Remote Sensing Analysis On The Landscape Of The Mu Us Desert

In this paper, surface information is obtained through remote sensing technology. Using the methods of remote sensing and landscape ecology, the author analyzed the dynamic changes of sandy ecological landscape in research area.Comparing the CBERS images in May 2006 with the TM images in 1997, pretreatment and geometric correction of the remote sensing images are finished by software EDARS. Further analysis on characteristics of vegetation, It was classified into eight landscape elements. Through human computer interaction, the 2006 landscape classification diagram is completed. Moreover, classification results are put into the software FRAGSTATS, and output is landscape pattern index. Based on the horizontal comparison of characteristic index, in land check-up, and the analysis of the distribution of landscape elements, the character of type plaque is systematically discussed. Finally, the author summarized the major elements of sandy landscape, furthermore, made the suggestion that heterogeneous symbiosis should be promoted and the stability of landscape improved.The main research results are as follows:(1) Processing mode of the CBERS images is established, the studies of band register, fourier analysis and crisp enhancement blends in treatment, which showed a good effect. After the processing image carries on the geometry correction, increased the control point selection precision, is suitable to the processing of desert area.(2) About 20 control points can reach the precision demand on uniform distribution, especially a desert region border part. There is effect of precision by choosing more than 20 control points in the process.(3) Landscape element remote sensing interpret tag is established, through supervised classification in desert area. The author obtained classification information based on the character of desert, mainly with the cursor.(4)According to description and comparison the Landscape index, Firstly, Patch of sand play an important part in landscape element absolutely. As a result, it is a long term to control desert, which is in integrity. Secondly, landscape grain of shrub and grass is bigger than others, the main type of sand-fixing vegetation are shrub and grass with broadly distribution. Thirdly, The artificial patch, water patch and woodland have opposite distribution trend, owned the smaller landscape fragmentation index. The forestland type connectivity is low, the result indicates that There is no effect on ecological damage and destruction by activity of human while water landscape is dotted among other landscapes. Finally, landscape diversity index, landscape evenness index and dominance index are calculated in this paper.