributed hydrologic model is built in the dissertation, based on which the effect of DEM resolution and spatial-temporal distribution of rainfall on runoff simulation is discussed in three sub-basin: Jiangji, Huangnizhuang and Qilin of Shiguan river basin.The object of spatial analysis is to distract and transport the spatial information. Analyzing hydrologic data and DEM data is the foundation of application of distributed hydrologic model. This paper deals with hydrologic data and DEM data using different interpolation and aggregation methods. The results show that Cressman object analysis method is the best aggregation method of all selected aggregation methods. Trend method is the best interpolation method for DEM. Accuracy of aggregation and interpolation for DEM is related with complexity of terrain. The more complex of the terrain is, the larger deviation is .the best interpolation methods for rainfall is Kriging methods. However, kriging methods are no advantage for rainfall interpolation when the density of rain station network is very low.The method that combines D8method with digitizing the main riverways, lake and reservoir (DMRLN) is used to extract river network form DEM. The river network extracted with this method reproduces the skeleton of true river network. The Concept distributed model based DEM is proved reasonable by calibration and validation of the model.With the increasing application of distributed hydrological model based on DEM the accuracy and resolution of DEMs is paid more attention. Topographty parameters and hydrological parameters extracted from DEMs are analyzed when DEM resolution is varied. The information content of different resolution DEM is quantified using entropy and analysis the sensitivity of hydrological parameters extracted from DEM to DEM entropy. Based on these analyses the effect of different DEM resolution on the runoff simulation and hydrologic parameter is analysis.Rainfall is one of the largest uncertain factors. At the same time it is the most important information for flood forecasting. The rain station network supplies the rainfall data for hydrologic analysis and simulation. However, raingage rainfall data acquired form point observations can not catch the distributing information of rainfall, especially in large scale of drainage area. Rainfall data form weather radars overcome these disadvantages. Effect of using spatially and temporally varied rainfall data on hydrologic simulation was analyzed in this study.