Study On The Desertification Prediction Based On CA-Markov Model

Desertification has been one of the most important issues that affect the global ecological environment.The expanding desertification has become an important factor hindering the development of human society. China has been affected by serious desertification, and desertification has been expanding to large scale area of North China in arid and semi-arid and semi humid areas. Based on the remote sensing images of the desert areas, by using the extraction indexs, the distribution and degree of desertification can be rapidly extracted. Combined with factors that can influence the development of desertification, the future development of the desertification can be accurately simulate, and this can slso provide the decision support on desertification governance and land resources planning.Based on the desertification extraction index, and the method of decision tree classification, this paper extractes the distribution of desertification in different periods in Shenmu County, analyses the spatial distribution characteristics of desertification and the factors of desertification change and desertification development. That using CA-Markov model to simulate and predict the future development of desertification.can provide decision support to desertification warning, desertification governce and ecological environment improvement. The main conclusions of this paper are as follows:(1) Shenmu County has suffered serious desertification. According to the desertification distribution map of the year of 2014. the total area of desertification in Shenmu County is 3791.63km2 and 52.35% for the total land area accounted. The main type of desertification is mild and moderate type. The desertification is mainly distributed in the northwestern region of the county, and the moderate and severe desertification has a increasing trend in the central and western regions.(2) In the last decade, the total area of desertification in Shenmu County has showed the trend of decrease. The total area of the desertification has reduced by 141.62km2. In the past, the trend of desertification change was reducing faster in mild and moderate desertification, but now it shows the situation that the reduction rate of mild and moderate desertification has slow down and the area of severe desertification has increased, which means it will be moret difficulty to further desertification governce(3) By using MS AVI, NDWI, soil brightness and texture features, the decision tree can be built. The overall classification accuracy reaches 80%, and the kappa coefficient is higher than 0.75, which represents the feasibility of desertification automatic extraction.(4) The desertification area by the predicting model in Shenmu County has increased in the year of 2020, and the severe desertification increases the fastest that in the future.Because severe desertification will increase fastest,enough attention should be given to this.