Remote Sensing Recognition Of Black And Odorous Water Bodies Based On Landsat-8 Images

With the rapid development of urbanization in the country,the phenomenon of black and odorous water bodies caused by excessive discharge of industrial,agricultural and domestic wastewater has become increasingly serious.Black and odor water not only destroys the river ecosystem,but also affects the safety of drinking water and agricultural products,and ultimately threatens human health.Monitoring the spatial distribution of black and odorous water bodies is a prerequisite for governance.Compared with traditional monitoring technologies,remote sensing technology has the advantages of wide coverage,good prescription,a large amount of information,and less impact on the geographical environment.Using remote sensing technology to identify urban black and odorous water bodies,and through long-term monitoring,we can obtain the distribution and change rules of black and odorous water bodies,and evaluate the effectiveness of their treatment.This is of special significance for the scientific treatment of black and odorous water bodies.This paper takes the main rivers and lakes in the main urban area of Changsha City as the research object.Based on the Landsat-8 SR multispectral data,combined with the water spectrum classification,supervised classification and unsupervised classification methods,the black and smelly waters are classified and identified.The main research work and results of this paper are summarized as follows:(1)According to the spectral characteristics of different water bodies and based on the similarity measurement of the spectral vectors,the spectral vectors of the water bodies in the study area are divided into four categories:non-black and odorous class 1 and class 2,and black and odorous class 1 and class 2.(2)Establish two template samples based on the water spectrum classification,namely template 1:black and odorous water body and non-black and odorous water body;template 2:non-black and odorous water body,nonblack and odorous type 2,black and odorous type 1,and black and odorous water body 2.class.Three common supervised classification methods,maximum likelihood method,minimum distance method,and BP neural network method,and two common unsupervised classification methods,ISODATA method and K-means method,are used to classify water bodies.The BP nerve of template 2 is found through accuracy verification.The overall accuracy of the network method classification results is 94.87%,the Kappa coefficient is 0.93,and the overall classification accuracy is the highest.(3)The three-phase Landsat-8 SR data from 2013 to 2019 was used to classify water bodies using the BP neural network method,and the temporal and spatial evolution characteristics of the black and odorous water bodies in Changsha were found to be as follows:1)From 2013 to 2016,the overall area of black and smelly water bodies did not change much;from 2016 to 2019,a large number of black and smelly water bodies turned into non-black and smelly water bodies.2)Small lakes are more prone to black and odor than large lakes.The downstream and narrow areas of the river are prone to black and odor.