Ocean Color Algorithms for Estimating Chlorophylla, CDOM Absorption, and Particle Backscattering in the Arctic Ocean

As the Arctic Ocean (AO) is increasingly altered by anthropogenic climate change, it is critical that we accurately assess ongoing changes in its capacity to support marine life. Ocean color remote sensing provides an effective tool to estimate phytoplankton biomass and net primary production in the remote and undersampled AO. However, standard algorithms have been parameterized using global data sets that severely underrepresent the AO. Because the optical properties of AO waters differ markedly from those of lower-latitude waters, standard algorithms perform poorly in the AO. Here, we use the largest bio-optical database ever assembled for AO waters to examine seasonal and regional patterns of AO optical properties and to quantify their impact on ocean color algorithms. We find that the standard algorithms fail to accurately retrieve the primary photosynthetic pigment chlorophylla(Chla) in the AO; it is overestimated at low phytoplankton biomass due to exceptionally high absorption of colored dissolved organic matter (CDOM), particularly on the interior shelves, and underestimated at high biomass due to the photoacclimation of phytoplankton growing in low light leading to high pigment packaging effect. Using this large bio-optical database that includes in situ measurements from across the AO and through the entire growing season, we parameterize a new empirical (AOReg.emp) and semianalytical (AO.GSM) algorithm which represent the unique bio-optical properties of the AO to successfully retrieve Chla, CDOM absorption, and particle backscattering with better accuracy than any algorithm yet applied to the AO.