Changes Of Slope Statistical Distribution With The Complexity Of The Terrain

Slope is one of the important variables of surface morphology and the main factors in soil erosion. Previous studies on the slope are extracted directly with DEM, and using the statistical histogram to represent the slope distribution. It does not accurately describe the impacts of changes of terrain complexity and DEM resolution on the slope distribution. The thesis research the slope distribution varying with the complexity of the terrain in 15 research sites with varied terrain types based on slope statistical distribution model and the common slope extraction algorithm. The study also analyzes the changes of the key parameters in the slope model with the different terrain types and DEM resolutions. The main results are as follows:(1) The slope statistical distribution model can be used to express the slope distribution of different terrain types and different DEM resolutions. The histograms similarity of the slope model and the actual data were bigger than 0.8. The simulation effects of the slope distribution were well.(2) The model parameter σg tend to be smaller with the terrain complexity becoming simpler or the DEM resolution becoming coarser. The peak interval of slope frequency curves became narrow and moved to gentle slope. In Hengduan mountain site, the slope was mainly presented as extremely steep slope. The slope was mainly presented as relatively steep slope in Loess Hilly area and Purple Soil Hilly area in Sichuan province. And the peak interval of the slope frequency curves were relatively wide in these two areas. In Northeast Gentle Hilly Black Soil area, the slope was mainly presented as low slope The peak interval of the slope frequency curve was relatively narrow in this area. The slope frequency curves of Tianshui site with complex terrain varyed greatly, while the Baiquan site with simple terrain varyed slightly.(3) The accuracy of the model depends on the terrain types and the data precision. The simulation effects were great in medium slope, like Suide, Tianshui, Gushanchuan, Huining, etc; But not good enough in extremely gentle slope or extremely steep slope, like Baiquan and Hengduan mountain. The analysis showed that the model had a good simulation effect in all kinds of DEM resolutions. While the histograms similarity declined slightly with DEM resolutions becoming coarser.