| Urban water environment is an important part of urban ecological environment.However,the pollution of urban rivers is becoming more and more serious,and many urban rivers have perennial or seasonal black and odorous water.Remote sensing monitoring of black and odorous water is an important method for understanding the current status of urban water quality,and comprehensively evaluating the effect of urban water environment treatment.This study takes the built-up area of Nanjing,Changzhou,Wuxi,and Yangzhou cities as the research areas,to clarified the optical characteristics of black and odorous water corresponding to different water color categories,data of field experiments and indoor water quality parameter were analyzed,combined with water color classification.Therefore,according to the differences in remote sensing reflectance of various types of water bodies,the remote sensing classification model of black and odorous water based on a decision tree was constructed and also applied to high spatial resolution images.Meanwhile,in order to reduce the impact of the surrounding environment of urban rivers on remote sensing monitoring results,the adjacency effect correction method based on the estimation of the background reflectivity of the point spread function model was established,so as to improve the accuracy of extracting remote sensing reflectance of water bodies from remote sensing images.Applying the model to high-resolution images in Yangzhou city can reveal the temporal and spatial distribution changes of water bodies with different degrees of black and odorous,which also can provide scientific basis and technical support for the monitoring,treatment and evaluation of urban black and odorous water bodies.This paper mainly gets the following conclusions:（1）Adjacency effect correction method for high-resolution images of urban riversBy improving the existing methods,a adjacency effect correction method based on the point spread function for estimating the reflectance of background pixels of the high-resolution image is established.After the adjacency effect is corrected,the"flying high"phenomenon of the water pixel in the near-infrared band is improved.The correction values of the four bands are close to the 1:1 line with the measured value,and the near-infrared band has the best correction effect.（2）Combining the relationship between the color of the water body and the degree of black and odorous,the water color classification rules have been establishedThe water color grading table is established by analyzing the water color of all sample data and referring to the international standard Laura color chart.The water bodies are divided into six types according to their chromaticity:type 1 water（gray black）,type 2 water（dark gray）,type 3 water（gray）,type 4 water（light gray）,type 5water（green series）and type 6 water（yellow series）.Among them,types 1-4 are gray water bodies,all of which are black and odorous water.Type 1 is severe black and odorous water,types 2-4 are mildly black and odorous water,and types 5-6 are general water.（3）Urban black and odorous water classification model based on decision treeCombining the relationship between the water color classification and the degree of black and odorous,the water quality parameters and the contribution rate of water components of water with different degrees of black and odorous are analyzed.After the classification of the water chromaticity classification table,the parameters of different water have obvious differences.A decision tree model was constructed based on the spectral differences of six types of water.1)DBWI（Difference Back and odorous Water Index）,that is,the difference in reflectance between blue and green wavelengths,is used to distinguish type 1 water and other types water.2)The three-band area water index G-R-NIR AWI（Green-Red-Near infrared Area Water Index）,that is,the triangular area enclosed by the reflectance spectrum at 542,631and 813 nm of the reflectance spectrum of different types of water,is used to distinguish the water dominated by pigment particles（type 2 and 5 water bodies）whose value of G-R-NIR AWI is greater than 0 nm·sr-1,and the water dominated by non-pigment particles（type 3,4 and 6 water bodies）whose value of G-R-NIR AWI is less than 0 nm·sr-1;3)The green band reflectance（Rrs Green）is used to distinguish type2 water and type 5 water;4)NDBWI（Normalized Difference Black and odorous Water Index）,that is,the ratio of the difference between the reflectance of the green band and the red band to the sum,is used to distinguish type 6 water from type 3 and4 water.After the adjacency effect of image were corrected,the threshold of each index need to be re-calibrated.The threshold of DBWI is 0.00088 sr-1,G-R-NIR AWI is 1.92 nm·sr-1,Green is 0.033 sr-1,and NDBWI is 0.032.There were 57 independent measured points are used to evaluate the accuracy of the decision tree model.The result shows that the overall recognition accuracy is75.40%,and the K value is 0.63,indicating that the data has high consistency and the model recognition accuracy is high.（4）Spatial distribution of black and odorous water bodies in Yangzhou City and evaluation of treatment effects of typical river sectionsThe model was applied to the Planet Scope image of Yangzhou City on April 17,2019,and the rivers in the built-up area were classified into black and odorous water.A total of 12 black and odorous river have been identified,with a total length of 47.03km.The black and odorous water are mainly distributed in narrow river sections and river sections flowing through factories or residential areas.Using satellite remote sensing data from 2017 to 2019,the monthly changes in the degree of black and odorous in the three river sections of Yangzhou（Two urban river sections that have been artificially governed,namely the Andun River and the ancient canal;and another suburban river section that has been slightly governed,namely Huangni ditch）were monitored to evaluate their governance effects.Two urban river sections that have been artificially governed,namely the Andun River and the ancient canal and one suburban river section that has been slightly governed,namely Huangni Ditch.The results show that the water quality of the Andun River and the Ancient Canal in the urban area has been improved after long-term dredging and sluice reconstruction,and there has been no rebound in the short term,almost fundamentally controlling the black and odorous water.However,the treatment of Huangni Ditch in the suburban rivers is not enough.It only takes the hardening of the river and the slope shaping,without dredging the sediment,and the effect of improving the black and odorous phenomenon is not obvious. |
PDF Full Text Request