Comparison of an Absorption-based and a Backscattering-based Algorithm for the Retrieval of the Particle Size Distribution

Ekaterina Kochetkova, University of Pennsylvania, Department of Earth and Environmental Science, Philadelphia, PA, United States, Tihomir S Kostadinov, California State University San Marcos, San Marcos, CA, United States, Shovonlal Roy, University of Reading, Geography and Environmental Science, Reading, United Kingdom and Irina Marinov, University of Pennsylvania, Department of Earth and Environmental Sciences, Philadelphia, PA, United States
The size structure of marine phytoplankton communities is a key characteristic that influences ecosystem functioning and the system response to climate change. Recently, satellite ocean color algorithms have been developed to retrieve the particle size distribution (PSD); the PSD quantifies the size structure of surface oceanic waters and can be used to define size-based phytoplankton functional types (PFTs). The PFTs are defined here as diameters of 0.5-2 𝜇m, 2-20 𝜇m, and 20-50 𝜇m for pico-, nano- and microplankton, correspondingly. We use monthly SeaWiFS data from 2003 to 2007 to compare the power-law PSD and PFT abundances obtained with two independent PSD/PFT algorithms, namely KSM09 (Kostadinov et al.,2009) and ROY13 (Roy et al.,2013). The algorithms retrieve the PSD slope and scaling factor, given spectral bbp(𝜆) or achl(𝜆=676), respectively. A fundamental difference between these algorithms is that KSM09 is sensitive to all backscattering particles (and interpreted as PFTs under certain assumptions), whereas ROY13 is phytoplankton-specific. We hypothesize that the KMS09 PSD represents the total particle assemblage, and the ROY13 PSD represents just the phytoplankton fraction. Indeed, generally KSM09 total abundances are higher than ROY13, and the ratio of ROY13 to KSM09 particle concentrations indicates that the oligotrophic subtropical gyres are dominated by phytoplankton, whereas in the higher latitudes and coastal areas phytoplankton composes a smaller fraction of the total particle count. However, discrepancies and inconsistencies remain and are likely due to the treatment of the index of refraction in KSM09 among other assumptions. We also discuss and compare the retrieved fractional contributions of the 3 PFT groups across the two algorithms in major ocean regions. We will discuss how using these two different PFT products will ultimately impact net and export production in the ocean.