Research On Mine Ecological Disturbance And Vegetation Restoration Monitoring Based On LandTrendr Algorithm

The development of mines has greatly changed the landscape type and spatial structure of mines,resulting in ecological and environmental problems in mines and affecting human beings.Especially the changes of vegetation coverage,so analyzing the dynamic changes of vegetation coverage during the development of the mine and further analyzing the change characteristics have an important role in studying the ecological environment of the mine.Remote sensing means to monitor the dynamic changes of vegetation is of great significance to indicate ecological and environmental changes,but the comparative study of remote sensing images between different years is not enough to reflect the dynamic changes of mine ecology,so it is necessary to use high-density remote sensing images under long time series to carry out mine ecological disturbance And recovery monitoring studies.This article introduces the Landtrendr algorithm that can simultaneously detect change trends and disturbance events.Taking the fast-developing Zijinshan gold and copper mine as an example,the Landsat image from 1990 to 2018 is selected as the remote sensing data source.Using the Google earth engine cloud platform,the first study Landtrendr algorithm monitors the process of mine ecological disturbance and vegetation restoration.On this basis,the year of occurrence of ecological disturbance and vegetation restoration of Zijinshan gold-copper mine,the amplitude of spectrum index change and the duration of change were extracted,and the dynamic changes of mine ecological disturbance and vegetation restoration were further analyzed.This article mainly achieved the following research results:(1)Based on the Google earth engine cloud platform,the image is preprocessed,and the multi-dimensional median synthesis is performed on the image of the same season to form a 1990-2018 standard long-term series data set suitable for processing by the Landtrendr algorithm.This data set is the basis for subsequent dynamic change data extraction and is also an important reference data for accuracy verification.(2)According to the characteristics of the mine ecosystem and the changes in vegetation coverage during the mine development process,a mine ecological disturbance and vegetation restoration model is established.On this basis,the Landtrendr algorithm has been developed to monitor the mine ecological disturbance and vegetation restoration,including determining the spectral index and debugging The relevant parameters of the algorithm enable the algorithm to accurately extract information related to the ecological changes of the mine.The Landtrendr algorithm is used to segment the time series,and the trajectory is appropriately segmented and linearly fitted,so that the complex spectral time series is modeled as a series of straight line segments,so as to accurately and efficiently extract the change time of ecological disturbance and recovery in the study area Variation and duration information.The accuracy analysis of the extraction results shows that the overall accuracy of disturbance events obtained by Landtrendr algorithm is 83.42%,and the overall accuracy of recovery events is 82.96%.(3)Based on the Landtrendr algorithm,the comprehensive analysis of the monitoring results of the ecological disturbance and vegetation restoration of the Zijinshan gold and copper mine shows that:From the perspective of the year of occurrence,the characteristics of ecological changes in the study area gradually expanded from the center to the surrounding areas,and the ecological vegetation restoration in the southwest part of the study area was in good condition.And the ratio of cumulative disturbance to vegetation restoration in the study area reached 75.37%,which may indicate that Zijinshan Gold and Copper Mine is undertaking more active ecological vegetation restoration projects.In addition,from the comprehensive view of the change of the spectral index and the duration of the change,the disturbance presents a rapid and violent disturbance state with a short duration,and the magnitude of the decrease in the spectral index is related to the type of growing vegetation;while the recovery shows a relatively stable state for a long time Longer areas with larger rises are related to man-made ecological restoration projects.