Retrieval of phytoplankton biomass from simultaneous inversion of reflectance, the diffuse attenuation coefficient, and Sun-induced fluorescence in coastal waters

Abstract

[1] A model has been developed to retrieve phytoplankton absorption, a proxy for phytoplankton biomass, from observations of reflectance ( R) and the diffuse attenuation coefficient (K-d) collected by moored radiometers in coastal waters, where high concentrations of chromophoric dissolved organic matter ( CDOM) confound conventional ocean color algorithms. The inversion uses simultaneously two forward models: ( 1) a look-up table (LUT) that accounts for instrument geometry and the effect of the solar angle on both R and Kd and ( 2) an analytical function describing the effects of Sun-induced fluorescence (SIF) of chlorophyll on R. Estimates of phytoplankton biomass ( mostly from SIF), the absorption by colored matter, and the particulate backscattering coefficients ( mostly from the LUT) are obtained by optimizing the amplitude and shape of the absorption and backscattering coefficients in the forward model to best match the observations. An equation describing the quantum yield of fluorescence ( photons fluoresced/photons absorbed) as a function of incident irradiance constrains the model and allows estimates of phytoplankton absorption. Innovations include: the utilization of both R and K-d, providing good separation of the effects of backscattering from absorption; avoidance of reflectance bands between 400 and 600 nm, thereby avoiding interference from bottom reflection and CDOM fluorescence; utilization of the full emission band of SIF; and accounting for the irradiance-dependence of its quantum yield. The model effectively retrieved chlorophyll concentration from an independent data set - r = 0.76, n = 93, with a mean absolute percent error (MAPE) of 24%; this is better than a modern ocean color algorithm (OC4V4) on its validation data set, when restricted to the same range of chlorophyll ( r = 0.67, n = 384, MAPE = 51%).