Bio-optical and remote sensing investigation of phytoplankton community size structure

This dissertation examines the optical and ecological role of phytoplankton cell size in the context of chlorophyll concentration and the presence of other optically active constituents. The first study isolates the effect phytoplankton cell size has on varying spectral remote sensing reflectance (Rrs(lambda)) through the use of optical and radiative transfer models that are linked in an off-line diagnostic calculation to a global monthly numerical model. Two separate simulations of the optical model were run with cell size-dependent and cell size-independent phytoplankton absorption with all other optical constituents treated identically. The consideration of size-varying phytoplankton absorption resulted in more realistic estimates of remote sensing reflectance at 443 nm (Rrs(443)) and ocean color four-band ratio algorithm chlorophyll concentration ([Chl]). When not considering phytoplankton size effects, normalized Rrs(443) will be underestimated in both the Northern Hemisphere (NH) and Southern Hemisphere (SH) summer and overestimated in the winter months at mid to high latitudes.;Phytoplankton cell size distribution (Sf, percent microplankton) is estimated from satellite Rrs(lambda) imagery through implementation of a forward optical model look-up-table (LUT) that incorporates the range of absorption and scattering variability in the global ocean. Satellite imagery of [Chl] and absorption due to dissolved and detrital matter at 443 nm (a CDM/NAP(443)) are used to narrow the LUT search space, leaving R rs(lambda) to vary only due to Sf. SeaWiFS Rrs is matched with the closest LUT Rrs option and the associated S f is assigned. Thresholds for [Chl] and aCDM/NAP(443) are determined. Global monthly maps of Sf are generated. The spatial and temporal patterns of Sf that emerge validate well with in situ observations.;The global monthly maps of Sf for the first ten years of the SeaWiFS mission (September 1997--August 2007) are analyzed with singular value decomposition analysis. [Chl] and Sf can be decoupled spatially and temporally. The NH high latitudes and equator experience more variance due to Sf, while the NH subtropical and SH high latitudes experience greater variability due to [Chi]. A strong El Nino Southern Oscillation response in the equatorial Pacific is more apparent in Sf than [Chl].