Analysis Of Landscape Features Based On Information Entropy And Terrain Information Content

The traditional landscape pattern research is generally based on landscape maps,land use vector data and professional landscape analysis software for corresponding analysis and discussion.DEM elevation data contains rich terrain information.At present,there are few related research methods of landscape pattern based on DEM terrain information content.Therefore,in order to deeply understand the relationship between DEM terrain information and landscape distribution,This paper selected the Nanliujiang River Basin,Qinjiang-Maolingjiang River Basin and Baise Youjiang River Valley as the study areas in Guangxi,and built a quantitative expression model of terrain information based on the theory of information entropy,comprehensively used Arc GIS and Fragstats software to conduct research on the change characteristics of landscape index and landscape pattern under different terrain information levels.On this basis,a landscape index regression model based on DEM topographic information content was established for the Nanliujiang River Basin to predict the landscape index of Qinjiang-Maolingjiang River Basin with similar topographic features and Baise Youjiang River Valley with different topographic features,and compared with the real landscape index of the predicted areas to find out the adaptability of the regression model.At the same time,it further built a regression model of DEM terrain information content and landscape index for the Qinjiang-Maolingjiang River Basin to reversely verify the correctness and adaptability of the regression prediction analysis method.The research is conducive to simplifying the analysis of landscape features under the influence of multiple terrain factors,improving the efficiency of multi-regional landscape feature analysis under similar terrain features,and overall understanding the landscape utilization status under different terrain information.It is of great significance to improve the value of land use and optimize the allocation of land resources.Through analysis,it can be known that the landscape distribution has a close relationship with the terrain information content.This article mainly draws the following conclusions:(1)In terms of SHDI,the Nanliujiang River Basin and Qinjiang-Maoling River shows roughly the same changing trend.That is,as the terrain information contentincreases,the overall SHDI shows a slight decrease,the types of patches that make up the landscape decrease,or the difference in the proportion of each patch type in the landscape increases,and the spatial heterogeneity of the landscape increases.SHDI of Baise Youjiang River Valley has no obvious change pattern.Generally speaking,in areas with low or medium terrain information content,SHDI shows a slight decrease with the increase of terrain information content,but in areas with high terrain information content,SHDI increased or decreased significantly.(2)In terms of SHEI,the Nanliu River Basin and Qinjiang-Maoling River shows roughly the same changing trend,and the same trend as SHDI.That is,as the terrain information content increases,the overall SHEI shows a slight decrease,the uniformity of the patches that make up the landscape decreases,the diversity decreases,and the landscape dominance increases.SHEI of Baise Youjiang River Valley has no obvious change pattern,and its change state is similar to that of SHDI.(3)In terms of PAFRAC,the three study areas all show the same changing trend.That is,as the terrain information content increases,the PAFRAC shows a small increase and then a small decrease trend,and reaches a high value in the median area of terrain information content,the normal distribution characteristics is obvious,which reflectes the complexity of the boundary and shape of the landscape patch increases first and then decreased.(4)In terms of dominant landscapes,the dominant landscapes in the three study areas are cultivated land and woodland.In areas with low terrain information content,the dominant landscape is mainly cultivated land,and the dominant landscape in the remaining regions is woodland.(5)According to the prediction results of landscape index,the regression model based on the terrain information content and the landscape index of the known area is constructed to predicts the landscape index of the area to be studied.The prediction results of similar terrain regions have smaller errors and better accuracy than that of different terrain region.The regression modeling method based on terrain information and landscape index can be used to predict the landscape index of similar terrain areas.