Spatial And Temporal Change Analysis Of Urban Green Space Based On Remote Sensing Data

This paper uses Landsat remote sensing image data to classify the green space of Beijing's built-up areas,and on this basis,analyzes the internal green space area,coverage,per capita green space and their changes in the urban built-up areas,and quantifies the changes in multiple indicators Study the changes in the green space inside the urban built-up area.The research results show that the area and coverage of green space in Beijing's built-up areas in 2019 have increased compared with 1990,but the per capita green space has declined to a certain extent,and the green landscape has been seriously fragmented.The main research methods and results of this article are summarized as follows:(1)Based on the normalized vegetation index(NDVI)time series,a random forest classification algorithm was used to classify Landsat images in Beijing area.Compared with the classification results of support vector machines,it was found that the random forest algorithm had the best classification effect.The Savitzky-Golay filter was used to reconstruct the NDVI time series feature set to classify the remote sensing images in the Beijing area.Four land use classification maps in 1990,2000,2010 and 2019 were obtained and their accuracy was verified.The highest accuracy reached 94.52%,and the lowest accuracy was 89.57%.(2)Based on the results of land use classification,analyzing the land use changes in Beijing from 1990 to 2019,it is found that the area of construction land has been increasing,while the area of woodland and arable land has declined,and the water body and bare soil have fluctuated greatly;The land use transfer matrix shows that the increase in construction land area has led to the decrease of cultivated land area;the dynamics and comprehensive dynamics of five land types are analyzed.Among them,the change rate of water body and bare soil is relatively fast,and the changes of other types are relatively stable.(3)Analyze the urban green space landscape pattern of Beijing area based on population data and landscape index,and obtained several basic landscape pattern indexes,and found that the fragmentation of Beijing's landscape has increased in recent years;combined with population data and green space The area,the amount of green space per capita and the percentage of green space reachable per person in each year are obtained.It is found that the percentage of green space per capita and reachable green space per capita declined before 2010 and increased from 2010 to 2019.This paper establishes an automatic land use classification method based on remote sensing time series data,and obtains a satisfactory classification accuracy in Beijing area,and studies the land use changes in the Beijing area in the past 30 years and the temporal and spatial changes of the green space in the past decade.It can provide a constructive basis for the planning and construction of Beijing urban green space and the sustainable development of the city.