Improved Approaches Variation in Grassland Desertification Based on Landsat Images and Spectral Mixture Analysis and Decision-Tree Method

China has a very large area but most of the population in China is living in the Northern Area because south area has mountains and hill and Northern Area has plan land. But from very slow process china is facing desertification in China and for find the desertification Landsat images study and Spectral Mixture analysis. We use a new approach to for this study Decision-Tree Method. In Past 4 year’s desertification is increasing very fast and we are using Yanchi County, China as a study area. The analyse area were covered by free could Landsat MSS, TM, ETM+. The series of images will analyse by the Landsat TM/ETM+. On the basis of the result grading system will construct and then we will use spectral mixture and then decision free method will use to interpret the images. By the methods RMS errors computed. By direct detection firstly find out the fraction change in images and then we apply change vector analysis to find out the land covert change. To find out the soil erosion and for mapping index we use EMI (Eolen Mapping Index), statistical measurement will analyse by ANOVA.