Remote sensing method for detecting green tide using HJ-CCD top-of-atmosphere reflectance

• Published in: International journal of applied earth observation and geoinformation 2021-10, Vol.102, p.102371, Article 102371
• Main Authors: Zhang, Hailong, Yuan, Yibo, Xu, Yongjiu, Shen, Xiaojing, Sun, Deyong, Qiu, Zhongfeng, Wang, Shengqiang, He, Yijun
• Format: Article
• Language: English
• Subjects: Macroalgae bloom; TCG; TCT-like; Top-of-atmosphere reflectance; Ulva prolifera
• ISSN: 1569-8432; 1872-826X
• DOI: 10.1016/j.jag.2021.102371

•A simple TCG index was designed to detect green tide directly using satellite Rtoa.•New GMB-specific TCT-like matrix was obtained for four-band satellite sensors.•TCG index performs without atmospheric correction and cloud and sun-glint mask.•TCG index is less sensitive to observing conditions and "see" through thin clouds. Accurate and rapid monitoring of green macroalgae blooms (GMB; also known as green tide) provide valuable information in decision making for mitigating its environmental impacts and economic losses. Plenty of spectral-based algorithms mainly rely on Rayleigh-corrected or atmospherically-corrected reflectance. However, this inherently requires additional information beyond satellite observations for atmospheric correction. This work developed a new method for detecting GMB just by using satellite top-of-atmosphere reflectance (Rtoa) without any atmospheric correction scheme based on Tasseled-cap-like transformation (TCT-like). Here we conducted a case study on Chinese four-band HJ-CCD sensor for illustrating the reliability and sensitivity of this method for detecting GMB of Ulva prolifera. The performance of the method was validated statistically via cross-sensor and cross-index comparisons with existing algorithms from atmospherically-corrected reflectance. Results showed that the Rtoa-based method can effectively detect GMB with high accuracy, and achieved high robustness against observing conditions such as various background water conditions, thin cloud cover and sun-glint. In addition to the findings presented here, this study may provide a promising way to detect massive macroalgae blooms of other species using optical satellite data.