Using Particle Size Distribution and Habitat Models to Predict Phytoplankton Functional Types in the Atlantic Ocean

Christopher Brown, NOAA, Center for Satellite Applications and Research, College Park, MD, United States and Timothy Moore, The Joan and James Leitzel Center for Mathematics, Science, and Engineering Education, University of New Hampshire, Durham, NH, United States
Abstract:
Phytoplankton community composition in the oceans is a critical determinant to both marine ecosystem structure and biogeochemical processes. A model was constructed to predict the dominant of five phytoplankton functional types (PFT) in the surface layer of the Atlantic Ocean as a fraction of biomass using a combination of particle size distribution (PSD) retrieved from satellite ocean color radiometric measurements and empirical niche models developed for each PFT. Presence of dominant PFTs were developed from a pigment data set from the Atlantic Ocean and co-located with gridded satellite, modeled and climatological data. The PFT model employs PSD, sea-surface temperature, photosynthetically available radiation, mixed-layer depth, wind and nutrients as inputs, and successfully classified the dominant groups 85% using data from the training set and 70% using an independent data set. The predicted monthly distribution patterns of the dominant PFT in the Atlantic during the SeaWiFS era appear reasonable yet require additional validation.