Optimizing models for remotely estimating primary production in Antarctic coastal waters

Primary productivity and associated biogeochemical fluxes within the Southe rn Ocean are globally significant, sensitive to change and poorly known com pared to temperate marine ecosystems. We present seasonal time series data of chlorophyll a, primary productivity and in-water irradiance measured in the coastal waters of the Western Antarctica Peninsula and build upon exist ing models to provide a more optimum parameterization for the estimation of primary productivity in Antarctic coastal waters. These and other data pro vide strong evidence that bio-optical characteristics and phytoplankton pro ductivity in Antarctic waters an different from temperate waters. For these waters we show that over 60% of the variability in primary production can be explained by the surface chlorophyll a concentration alone, a characteri stic, which lends itself to remote sensing models. if chlorophyll a concent rations are accurately determined, then the largest source of error 13-18) results from estimates of the photoadaptive variable (P-opt(B)). Further, t he overall magnitude of P-opt(B) is low (median 1.09 mg C mg chl(-1) h(-1)) for these data compared to other regions and generally fits that expected for a cold water system. However, the variability of P-opt(B) over the cour se of a season (0.4 to 3 mg C mg chl(-1) h(-1)) is not consistently correla ted with other possible environmental parameters, such as chlorophyll, sea surface temperature, incident irradiance, day length, salinity, or taxonomi c composition. Nonetheless, by tuning a standard depth- integrated primary productivity model to fit representative P-opt(B) values and the relatively uniform chlorophyll-normalized production profile found in these waters, w e can improve the model to account for approximately 72-73% variability in primary production both for our data as well as for independent historic An tarctic data.