Monitoring The Dynamics Of Desertification In Northern China And Analysis Of Its Driving Forces

Global and regional climate change and human activities are the main causes of desertification,and land desertification has become an important ecological and environmental problem that human beings must face,which seriously restricts the regional ecological balance and human economic and social stability.In the context of global warming,remote sensing dynamic monitoring of desertification and its driving factors are of great theoretical and practical significance for scientific understanding of the desertification process and its spatial and temporal patterns in northern China and adjusting the policies and measures for prevention and control.This paper takes northern China as the study area,based on MODIS 09A1 data from 2000 to 2020,analyzes the applicability of different land desertification remote sensing monitoring models to the study area,and searches for the best monitoring model to study the spatial and temporal changes and driving forces of land desertification in northern China;in addition,based on GIS geostatistical analysis method,the change characteristics of land desertification,spatial distribution pattern and evolution law of center of gravity are comprehensively analyzed from time and space dimensions;finally,Pearson correlation coefficient method,geographic probe and geographic weighted regression model are applied to explore the key factors affecting spatial and temporal changes of land desertification by combining 11 factors in climate,physical geography and socio-economic aspects and The main findings of this paper are as follows.The main research findings of this paper are as follows:(1)Based on Thornthwaite’s formula combined with the data from ground meteorological stations,the potential occurrence range of desertification in the northern region of China was determined,providing a theoretical basis for the establishment of the subsequent desertification monitoring system.In this paper,the climatic zoning of desertification based on the calculated wetness index finds that the area share of each climatic type shows a decreasing trend between 2000 and 2020,among which the area share of semi-arid zone decreases more significantly,and its average growth rate is-0.302/a;among the climatic zoning types,the area share of sub-humid arid zone is the smallest,followed by arid zone,and the area share of semi-arid zone is the largest;up to 2020,the area of potential occurrence of desertification in northern China accounts for 34.42%of the total area of northern regions in China.(2)In this study,the Remote Sensing-based Desertification Index(RSDI)was constructed using principal component analysis,and the index was compared with the existing image dichotomous model,Albedo-NDVI feature space(line-to-line and point-to-point)and Albedo-MSAVI feature space(line-to-line and point-to-point).point-to-point)extraction results for a comparative analysis of full-area and regional scale accuracy.In the northern region of China,the RSDI-based model not only compensates for the deficiencies in the existing desertification monitoring models,but also has the highest model accuracy,with an overall accuracy of 87.11%.(3)In the trend of temporal change,the land desertification in the northern regions of China shows a decreasing trend,and its average growth rate is-0.0007/a.Spatially,the spatial change trend of desertification in the northern regions of China is mainly based on the trend of gradual improvement,and the regional proportion is 48.79%,which is mainly distributed in the key areas where ecological restoration projects are implemented in China.In the future trend analysis based on Hurst index,it is found that the reverse persistence of desertification in the future is stronger than the positive persistence,which also indicates that the future desertification degree in the northern regions of China shows a gradual improvement trend.The center of gravity of each desertification type has the largest migration distance(553.86km)for the moderate desertification type and the smallest migration distance(36.93km)for the severe desertification type,and the influence range of extreme,severe,mild and non-desertification is gradually decreasing,while the influence range of moderate desertification is gradually expanding.(4)Using Pearson correlation analysis to qualitatively explore the spatial and temporal correlations between different driving factors and RSDI,the average wind speed,average temperature and sunshine hours showed positive correlation trends,and the precipitation and average relative humidity showed negative correlation trends.The influence of different driving factors on RSDI was further quantitatively investigated from the global scale by using the geographic probe model in terms of factors and interaction detection,respectively.There was no significant change in the influence of natural geographical detector on desertification in northern China,while climatic and socio-economic factors had an increasingly strong influence on desertification,and total precipitation and mean relative humidity were the main factors affecting desertification in northern China.Using a geographically weighted regression model,the influence of different driving factors on RSDI was further studied quantitatively at the local scale.The effects of different factors on RSDI in different periods and spaces showed variability,with total precipitation showing a positive contribution,and average temperature,sunshine hours and elevation having both positive and negative effects.