Change Detection Of Vegetation In Mining Area On Spatiotemporal Fusion Data

The land use change in coal mining has caused great disturbance to the ecological environment.In order to study the local ecological conditions represented by vegetation growth in the mining and surrounding areas,land use change detection is needed.The real-time update of land cover and its dynamic changes on a higher spatial and temporal scale contributes to the detection of ecological continuity changes in the mining area,analyzes the changes in a timely manner and makes feedbacks that are beneficial to the local ecology.Ningdong Coal Base is a typical fragile ecosystem in the arid desert region of Northwest China,which is particularly sensitive to disturbances caused by mining.The detection of high temporal and spatial continuity changes not only helps to discover the differences of vegetation changes in different seasons,but also explores the disturbance effects of human activities on fragile ecology.It has certain reference significance for the detection of vulnerable ecological regions in other mining coal mines.In this study,Ningdong Coal Mine Base of Ningxia Hui Autonomous Region was used as a research area for land use change detection.Object-based spatiotemporal data fusion,NDVI time series based on spatiotemporal fusion data and time series change detection method were used to improve the accuracy of change detection.It also enhances the continuity of the time series,and obtains the NDVI time series with high spatial resolution and time resolution.Simultaneously,this study explores land cover changes and the impact of coal mining on ecology from both qualitative and quantitative aspects.The main research methods and conclusions of this paper are as follows:(1)the object-based method is used to integrate the Landsat and MODIS images in 2007 and 2017.the NDVI time series with spatial resolution of 15 meters and time resolution of 16 days were obtained.The correlation data of the fusion data were evaluated,and the correlation is 0.9.In the above,the time-space fusion NDVI time series which can accurately reflect the changes of surface vegetation is obtained.In this study,the changes of coal mining bases in this fragile habitat in the arid desert region of Northwest China are continuously tested.The time-space fusion data NDVI time series can be used to obtain the continuous changes under time characteristics and changes of land types under spatial characteristics.(2)Using the two-year NDVI time series to study the seasonal variation and the change of land cover type in continuous time,the accuracy of the change detection is 0.85.Through the change test,it was found that the NDVI value of the study area changed in a zigzag pattern with the season.The peak period of NDVI corresponds to the abundant rain from July to September,which is characterized by obvious arid regions.Comparing the two-year NDVI time series,it can be seen that the overall NDVI in 2017 is higher than that in 2007,indicating that the overall ecological situation of the region is improving.On the other hand,the change of vegetation area in the two years can be obtained.Although some grassland was converted into forest land,there was no significant improvement overall,and the green area did not increase significantly.Although the study area has continuously improved the overall ecological level under the policy of ecological restoration,there is still room for improvement in the growth of vegetation area.(3)Spatial superposition analysis of the changed area and the detected mining subsidence area shows that the land use change at the subsidence boundary presents a vegetation growth trend.This correlation was not found in other regions.It shows that the subsidence caused by mining may be positively correlated with vegetation growth,and this discovery needs further research in the future.In summary,object-based spatiotemporal data fusion effectively improves data utilization efficiency,simplifies fusion steps,and improves fusion accuracy.The NDVI time series reflects the changes of vegetation seasonality and land cover types in the study area,and improves the detection accuracy of changes.The subsidence caused by coal mining may be related to the vegetation.It is necessary to further explore the impact of subsidence on the ecology of the study area in the next study.