Study On Urban Black-odor Water Identification Model Based On Gaofen Multispectral Images

In recent years,with the rapid development of urbanization,a large number of domestic sewage and industrial wastewater discharge to the urban river,resulting in serious organic pollution.The river self-purification ability is insufficient,and the water quality of urban rivers is deteriorating.At present,in most urban areas of China,the urban rivers have been polluted in varying degrees,and the urban black and odorous water is especially prominent.With the development of remote sensing technology,its advantages of convenience,efficiency,continuity and coverage gradually make up for the deficiency of traditional monitoring methods of black and odorous water.By using satellite remote sensing to identify the black and odorous water in the city,the distribution and variation of black and odorous bodies can be obtained through long-term monitoring,so as to evaluate its effectiveness.The rapid development of Gaofen satellite(China’s high resolution satellite)has made it possible to monitor black and odorous waters by remote sensing,which is of great significance to national regulation and evaluation of black and odorous water.This paper took seven cities with typical black and odorous water as the study area,design and carry out experiments in the field for many times,and obtain a lot of black and odorous water quality parameters,the apparent optical properties as well as the inherent optical properties.From the perspective of remote sensing operation to monitor the black water body of the city,the remote sensing classification of the urban black and odorous water was carried out.On the basis of classification result and spectral characteristics of black and odorous water.,this paper analyzes the limitation of using GF multispectral satellite to identify this water,and establishes BOI index to identify the black and odorous water in the city.A accurate pretreatment process of GF multispectral data was proposed for business requirement of the identification of black and odorous water.And taking the key areas of Shenyang city and Beijing city,the application of GF multispectral data pretreatment process and the identification of BOI model in the urban black and odorous water body have realized the monitoring of the urban black and odorous water in the regions from 2015 to 2017.The main achievements and results of this paper are:1)A remote sensing classification system of the urban water was established,including two types of general water,three types of mild black and odorous water and three types of heavy black and odorous water.The spectral characteristics of eight types water were analyzed by classification system,and two types water had the spectral characteristics of clean water and eutrophic water respectively.The mild and heavy black-odor water which the reflectance characteristics are similar to ordinary water has the spectral characteristics of eutrophic water.The high reflectivity of the mild and heavy black-odor water has the spectral characteristics turbid water.The extremely low reflectivity of the mild and heavy black-odor water has special spectral characteristics,namely,no obvious spectral characteristics of water.2)In order to realize the remote sensing identification of urban black-odor water bodies,urban water classification based on GF multispectrum was carried out according to urban water classification based on the remote sensing classification system,including general water and two types of black and odorous water.The limitations of GF multispectral satellite identification black and odorous water: due to factors such as GF multispectral satellite broadband sensor and band setting,it is difficult to distinguish between heavy black and odorous water from mild black and odorous water,and it is difficult to distinguish between general water and black and odorous water which is similar to eutrophic water.3)Using the measured remote sensing reflectance data simulate GF multispectral data and comparing the general water and black-odor water spectral characteristics,we found that the reflectance change of black and odorous water in the red-green band is more gradual than that of general water,and proposed a normalized ratio model of BOI(Black and Odorous water Index)model to identify black-odor water,and identification accuracy is higher.4)GF multispectral data is affected by the data itself.There are some problems in practical application.This paper proposes a pre-processing method for GF data,including restoration fusion,geometric correction,and atmospheric correction.Applying the BOI model to the pre-processed image of the key regions of Shenyang City and Beijing City from 2015 to 2017,the trend of black and odorous water in two cities is decreasing year by year.This has a direct relationship with government regulation and supervision.This paper on the basis of water color remote sensing classification of knowledge,building a remote sensing classification system of urban water,and on the basis of multispectral images is proposed black-odor water body identification technology process based on GF,monitoring for effective regulation black-odor water to provide technical support.