Remote Sensing Information Extraction For Solid Waste Landfills Based On Multi-source Remote Sensing Data

At present,landfill is one of the main ways to deal with domestic waste worldwide,which has the advantages of low cost and health.However,the heat,leachate and harmful gas are produced during the landfill process,which seriously endangers the surrounding environment and the health of residents.With the advantages of low cost and high efficiency,remote sensing technology has been gradually applied to all fields of the earth’s ecological environment,which can provide data and technical support for large-scale dynamic monitoring of landfill sites.Suzhou has developed river network,flat terrain and high level of urbanization.The leachate produced by domestic garbage landfill is easy to pollute the groundwater environment,and the harmful gas produced at the same time is harmful to the health of the surrounding residents.Therefore,it is of great significance for the location and dynamic change monitoring of landfill sites.However,at present,the research on the identification and extraction of landfill sites based on remote sensing technology is limited.At the same time,due to the small area of landfills and the complex components,the image resolution and identification methods have very strict requirements.Taking Suzhou as an example,this study constructed a feature set based on spectral features,texture features and shape features by using multi-source data with high temporal and spatial resolution(Landsat 8,Sentinel-2A and GF-1 multi-spectral data),and studied an effective method to identify and extract landfills.It provides a new method and technical support for dynamic monitoring and scientific management of landfill.The main contents and conclusions of this study are as follows:(1)Landsat 8 based surface temperature law study of landfill sites.In suzhou city,covering an area of the largest mountain landfill site as an example,the seventh,peulthai the selection of 2015-2020 Landsat 8 remote sensing image,using the method of atmospheric correction inversion of the seven hills landfill site and the surrounding environment of the earth’s surface temperature,and the different years and different land cover and different seasons(summer and winter)under the condition of the surface temperature has carried on the contrast analysis,This paper discusses the variation of landfills surface temperature and its comparison with surrounding surface temperature,which provides a theoretical basis for large area identification of landfills.(2)Large area identification of landfill sites based on Sentinel-2A.Taking the whole city of Suzhou as the study area and Qizishan landfill site and Zhangjiagang landfill site as the sample area,the spatial distribution of landfills in the whole city of Suzhou was identified by decision tree method.(3)Research on fine extraction of landfill site based on GF-1.Due to the large area identification error of landfill based on Sentinel-2A data,taking Zhangjiagang domestic waste landfill site as an example,the separability between the film-covered area and other features of the landfill site was calculated based on the feature set of spectral features,shape features and texture features to select the classification features.The MSW extraction index was constructed to increase the separability of the marsh and MSW with low separability,and the random forest classifier was selected to extract the MSW boundary accurately.