Different Resolution Dem Extraction Of Terrain Factors Of Appropriate Analysis Window

Digital elevation model (DEM) is the basis of GIS digital terrain analysis, it contains rich topography and landforms and the related geological information. Terrain factors extraction based on the DEM is the foundation of terrain analysis and terrain visualization. In GIS, terrain factors generally cannot get by a single grid, but be calculate through some neighborhood grids within a certain scope. According to the regional space range, terrain factors can be divided into microscopic terrain factors and macroscopical terrain factor, which have different calculation principle and method. Microscopic terrain factors usually get by a mathematical model of the local neighbourhood. For example, the current slope calculation neighborhood common based on the3x3rectangular windows, and set up the certain gradient calculation model, but which has not been further universality validation. Macro terrain factors are typically based on the statistical features of the local neighbourhood. For example, the relief is usually calculated by the difference of the maximum elevation and the minimum elevation within a certain scope, but there is a great differences in the neighborhood window size of different areas and different resolution DEM data, so we lake a unified standards and specifications in the whole of our county.This paper chooses different resolution DEM data of several typical terrain areas, the slope gradient as micro terrain factors and the relief as macro terrain factors. The two factors were extracted by different rectangular analysis window, the author mainly research the impact of window size on the terrain factors, and given the suitable analyze window for he calculation of the terrain factors based on DEM data through experiment and analysis. The main research contents and conclusions are as follows:1. Based on10m resolution and25m resolution DEM data of four typical sample regions—Binxian County of the Heilongjiang Province in northeast black soil, Ansai County in the northwest loess plateau, Yanting County in Sichuan Province karsts purple soil and Xingguo County of Jiangxi Province in the south red earth rolling terrain, using constrained quadratic surface to fit actual surface and calculate slop gradient through different size window, the consequence is that the impact of different window size on Ansai is most strong, the next in order as Xingguo, Yanting, Binxian, and the impact of different window size on25m resolution DEM is more sensitive than10m resolution DEM.2. Within Matlab, Using half the concave ellipsoid and gaussian synthetic surface to imitate actual surface, and comparing the accuracy of slop gradient within different sizes window, the consequence is that when the simulative DEM data was accurate,3x3analyze window can get the highest precision of slope; But when random error was join the simulative DEM data,5x5,7x7even more analysis can get the highest precision of slope. Comprehensive consideration of the fitting precision and accuracy of the slope, the suitable analyze windows of slop in4sample regions are that:the region of10m resolution in northeast black soil is7x7, the region of25m resolution in northeast black soil is25x25, the region of northwest loess plateau is3x3, the region of southwest karsts purple soil is5x5, the region of the south red earth rolling terrain is3x3.3. Based on90m resolution SRTM DEM of33typical sample regions in China, extracting relief through different size window, the consequence is that with the expansion of the window, the relief of the three plains in our country appeared straight line distribution or ladder type distribution, the relief of other regions, such as Hills, basins, platform, low mountain, high mountains, and very high mountains, appeared logarithm increasing trend, and the impact of different window size on mountain is most strong, and the impact on Hills, basins and platform is low. Moreover, based on500m resolution and1000m resolution DEM data of China, extracting China relief, the consequence is that the impact of different window size on1000m resolution DEM is more sensitive than500m resolution DEM.4. With the growth strength of relief to quantitative description of the impact of different analysis window and different landform areas on relief. When it reduced to zero value and think that change of relief is attain stable state. The paper finally given the suitable analyze window of relief for33different landform areas of our country, which are11levels and from7x7to27x27. and the whole country area when the growth strength of relief reduce to zero. At the same time, the suitable analyze window of500m resolution DEM in the whole country is15x15, and the1000m resolution is9x9.