[Comprehensive Classification Method of Urban Water by Remote Sensing Based on High-Resolution Images].

2021 
Urban water is a significant part of the urban ecosystem. Therefore, a comprehensive evaluation method of the water environment was proposed based on domestic high-resolution images. The relationships between the spectral characteristics and water quality parameters of urban water were analyzed based on sampling in Nanjing, Wuxi, Changzhou, and Yangzhou from 2017 to 2019. An index named the U-FUI (urban Forel-Ule index) suitable for urban water based on GF-2 images was proposed to achieve the classification of urban water on the basis of the international standard chroma conversion model and the Forel-Ule index. Independent verification data showed that the recognition accuracy of the classification model could reach 72%. The results indicated that urban water can be classified into six classes from Ⅰ to Ⅵ, which represent water colors of blue, light green, dark green, yellow, yellowish brown, and dark grey, respectively, according to the U-FUI. Among them, the water quality of U-FUI Ⅰ water is good, but is rarely distributed in urban water. The concentrations of chlorophyll-a in U-FUI Ⅱ-Ⅲ water are higher than those of the other classes; the concentrations of total suspended solids, particularly inorganic suspended solids, of U-FUI Ⅳ-Ⅴ water are higher than those of the other classes; and the water quality of U-FUI Ⅵ water is poor and the water quality parameters are different from those of the other classes. Meanwhile, the method was successfully applied to the GF-2 image of Nanjing on April 9, 2018. The results showed that the urban water in Nanjing is mainly composed of U-FUI Ⅱ-Ⅳ water, whereas the distribution of U-FUI Ⅰ, Ⅴ, and Ⅵ water is lower in the city. The spatial distribution characteristics were consistent with the results of in-situ sampling in the same period.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    1
    Citations
    NaN
    KQI
    []