Dynamic change of red edge vegetation index within a growth cycle in arid area under coal mining stress

Red Edge Vegetation Index (REVI) is sensitive to chlorophyll concentration,leaf surface canopy,leaf cluster and canopy structure.The research on the response characteristics of REVI under mining stress can provide quantitative ecological indicators for vegetation damage diagnosis,natural vegetation restoration and artificial reclamation effectiveness evaluation in mining subsidence area.Taking the first mining face (No.113101) of Bojianghaizi Mine in Ordos City along the upper and middle Yellow River Basin as research instance,the mining-induced deformation field was obtained by DInSAR technology using Sentinel-1A SAR remotely sensed data,meanwhile,based on the traditional vegetation index,the red edge band was introduced to improve the existing vegetation index,the spatiotemporal variation of the vegetation indices in subsidence was analyzed with the help of normalized difference vegetation index (NDVI) and 5 typical VIre:RENDVI (Red Edge Normalized Difference Vegetation Index),NDre(Normalized Difference Red Edge),MCARI (Modified Chlorophyll Absorption in Reflectance Index),TCI (Triangular Chlorophyll Index),MSRre(Modified Red Edge Simple Ratio)-inversed from 11 periods of Sentinel-2A multi-temporal remote sensing data from October 2016 to October 2017.The results showed that ① the vegetation indices within a life cycle exhibit a single peak change of GAUSS.In this research area,the starting date of growing season (SOS) is in April,the peak is in August,and the ending date of the season (EOS) is in October.② The linear regression of 6 vegetation indices in subsidence area is smaller than that in check area (CK),which indicates that the growth rate of vegetation in subsidence area is slower than that in non-subsidence area.③ By studying the decline of subsidence area relative to non-subsidence area,it is found that affected by the subsidence disturbance,the increment of vegetation indices (δVIs) in the subsidence area is smaller than that in the non-subsidence area,and the average decrease of 6 vegetation indices is 18.8%-32.8%.The fitting effect of TCI is the best (R2=0.982 1).It further shows that the vegetation growth speed above the subsidence area is slower due to mining,and the red edge index is more sensitive to the vegetation growth in the mining subsidence area.④ Binary logistic regression shows that 4 vegetation indices (except for MCARI and TCI) are significantly correlated with coal mining activities in April,May,June and August (Significant level 0.001-0.1).However,in August,the 6 vegetation indices are significantly correlated (Significant level 0.001-0.026),and the significance level of TCI model is the highest.In the peak season of vegetation growth,coal mining activities have a significant impact on vegetation index above the working face.The vegetation indices based on red edge band combination can sensitively reflect the subtle changes of vegetation growth under mining stress,and can be used as the dominant indicators for the vegetation damage diagnosis and eco-environmental assessment of loess-aeolian coal mine in the Yellow River Basin.