Application of red edge band in remote sensing extraction of surface water body: a case study based on GF-6 WFV data in arid area

Abstract

This paper mainly researches the application method of red edge band in water body remote sensing extraction. Gaofen-6 (GF-6) WFV data were chosen for the high spatial resolution, more bands, and wide width. Two new methods were proposed: the single-band threshold method based on the red edge 2 band and the decision tree model method based on the combined operation of the green band, red band, near infrared band, red edge 1 band, and red edge 2 band. Four traditional methods were used for comparing the extraction accuracy. Two study areas with different characteristics were chosen to analyze the reliability of the proposed method, one mountainous and one urban region, both located in Minqin, Gansu, China, which is a typical arid area. The results showed that the two red edge bands of the GF-6 WFV data can be utilized to extract water body information. Kappa coefficients extracted from the single-band threshold method based on the red edge 2 band in water bodies in mountainous and urban areas reached 96.18% and 93.21%, respectively. The decision tree method has the best extraction effect. Kappa coefficients of this method in mountain and urban water bodies were 97.73% and 94.41%, respectively. HIGHLIGHTS The red edge band of Chinese Gaofen-6 WFV date was found to be applicable to the extraction of surface water.; A decision tree model based on the joint operation of five bands was established, which can effectively extract surface water.;

Keywords

• decision tree
• gf-6 wfv data
• red edge band
• surface water body extraction