Particle Size Distribution and Size-partitioned Phytoplankton Carbon Using a Two-Component Coated-Spheres Bio-optical Model: Monthly Global 4 km Imagery Based on the OC-CCI v5.0 Merged Ocean Color Satellite Data Set

Monthly global 4km satellite products spanning September 1997 to December 2020. The data contains Particle Size Distribution (PSD) parameters of an assumed power-law PSD, absolute and fractional size-partitioned phytoplankton carbon and associated variables such as particulate organic carbon (POC) and Chlorophyll-a as derived from the PSD algorithm. The retrieval is based on a backscattering bio-optical model using two particle populations and coated spheres for phytoplankton inherent optical properties (IOP) modeling, and a retrieval using spectral angle mapping (SAM - where satellite spectra are classified using a comparison to a collection of modeled end-member spectra, by treating spectra as vectors and using their dot product). Partial uncertainties are given as standard deviation and are estimated using a combination of Monte Carlo simulations and analytical error propagation. An empirical tuning factor is given for attaining more realistic estimated model concentrations of POC and Chlorophyll-a. The tuning factor is multiplicative, to be applied in linear space. This tuning factor has not been applied to the monthly data, users can choose whether or not to apply it to absolute carbon and Chlorophyll-a concentrations. The factor does not affect retrievals of fractional contributions of phytoplankton size classes to total phytoplankton carbon. Monthly climatologies files and an overall climatology file are also provided, and in those files, both untuned (tuning factor not applied) and tuned (tuning factor applied) variables are provided, for user convenience. Input remote-sensing reflectance data are v5.0 of the Ocean Colour -Climate Change Initiative (OC-CCI) of the European Space Agency. The OC-CCI general reference is Sathyendranath et al. (2019; doi:10.3390/s19194285), and for v5.0 of the dataset, the reference is Sathyendranath et al. (2021; doi:10.5285/1dbe7a109c0244aaad713e078fd3059a). More detailed metadata, including geospatial metadata, are given in the netCDF files. Variable names should be self-explanatory. Quick browse images are provided as well. Coastlines in these quick browse images are from v2.3.7 of the GSHHS data set - see Wessel and Smith (1996) (doi:10.1029/96JB00104). Modeling and data processing was done in MATLAB ®.