Open AccessEstimation of chlorophyll concentration in waters near Hokkaido using the linear combination method
Author(s) -
Chong Shi,
Teruyuki Nakajima
Publication year - 2017
Publication title -
optics express
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.394
H-Index - 271
ISSN - 1094-4087
DOI - 10.1364/oe.25.00a963
Subject(s) - rayleigh scattering , residual , remote sensing , radiative transfer , atmospheric correction , linear regression , environmental science , scattering , standard deviation , correlation coefficient , atmospheric radiative transfer codes , chlorophyll a , leaf area index , polarization (electrochemistry) , mean squared error , chlorophyll , inversion (geology) , optics , mathematics , physics , statistics , algorithm , geology , chemistry , reflectivity , ecology , biology , paleontology , biochemistry , organic chemistry , structural basin
An inversion algorithm is implemented to retrieve the surface chlorophyll a (Chl) concentration using satellite observation data from the MODIS instrument. The algorithm employs a simple and flexible index (LCI) to combine with the Chl without explicit correction for aerosol scattering. To investigate the sensitivity of LCI to Chl and other influence factors, an oceanic radiative transfer model coupled with a comprehensive bio-optical module is developed. It is studied that the LCI is significantly linear to Chl and not sensitive to other influence factors, except in very low oceanic salinity or scattering angle conditions, where over a 12% relative difference of derived Chl exists. Inversion results show the retrieved Chl are highly consistent with the MODIS operational data products in waters near Hokkaido, with the correlation coefficient, root mean square deviation, and average percentage difference of 0.9702, 0.3756 mg m -3 , and 13.89%, respectively. Investigation of the validity of this algorithm with a variety of atmospheric conditions indicates that the residual influence of atmosphere on the LCI index, after Rayleigh scattering correction, is generally within ± 0.001, allowing the retrieval error of Chl at less than 25% in most cases. A good comparison between retrieval and in situ measurements is also identified and implies that the retrieval accuracy via the LCI method depends on the linear combination coefficients used and the bio-optical module selected, while effects of polarization can be ignored.