Study On The Relationship Between Snow And Vegetation Change In Qinghai-Tibet Plateau Based On Multi-source Remote Sensing Data

The Qinghai-Tibet Plateau is very sensitive to climate change.In recent years,the land cover such as snow and vegetation has changed rapidly.Exploring the temporal and spatial changes of snow and vegetation on the Qinghai-Tibet Plateau and the differences in the impact of snow on vegetation will not only help enhance the understanding of the ecological environment of the Qinghai-Tibet Plateau itself,but also have important guiding significance for improving the Qinghai-Tibet Plateau's adaptability to climate change..Based on multi-source remote sensing data and meteorological data,this study explored the temporal and spatial changes of snow cover on the Qinghai-Tibet Plateau and the response mechanism of vegetation to snow cover changes.It has research characteristics in the use of multi-source remote sensing products,the trend identification of a variety of snow parameters,and the cognition of the interaction of influencing factors.The main research contents and results are as follows:(1)Three algorithms: Random Forest(RF),Back Propagation Neural Network(BPNN),and Convolutional Neural Network(CNN)were used to classify and analyze the surface types of typical areas on the Qinghai-Tibet Plateau.The results found that: First,the BPNN algorithm had the highest classification accuracy.The overall accuracy in 1990,2000,2007,and 2016 were97.82%,98.92%,98.92%,and 94.63%,respectively,and the kappa index was 0.958,0.959,0.980,and 0.918,respectively;Second,the snow cover in the study area was the main surface type from 1990 to 2000,the fragmentation of the landscape decreased,and the degree of landscape aggregation increased.From 2000 to 2016,the degree of aggregation of vegetation increased,while the degree of snow accumulation decreased,and vegetation was the main surface type;third,as the altitude increases,the possibility of conversion between vegetation and snow increased.(2)Using multi-source snow cover data to analyze the temporal and spatial changes and trend of snow cover on the Qinghai-Tibet Plateau,it was found that the snow depth of the Qinghai-Tibet Plateau from 1979 to 2019 decreased with the increase of the year,and the spatial distribution was very different.The snow was deep in the Tanggula mountain range,while the snow was small in the Qaidam Basin and southern Tibet.In some parts of the northeastern part of the Qinghai-Tibet Plateau,the snow depth showed an upward trend,while it showed a significant downward trend in most other areas.The snow cover on the Qinghai-Tibet Plateau last long in the west,central and northeast,and short in other places;the area with a downward trend in snow cover days(43.837%)was slightly lower than the area with an upward trend(46.053%),mainly in Qinghai-Tibet The plateau area of northern Tibet and the valley area of southern Tibet showed a downward trend.(3)Analyzing the relationship between snow depth data and the Normalized Difference Vegetation Index(NDVI),it was found that the snow cover of the Qinghai-Tibet Plateau from1983 to 2015 had an important impact on vegetation growth,and the area where snow depth was significantly positively correlated with NDVI The proportion was 19.061%,and the area with a significant negative correlation was 9.667%;the greater the altitude,the more obvious the negative correlation between NDVI and snow depth,but the smaller the altitude,the more obvious the positive correlation between NDVI and snow depth;five driving factors(Altitude,temperature,precipitation,slope and aspect)all affect the correlation between snow depth and NDVI,and altitude was the most important control factor for the spatial differentiation of the correlation between tne NDVI and snow depth.