| Accurate and effective monitoring of spatial and temporal changes of vegetation fraction in the Tibetan Plateau region is important for analyzing the sensitive and fragile ecological environment of the region.It is difficult to acquire image data with high temporal and spatial resolution due to the influence of satellite sensors and cloudy and rainy weather.Using spatiotemporal fusion model to generate high-resolution remote sensing images of ecologically sensitive areas provides a new idea and technical support for the study of vegetation fraction in this region.In this paper,Landsat and MODIS data are used as the basis to generate high-resolution remote sensing data from 2005-2019 by the spatiotemporal fusion method(FSDAF),and NDVI data are calculated according to the cumulative frequency,and the vegetation fraction of the region is calculated by the dimidiate pixel model,taking the Sangzhuzi district of Rikaze city as the study area.Then,the spatial and temporal trends of vegetation fraction were studied by using unary linear regression analysis,Hurst index and other methods,and correlation analysis was conducted between meteorological data and vegetation fraction to reveal the correlation degree among factors and elucidate the influencing factors of vegetation fraction changes in the region.The main research results are as follows:(1)Accuracy verification of spatio-temporal fusion results.The accuracy of the fusion results was evaluated using five commonly used indicators,such as R^2,CC,RMSR,PSNR and SSIM,with the range of R^2 values from 0.4270 to 0.7763,CC values from 0.6472 to 0.8761,RMSR values from 0.0195 to 0.0308,PSNR values from 43.4632 d B to 47.032 d B,and SSIM values was from 0.7991 to 0.8868;the best performance of all indicators was in band 1,and the worst performance was in band 4;the fusion result on September 7 was the best and the closest to the real image,and the fusion result on March 15 was the worst;from the accuracy verification of NDVI,the range of R^2 values was from 0.5312 to 0.8763,and the range of CC values is 0.7284 to 0.8984;therefore,the fusion results from March to October can basically meet the needs of the study,and the input images near the prediction date can be selected according to the actual situation of the study area.(2)Analysis of spatial and temporal variation of vegetation fraction.2005-2019 annual average FVC in the study area fluctuated widely between years,with values ranging from 0.3 to 0.54,but the vegetation fraction showed an increasing trend in the past 14 years,indicating that the ecological environment in the study area is more fragile,but overall the vegetation in Sangzhuzi district showed an improving trend;the study area was mainly dominated by the changes of medium to high vegetation fraction types,mainly the changing vegetation fraction in the study area is distributed in the mountainous areas in the south,north and east,and the rest of the vegetation fraction types have slightly less changes;the vegetation fraction in Sangzhuzi District is mainly in the trend of stability and improvement in general,with the area share of46.08% and 44.96%,and the areas in the trend of stability are mainly distributed in the river flowing areas and some mountainous areas,the areas in the trend of improvement are mainly distributed in the mountainous areas,and the declining trend is mainly located in the valley areas adjacent to the settlements;the study area is dominated by weaker persistence,with weakly antipersistent ? stable and weakly positive persistent? stable areas mainly scattered in mountainous areas,with area shares of 19.17% and17.79%,and weakly positive persistent ? improving and weakly antipersistent ?improving areas mainly distributed in valley and valley-mountain transition areas,with area shares of 25.27% and 19.02%.(3)Response analysis of vegetation change and climate factors in the study area.The overall average annual precipitation in the study area from 2005 to 2019 showed a weak fluctuating upward trend,and the average annual temperature showed a weak fluctuating downward trend;vegetation and climate factors in the study area were mainly insignificantly correlated,among which,precipitation was mainly positively correlated(76.28%),distributed in mountainous areas and valley-mountain transition areas,and temperature was mainly negatively correlated(61.04%),distributed in most mountainous areas.As the elevation increases,the percentage of areas with significant correlation with precipitation gradually decreases,while the area with insignificant correlation gradually increases,indicating that the significance of precipitation on vegetation gradually decreases as the elevation increases;the significant effect of temperature on vegetation is lowest in the area from 4000 to 4500 m,and the rest of significant effects show a gradual increase from low to high elevation. |
PDF Full Text Request