A new semi-analytical model for retrieving chlorophyll a concentration and magnitude of fluorescence in the South China Sea

Abstract

The red band signal of the color spectrum is of significant interest because its remote sensing reflectance yields valuable information on phytoplankton. In this study, we combined a fluorescence model with a modified version of the Garver-Siegel-Maritorena (GSM) algorithm to retrieve the chlorophyll a concentration (Chla) and the magnitude of fluorescence in coastal waters. An analytical inverse optimization procedure was performed using the measured reflectance over the whole visible spectrum with a focus on the Ocean and Land Color Instrument (OLCI) sensor (from 413 to 709 nm). The coupled visible-infrared model (GSM&Fluo) was validated by a comparison with Chla that was measured in situ in a wide variety of productive coastal water conditions in the South China Sea. The coefficients of determination (r2) were 0.79 and 0.88 for the GSM and our algorithm, respectively. The combined approach produced more accurate results in our study location. The coupled GSM&Fluo model was able to avoid being trapped in a local optimum by adding an estimation of the fluorescence signal. The results also showed that the suspended particulate matter (SPM) played a major role in the magnitude of the reflectance peak in the chlorophyll fluorescence emission band, and the Fluorescence Line Height (FLH) calculated with a simple baseline approach tended to overestimate the true magnitude of the fluorescence emission as SPM increased.