Analysis Of Land Use Change And Suitability On Geomorphic Type Units In Shanxi Province

The loess landform is an external representation of the soil erosion on the Loess Plateau,reflecting the material migration and energy conversion of the surface.It is the basis of the quantitative evolution of the loess plateau that its qualitative and quantitative methods to analyze loess landforms.In this paper,we take the Loess Plateau,Loess Hill and Loess ridge in Shanxi Plateau as the study area.First of all,the accuracy of the digital elevation model data(ASTER GDEM V2)in the study area was evaluated by ICESat GLA14 v34 data.Then,based on the analysis of the basic geometrical characteristics and surface morphological characteristics of the Loess Plateau,Loess Hill and Loess ridge entities,using the ASTER GDEM V2 data to construct the corresponding digital index system and selecting the basic factors for the quantitative expression of three types of landforms.Finally,three types of the loess landform recognition functions in the study area are fitted by template matching method.The main contents and conclusions are as follows:(1)In the study area,the mean and standard deviation of ICESat / GLA14 v33 and v34 data in different elevation ranges were analyzed,and the results showed that the vertical accuracy of the ICESat / GLA14 v34 and v33 data was centimeter level,and v34 was higher than v33.We took the ICESat / GLA14 v34 as the standard elevation,calculated the mean value of the elevation difference with the ASTER GDEM V2,the value is less than 1m,within the error range,and the mean standard deviation is concentrated.It means that the accuracy of the ASTER GDEM V2 data can meet terrain factor extraction requirements.(2)Based on the ASTER GDEM V2 data,the terrain factors such as the elevation,slope,aspect,relief,curvature(plane curvature and profile curvature),slope variability,aspect variability,topographic roughness,terrain wetness index and erosion of surface runoff are extracted by ArcGIS platform.Using the SPSS tool to take the correlation analysis and T test about the terrain factors of the Loess Plateau,Loess Hill and Loess ridge.The basic factors for the quantitative study of three types of landforms were selected,including area,slope,aspect,elevation,curvature and erosion of surface runoff.(3)Using ArcGIS statistical analysis module,three typical topographic data of Shanxi Plateau have been randomly classified according to the ratio of 95: 5,which are used for modeling and verification respectively.The recognition function of the Loess Plateau,Loess Hill and Loess ridge was established by template matching method in pattern recognition,and the calibration results show that the accuracy of the model is 82.61%.Therefore,the function established by the pattern recognition template matching method can be used to quantitatively identify three typical topography of the Shanxi Plateau.Above all,in this paper,ICESat/GLA14 data in different periods were explored and evaluated and ASTER/GDEM V2 data were analyzed based on the former data in Shanxi Plateau,which provided a reference for the application of two kinds of basic data.We take the accuracy analysis of the data as a prerequisite,the establishment of numerical index system and the fitting of quantitative expression are carried out for the three types of typical landforms in the study area,which is of great significance to the quantitative study of geomorphology in this area.