The Spatial-Temporal Variation And Driving Force Research Of Land Desertification In Fuxin

Desertification is one of the severe ecological problems that catch the attention of the world,which directly affects the developing of the economic development and the stability of the society.The northwest of Liaoning province is one of regions with more serious land desertification in North China and even in the whole country.This paper uses Fuxin City in northwestern Liaoning Province as research area,and takes the Spatial-Temporal Variation of land desertification as the research objective,based on the Landsat remote sensing image data in 2000/2005/2010/2015,also according to the regulations of the monitoring technique about the desertification of china.The use of object-oriented and decision tree classification methods to obtain typical object categories,define modified vegetation cover,and achieve more accurate sand extraction information.On this basis,a targeted study was conducted around the spatial-temporal Variation law of land desertification,prediction of land desertification,driving force for land desertification,and prevention and control in Fuxin,as follows.(1)The Spatial-Temporal Variation research of land desertification in fuxin is Combined object-oriented and modified vegetation coverage inversion mode.For typical objects in Fuxin,taking advantage of the large quantities of the information and far and wide ranges of the observation as well as the strong timeliness about the remote sensing technology,on the basis of field investigation,the spectral and geometric and texture analysis,combined with the method of object oriented and decision tree model based on classification and extraction of typical desertification information with high efficiency and high precision,so as to identify the desertified land and land desertification trend.The modified soil adjusted vegetation index(MSAVI)is introduced into the vegetation coverage model as a parameter,and calculating the improved vegetation coverage value of land with sandy tendency.The modified vegetation cover was used as the main grading index,according to the desertification degree of the land;we divide the degree of desertification into mild desertification,moderate desertification,heavy desertification and extremely heavy desertification,and analyze the land desertification on two aspects of time and space.(2)Prediction of Fuxin land desertification based on Markov spatial-temporal evolution model.This paper used Markov spatial-temporal evolution model which has advantages such as low sample demand and high prediction precision to finish the prediction of land desertification,in order to realize the Fuxin land desertification simulation evolution and forecast.The experiment uses the Malkov method to constructed forecast model,five year as a step,calculates 2000 to 2015 Fuxin land desertification area transition probability matrix,conclude that for a year"weighted"step transition probability matrix and finish the forecast of the area about land desertification in 2015,after analyzing the feasibility of the fixed model based on the chi-square detection to forecast the Fuxin city land desertification situation in 2020.(3)Using the SRP model,we analysis the land desertification in Fuxin from the ecological angle,and Constructs the standard layer including the ecological resilience,the ecological sensitivity and the ecological pressure degree,the criteria layer contains 8 elements of landscape structure,vigor,function,terrain factor,meteorological factor,soil factor,population activity pressure and social environment pressure,and finally completed the analysis of land desertification in Fuxin using bythe land desertification evaluation System of the index layer,which is composed of 13 indices of landscape pattern index,biological abundance,point disturbance,population disturbance,slope,slope direction,elevation,average annual precipitation,average annual temperature,average annual relative humidity,soil erosion intensity,density and GDP density.(4)We replace the traditional resolution coefficient by using the adaptive coefficient matrix of iterative training,improved grey correlation algorithm,use the statistical data collected such as meteorological,gross domestic product,fixed asset investment data and population numbers and the area of extremely heavy desertification for correlation analysis,and then completed the quantitative description of correlation between impact factors and land desertification.Finally we can discover the main factors influencing land desertification were:sunshine duration,annual average relative humidity and total population number,in this,we finished the driving force analysis of land desertification in Fuxin based on grey relational analysis,according to desertification situationin 2020 we predicted based on development tendency.In addition to the climatic factors,we suggest strongly strengthening the construction of returning farmland to forests and weaken human activities to improve desertification control.