Vegetation Coverage And Dynamic Change Analysis In Karst Region Based On Spatio-temporal Fusion Of Multi-source Remote Sensing Data

With the development of the application of remote sensing technology,the field of remote sensing has entered a wider range,higher accuracy and deeper application stage.Under the complicated conditions of karst terrain fragmentation and ecological fragility,vegetation coverage and other data with high spatial and temporal resolution are urgently needed to support related researches such as vegetation dynamics and comprehensive prevention and control of rocky desertification.Due to technical and budgetary constraints,it is difficult for a single sensor to acquire data that satisfies both high spatial resolution and high temporal resolution.In view of the problem of missing Landsat data in the Karst area of Guizhou due to persistent cloud cover,cloud shadows and other reasons,the low spatial resolution of MODIS data is not suitable for small-scale applications,as well as the need for high spatial and high resolution vegetation coverage data in studies of rocky desertification in karst areas,taking karst area as the research area,based on the comparison of Landsat data and MODIS data to study the applicability of STARFM,ESTARFM,FSDAF three space-time fusion methods in karst area.the application is suitable for the research area.Reconstruct the NDVI data set of the long-term series from 1988 to 2018 in the Baifu River Basin and the vegetation coverage data set of high spatial resolution of the long-term series by applying the space-time fusion method and Savitzky-Golay filter algorithm suitable for the Baifu River Basin And analyze the characteristics and driving factors of the dynamic changes of vegetation coverage in the Baifu River Basin from 1988-2018 from a qualitative and quantitative perspective,and draw the following conclusions:1.The spatial details of the fusion image obtained by the ESTARFM method and the FSDAF method are the clearest and the layering is more obvious.The result image of the STARFM model is blurry,the layering is poor,some spatial details are lost,and there is plaque phenomenon.The five bands of Blue,Green,Red,NIR,and NDVI generated by the three space-time fusion methods of STARFM,ESTARFM,and FSDAF are generally the ESTARFM method and the FSDAF method.Compared with the STARFM method,they obtain higher R and SSIM,and a smaller RMSE.,And the accuracy of reverse fusion of the three spatiotemporal fusion methods is higher than that of forward fusion.The ESTARFM method and FSDAF method have better applicability than the STARFM method in the study area.The FSDAF model only needs to input a pair of initial data pairs,which is flexible to use,and has certain advantages for applications in regions where it is difficult to obtain the initial data pairs due to cloud coverage.2.The NDVI data processed by the Savitzky-Golay filtering algorithm has higher fidelity to the original NDVI data.Based on the existing data,after de-cloud processing,surface reflectance spatio-temporal fusion and Savitzky-Golay filtering,Landsat time series datasets can be constructed in areas where some Landsat series data is missing due to persistent cloud coverage,cloud shadows,and other reasons.3.The Savitzky-Golay filter algorithm can effectively retain the effective information of the original time series curve of vegetation coverage in the study area,and eliminate the influence of noise in the original data.From 1988 to 2018,the overall vegetation coverage of the Baifu River Basin showed a significant increase.The average vegetation coverage increased from 0.516 in 1988 to 0.699 in 2018,with a growth rate of 0.0079,an increase of 52.4%,and the linear growth trend is significant.From 1988 to 2018,the area of high vegetation coverage area and medium and high vegetation coverage area increased by 163.93% and 6.95% respectively;the area of medium vegetation coverage area,medium and low vegetation coverage area and low vegetation coverage area decreased by 47.71%,65.44% and 46.43% respectively.Vegetation coverage in the Baifu River Basin presents both improvement and degradation,the improvement area is greater than degradation,and the overall vegetation development shows a benign evolution trend.The vegetation coverage changes in the watershed is mainly affected by human activities such as returning farmland to forests and construction activities,and the correlation with temperature and precipitation is not obvious.