The Research Of Watershed Characteristics Based On Multi-scaled Digital Elevation Model

Digital elevation model is of great importance in Geographic Information Systemspatial data infrastructure. Nowadays,DEM provides a convenient and efficient approachin every aspect of hydrology. As is suitable for computer automation and digitalprocessing, and contains plenty of topographic information, DEM is widely used in thedescription of complex surface morphology and river basin water simulation. DEM isalso in widespread use in the extraction of terrain parameters and watershedcharacteristics. But in the process of applying digital elevation model to hydrologicalanalysis, there exists contradiction between scale and precision. On the one hand,high-resolution DEM can describe the topography in detail, but subjected to computerstorage capacity and data source constraints. Therefore,the application of high resolu-tion DEM is difficult.On the other hand, the low-resolution DEM influences the correctextraction of the watershed characteristics. Therefore, the study of what differencesbeween multi-scaled watershed characteristics from multi-resolution DEM can provide abasis for the choice of DEM at appropriate scale,and also can lay the foundation for thescale transformation.Based on the jiulong river watershed,digital terrain analysis technology is applied tothe study of automatic extraction of multi-scaled watershed characteristics. With thesupport of patial analysis method and statistical analysis the research is carryed out asfollows.(1) The research is on the extraction of drainage system and watershedcharacteristics based on digital elevation model. The1:5million DEM is selceted as thesource data of research.Based on digital terrain analysis technology, combined withhydrological module in the Arc/Info software,the drainage system of jiulong river andwatershed characteristics are extracted.(2) An efficient and fast method is proposed to determine the catchment areathreshold.In the first place, drainage systems are extracted at different catchment areathreshold. In the second place,fractal dimension is introduced to described the drainagestructure characteristics quantificationally. Though the discussion and analysis of therelationship of catchment area threshold and fractal dimension,a regression equation is given.By means of this equation,the catchment area threshold can be determined fast andprecise when the fractal dimension is in known.This method can avoid humansubjectivity,and improve the extraction precision of digital drainage system.(3) The scale effects on watershed characteristics at multi scales are discussed.Multi-resolution DEMs are obtained after the resampling of original DEM data.Subsequently different resolutions of watershed slope, topographic index, river networkdensity are extracted from the multi-resolution DEMs.With the application of spatialanalysis and statistical analysis method,the accuracy of these multi-resolution watershedcharacteristics at different scales are compared to find the discipline of changing scale onthe influence of them.