Remote Sensing Monitoring And Comprehensive Evaluation Of Alpine Grassland Degradation

The alpine grassland ecosystem on the Qinghai-Tibet Plateau is an important part of the grassland ecosystem in China and an important ecological barrier in Asia.In recent years,the gradual deterioration of the global natural environment and the unreasonable exploitation and utilization of human beings have led to the degradation of the alpine grassland on the Qinghai-Tibet Plateau to varying degrees,which seriously restricts the ecological sustainable development of the alpine grassland in these areas.Accurate monitoring and evaluation of regional grassland degradation based on remote sensing technology is the basis of strengthening grassland protection,restoration and improvement of grassland ecological quality.However,at present,the exploration and understanding of remote sensing monitoring and comprehensive evaluation of alpine grassland degradation are not comprehensive.In addition,most research works are carried out based on low-and medium-resolution remote sensing satellite data,which often affects the accuracy and scientific nature of evaluation results.Based on this,the Gannan Tibetan Autonomous Prefecture,a typical representative region of alpine grassland,was selected as the research area in this study.The remote sensing monitoring and evaluation index system of alpine grassland degradation was constructed by using multi-source spatial data,and the spatio-temporal evolution characteristics of grassland degradation in Gannan Prefecture in recent 21 years were graded and evaluated from two aspects of single index and comprehensive index.Finally,the driving mechanism of different degradation indexes was discussed in order to provide a scientific basis for the effective management of grassland degradation and the formulation of regional ecological environment construction.The main research results are as follows:(1)In the study of alpine grassland extraction in Gannan,the accuracy ranking of the four machine learning algorithms was: RF>GDBT>CART>SVM;The user accuracy and producer accuracy of SVM and RF extraction of grassland can reach90.55% and 93.03%,and the recognition accuracy of grassland is good.(2)The results of degradation evaluation of different single indexes are quite different.The proportion of non-degradation grades was as follows: soil erosion(93.51%)>SOC(41.25%)>NPP(38.51%)>AGB(36.94%)>FVC(33.59%);The ranking of mild degradation grade was FVC(49.23%)>AGB(45.32%)>NPP(42.77%)>SOC(33.63%)>Soil erosion(5.21%);The sum of the three grades of moderate,severe and extremely severe was SOC(25.13%)>NPP(18.71%)>AGB(17.74%)>FVC(17.18%)>Soil erosion(1.28%).(3)The grade division of the comprehensive evaluation index of degradation was consistent with the real situation of the study area.The non-degraded grade of alpine grassland in Gannan was mainly distributed in the east and north of Maqu County and Luqu County,and scattered in other cities and counties.The mild degradation grade is mainly distributed in Xiahe County,Zhuoni County and Hezuo City,and scattered in other counties.The moderate degradation grade mainly concentrated in the west of Maqu,the middle and north of Xiahe County.The severe and extremely severe degradation grades are concentrated in the west of Maqu and the north of Xiahe.From2000 to 2020,the degradation grades of alpine grassland in Gannan were mainly nondegraded and mildly degraded,accounting for 80.31%(non-degraded 37.57% and mildly degraded 42.74%)on average.Moderate degradation followed,accounting for about 13.58% on average,and severe and extremely severe degradation grades accounted for the least(severe degradation 5.16%,extremely severe degradation0.95%).(4)From 2000 to 2020,the degradation grade of alpine grassland in Gannan was mainly reduced by the fluctuation of degradation grade,accounting for 40.60% of the total area,while the area with the increase of degradation grade fluctuation only accounted for 6.25%.The proportion of areas with constant degradation grade and constant fluctuation grade were 31.72% and 21.43%,respectively.The overall results showed that the degradation grade of alpine grassland in Gannan showed a stable and declining trend.(5)The factors influencing the spatial distribution of alpine grassland degradation index were altitude,soil type,average annual temperature,landscape diversity,annual precipitation and grazing pressure.The effect of natural factors is stronger than that of human activities and landscape factors.