摘要
This study aims to develop new algorithms to retrieve sea surface parameters including concentrations of Chlorophyll a （Chl a） and Suspended Particulate Matter （SPM）, and absorbance of Colored Dissolved Organic Matter （aCDOM） by incorporating the contribution of red bands to make them more adaptable to case 2 waters. Optical remote sensing algorithms have demonstrated efficient retrieval of Chl a, SPM, and aCDOM, yet they are not very accurate especially for coastal areas. It has also been found that the default algorithm has overestimated Chl a in the Pearl River Estuary, and shown poor correlation for CDOM absorbance. By incorporating the red band ratios into the algorithm, a correction effect has been shown, which improves the accuracy of quantifying the actual concentration. Modeling and data fitting of the algorithm have been done based on 61 data samples collected in the Pearl River estuary during a cruise from 3 to 11 May 2014. The study also attempts to modify the aerosol correction bands used in SeaDAS to prevent saturation of these bands. The modified algorithms showed an R-Square value of 0.7289 for Chl a fitting, and 0.7338 for CDOM fitting, and corrected overestimation of Chl a concentration in the Pearl River estuary.
This study aims to develop new algorithms to retrieve sea surface parameters including concentrations of Chlorophyll a （Chl a） and Suspended Particulate Matter （SPM）, and absorbance of Colored Dissolved Organic Matter （aCDOM） by incorporating the contribution of red bands to make them more adaptable to case 2 waters. Optical remote sensing algorithms have demonstrated efficient retrieval of Chl a, SPM, and aCDOM, yet they are not very accurate especially for coastal areas. It has also been found that the default algorithm has overestimated Chl a in the Pearl River Estuary, and shown poor correlation for CDOM absorbance. By incorporating the red band ratios into the algorithm, a correction effect has been shown, which improves the accuracy of quantifying the actual concentration. Modeling and data fitting of the algorithm have been done based on 61 data samples collected in the Pearl River estuary during a cruise from 3 to 11 May 2014. The study also attempts to modify the aerosol correction bands used in SeaDAS to prevent saturation of these bands. The modified algorithms showed an R-Square value of 0.7289 for Chl a fitting, and 0.7338 for CDOM fitting, and corrected overestimation of Chl a concentration in the Pearl River estuary.
基金
This work is supported by the Hong Kong Innovation and Technology Fund under grants of ITS/272/11 and ITS/259/12, the General Research Fund of Hong Kong Research Grants Council （RGC） under grants CUHK 402912 and 403113, the National Natural Science Foundation of China （Grant No. 41376035）, and the direct grants of the Chinese University ofHong Kong. The authors are grateful to Dr. Chunyan Shen, who provided with substantial supports to the in-sire data collection.
