| The rocky desertification in the karst area of southwest China is a manifestation of the degradation of the karst ecosystem to the extreme,which seriously restricts the healthy development of the regional social economy.In the southwest Karst ecological environment is extremely complex and diverse karst niche partition takes full account of the karst basin,such as geology,landform,hydrology structure characteristics,ecological environmental conditions,rocky desertification causes and control measures of similarity,it plays an important role in depicting the internal characteristics of karst ecosystems and revealing the spatial heterogeneity of rocky desertification,supplementing and perfecting the understanding of the karst environment.Based on the Landsat data of 1995,2005,and 2015,this study selected one study area from each of the four habitats of the karst cluster-peak-depression region,dislocation basin,Karst Plateau and Karst Valley 4 habitats,on the basis of the original classification system of rocky desertification,the fuzzy reasoning system is used to construct the fuzzy division of the degree of rocky desertification,and the improvement and innovation of the classification method of rocky desertification is realized.The classification accuracy of this method is 86.16%,the Kappa coefficient is 0.83,and the coincidence rate with sampling results is more than 0.7,which shows that the fuzzy classification method of rocky desertification degree has high accuracy.In order to comprehensively explore and describe the details,characteristics and anomalies of the spatial and temporal evolution of different karst habitats in the past 20 years,the article perfects the framework for the analysis of the temporal and spatial evolution of rocky desertification.Macroscopically,the evolution and differentiation characteristics of rocky desertification in different habitats are analyzed from the spatial and temporal level.Microscopically,the Evolution Direction,trajectory,dominant type,rate,frequency and anomaly of rocky desertification are analyzed,A comprehensive and in-depth grasp of the evolution details of rocky desertification in different habitats.The results showed that the overall situation of rocky desertification showed reverse evolution,and the spatial differentiation of rocky desertification was obvious between different habitats and within habitats.In view of the shortcomings of quantitative and spatial analysis of driving forces of rocky desertification,spatiotemporal geographic weighted model(STWR)was introduced into the study of driving forces of rocky desertification for the first time.The regression effect of driving factors was evaluated by comparing the parameters of the global regression model OLS of different habitats and three geographically weighted regression models(GWR,MGWR,STWR),such as the determination coefficient R~2,AICc information criterion and residual analysis.The results show that the performance of the geographically weighted regression model is significantly improved compared with that of the global regression model,and the STWR model performs the best.On this basis,the driving mechanism of rocky desertification in different habitats was analyzed in detail.In the prediction part of rocky desertification evolution,the study combines top-down scenario setting and bottom-up local quantification and combination.Based on the basic data of rocky desertification in 2015,the occurrence probability and neighborhood conversion probability of rocky desertification were calculated grid by grid under three scenarios of natural evolution,positive mild intervention and positive strong intervention,and the spatialization of dynamic simulation and prediction of rocky desertification in different karst habitats in 2035 was realized for the first time.Based on the forecast results of different rocky desertification development in each habitat,various rocky desertification control and land management strategies are proposed. |
PDF Full Text Request