The Loess Plateau Serious Soil And Water Loss Multi-scale Dem Terrain Factor Analysis

Topographic factors are important variables in the model of watershed hydrology and soil erosion, it determines the spatial variation of soil erosion way and intensity. At present, the extraction of terrain factors are done based on DEM, but the different DEM could express the different surface because of the DEM’s own characteristics, then topographic factors extracted from DEM with the different grid-sizes would be different, so the analysis of topographic factors with different grid-sizes is the important basic work of evaluating of soil erosion.Based on summing up predecessors research results and research experience, this paper selects4macro terrain factors as terrain index of regional scale research, that are relief amplitude, surface roughness, gully density and coefficient of variation. Using the basic theory and method of geographic information system, geomorphology and soil and water conservation, this paper take the loess plateau in northern Shaanxi for sample region, take25m resolution DEM for basic information source, and research the variation rules of topographical factors derived from DEM with different grid-sizes, and study the best resolution of1:50,000DEM on the study area. The main research contents and results are as follows:(1) Based on the theory of geomorphology and soil and water conservation, this paper analysis the influence of topography on soil and water loss. based on domestic and foreign major relevant topographic study on quantitative indicators of achievement in soil erosion prediction model, it identify the representative factors as regional-scale topographic indexes on soil erosion evaluation to study, and that could extract from DEM.(2) Taken1:50,000-topographic map draw up by the national survey and mapping sector for basic information source, the study track the contour using an interactive vectorization method, and establish DEMs with different grid-sizes from contour data by constructing a triangulated irregular network (TIN) modeling method. And the accuracy of DEM is tested. The results show that the DEMs have higher accuracy, that can be applied to later research.(3) In the research area, this study explores the topographical changes from DEM with different grid-sizes. Results show that the selected topographic factors could reflect surface characteristics of the sample area, describe the different types of landforms and reflect the area’s soil erosion. As DEM resolution goes down, the relief amplitude and coefficient of variation increases, while gully density and surface roughness decreasing. And using a curve fitting regression models, a quantitative expression of the relationship between topography factors and the resolution is given.(4) Draw on the gradient errors of thought, this article selects the relief amplitude and surface roughness to determine the best resolution of the1:50000DEM in the research areas. By computing topographic and surface roughness, here draw variation tendency of topographic and surface roughness separately with different grid-sizes, the grid-size can be considered to be the best resolution when curve showed a significant turning point. The results show that24is the best grid-size of1:50000DEM, this size not only guarantee the best simulation of surface, but also has a moderate amount of data. This method overcomes the haphazard and blindness in spatial analysis, and ensures the accuracy of the spatial analysis based on DEM, also provides an important reference for related research.