Font Size: a A A

Spatial And Temporal Dynamics And Spatial Autocorrelation Of Desertification Land In Mu Su Sandy Land In Recent 30 Years

Posted on:2020-09-24Degree:MasterType:Thesis
Country:ChinaCandidate:X Y HanFull Text:PDF
GTID:2381330578955524Subject:Soil and Water Conservation and Desertification Control
Abstract/Summary:PDF Full Text Request
Seven remote sensing images of Mu Su Sandy Land from 1990 to 2017 were selected.Based on ENVI 5.3 and ArcGIS 10.2,the area and spatial distribution of desertified land in different periods of Mu Su Sandy Land were extracted and the spatial and temporal dynamic changes of desertification were analyzed by supervised classification-neural network method.Based on OpenGeoDa software,the spatial autocorrelation of desertification land in Mu Su Sandy Land was analyzed by taking villages and towns as research units.The main conclusions are as follows:1.Dynamic change of time:The process of desertification from 1990 to 2017 can be divided into three stages:development stage:1990-2000,Serious and extremely severe desertification land continues to increase;rapid reversal stage:2000-2010,moderate and mild desertification become the main body of desertification types;stable reversal stage:2010-2017,desertification land of different degrees showed a steady reversal trend,and the annual change rate was relatively low.In the past 30 years,the area of desertification land has reduced by 1684.09 km2 and the desertification index has decreased year by year,which is in a reversal trend in general,but the deterioration in some areas is serious.2.Evolution law of spatial distribution:During 1990-2017,the gravity center of extremely severe desertification land was relatively stable and only moved westward 3.42 km;the gravity center of severe desertification land moved northwest 8.80 km;the gravity center of moderate desertification land shifted northwest 5.42 km;and the gravity center of light desertification land extended southeast 9.90 km.Up to 2017,the desertification degree of Wushen Banner,located in the hinterland of Mu Su Sandy Land,was the most serious;Etok Qian Banner,Shenmu County,Yuyang District,Yanchi County and Yijinholuo Banner are mainly Mild and moderate desertification;Etok Banner is dominated by moderate and severe desertification;Dingbian County,Jingbian County and Hengshan County are mainly mild,moderate and non-desertied land.3.Spatial autocorrelation of desertification land in villages and towns:At the significance level of P<0.05,all of the global autocorrelation indexs Moran's I are greater than 0.2446,and positive spatial correlation is relatively high.The more serious the desertification degree is,the stronger the spatial correlation is.The different degrees of desertification are mostly clustered in H-H and L-L,and there is no L-H anomalous aggregation occurred in the study area.The H-H aggregation of desertified land mainly occurs in the villages and towns under Ushen Banner,Etok Banner and Etok Qian Banner,and L-L aggregation mainly occurs in Yuyang District,Jingbian County and Hengshan County.
Keywords/Search Tags:Mu Su sandy land, Desertification, Spatial-temporal dynamic change, Spatial autocorrelation, Supervisory classification-neural network method
PDF Full Text Request
Related items