Slope Scale Transformation Uncertainty Analysis

As an important and basic terrain parameters slope is being used as a key input for watershed hydrology and regional soil erosion modeling. At the regional scale, The slope is usually extracted from Digital Elevation Model (DEM) with a medium and lower resolution. However, slope extracted from DEM become reducing in the case that the resolution of DEM become coarser. In order to support regional scale soil erosion modeling, the slope from lower resolution DEM must be re-scaled to approximate higher resolution slope. This research studies slopes caling in different scale, and primarily analyzes the uncertainty of slope scaling. The main contents and results of this research are as follows:(1) Analysis of influence on slope caused by different DEM types. There are two different methods by which TINs are created from feature and Hutchinson algorithm in our experimental area, i.e. the Loess Plateau of Changwu County. The result has shown that DEM type is a key factor of slope extracting, two methods by which TINs are created from feature and Hutchinson algorithm have their own advantage, and DEM established by the improved Hutchinson algorithm by adding topographical feature lines has a same effect as TIN-DEM as well as the latter's advantages.(2) Analysis of slope reduction casused by the coarsing of resolution. A case study in Xiannangou Catchment analyzes the feature of slope reduction in the base of four different resolution DEMs, established by 1:10000,1:50000,1:100000, and 1:250000 topographic maps. This research creates 5m,10m,25m,50m, DEMs by using Hutchinson algorithm and extracts slope to analyze slope reduction feature. The result shows gentler slope area increase and moderate slope area decrease as resolution become lower. Average slope also decrease is shown by figure feature and statistical result.(3) Re-scaling the coarser slope with histogram matching approach:This research take an experiment of slope re-scaling in a base of four different resolution slopes in Xiannangou Catchment via slope histogram matching. It is analyzed from two aspects: different resolution (includes distance resolution and neighbor resolution) and different region. Different resolution slope re-scaling study shows that, slope reduction is corrected by slope scaling, but slope structure cannot comeback. Different region slope scaling begins with 1km2 as a model region in Xiannangou Catchment, extends to the whole catchment, and analyzes what influence will be brought when the region area is changed via the same re-scaling modeling. The result has shows no obvious variance appears while the area is spreaded.(4)Analysis of the uncertainty during the slope's scaling. We analyzes the uncertainty that will be resulted from relationshape between slope and DEM resolution, data type of DEM, and re-scaling region. At last, this research analyzes error in quantitative of slope scaling, and finds an error modeling.