Jumbo Content

ROSES Proposals

Improved Satellite Ocean Color Retrievals of Ocean Inherent Optical Properties and Biogeochemical Properties Utilizing the Capabilities of PACE

PI: Greg Mitchell - Scripps Institution of Oceanography
To support algorithm development for PACE, we propose to use a globally diverse and detailed optical and biogeochemical data set to develop new parameterizations of absorption and scattering, to advance our understanding of the scattering phase function, and to assess closure of forward and inverse model predictions compared to measured optical and biogeochemical variables. Our data spans the range of spectral relevance for PACE for ocean retrievals (350-750 nm) and includes inherent and apparent optical properties (IOP, AOP) and the most fundamental biogeochemical parameters of interest for remote sensing of ocean ecology. Contemporary satellite retrievals of several IOP variables (phytoplankton absorption, the sum of detrital and correlated spectral channels on orbit. These limitations restrict retrievals of IOPs at few wavelengths and limit our ability to accurately estimate chlorophyll a, particulate organic carbon, and adg, defined as the sum of soluble absorption and detrital particle absorption (CHLA, POC and adg). Using our global data set, we evaluate in this proposal the performance of current empirical and inverse algorithms demonstrating important limitations that can be greatly improved by the work we propose. Combining our uniquely detailed and global data set with UV-Vis numerical modeling in Hydrolight we propose to develop new parameterizations of relationships between the most important ecological variables that govern upper ocean IOP and AOP over the extended spectral range of PACE (350-750 nm). Due to the improved spectral range and resolution of PACE, we envision an ability to broaden the number of biogeochemical constituents to also include UV-absorbing mycosporine amino acids (MAA), phycobiliproteins (PBP) and particle size distribution (PSD ) that are needed to specify phytoplankton functional groups and plankton ecosystem structure. PSD data will be based on our global observations of Coulter Counter size distributions and flow cytometer (FCYT) analysis. FCYT data will provide important details of phytoplankton functional groups and size distributions, 2 µm. We will analyze our liquid nitrogen archived samples for FCYT, MAA and PBP collected over the past 15 years and integrate these new analytical results to our uniquely detailed and global data base to allow us to pursue our proposed goals. Our goal will be to develop forward models of IOP and AOP as governed by CHLA, POC, adg, MAA, PBP and PSD. We will implement a system of optimization combining Hydrolight code extended to the UV, and our large, detailed and globally distributed measurements, to develop forward radiative transfer models dependent on the expanded set of biogeochemical variables. We will use our optimized data to create synthesized ocean spectral reflectance 350-750 nm that will be combined with our expanded set of biogeochemical observations to develop novel inverse algorithms for PACE (MAA, PBP, Nd), spectral values of IOPs, and more robust algorithms for the heritage retrievals (CHLA, POC, adg). We will also utilize our hyperspectral global ocean phytoplankton absorption, HPLC pigments, MAA, PBP and Nd data to develop improved estimates of phytoplankton functional groups. As we did for SeaWiFS (SeaBAM; O'Reilly et al. 1998), as participants in the PACE Science Team we will evaluate our data, models and satellite algorithms in a collaborative way to contribute to the community goal of robust consensus algorithms for a dramatically expanded set of biogeochemical variables that will be enabled by the capabilities of PACE.