Phenology and Magnitude of Primary Production in the Eastern Bering Sea

Jens Nielsen1, Calvin W. Mordy2, Michael W Lomas3, Lisa B Eisner1 and Phyllis J Stabeno4, (1)NOAA - Alaska Fisheries Science Center, Seattle, WA, United States, (2)Joint Institute for the Study of the Atmosphere and Ocean, Seattle, WA, United States, (3)Bigelow Lab for Ocean Sciences, East Boothbay, ME, United States, (4)NOAA Pacific Marine Environmental Laboratory, Seattle, WA, United States
In sub-arctic systems, such as the Bering Sea, the timing and magnitude of phytoplankton production have high and long lasting effects on annual biological production and ecosystem functioning. Here, we use a combination of satellite, in-situ and high frequency mooring measurements to compare estimates of gross primary production (GPP), net primary production (NPP) and net carbon production in the Eastern Bering Sea. In-situ measurements from cruises and moorings provide detailed information on water column integrated primary production and incorporate subsurface maxima but are spatially restricted. Production estimates from satellite data provide good spatial coverage but these products are limited to measurements within the surface ocean and also have missing data due to ice and cloud cover. Combining and contrasting methods allow pinpointing advantages and limitations of each method and thus is a good way to better capture production dynamics. High frequency dissolved oxygen data from the surface to ~50 meter depth from a monitoring station on the Bering Sea middle shelf were used to estimate GPP, while satellite data provided estimates of NPP. Specifically, our aim was i) to compare production estimates from in-situ chlorophyll-a and dissolved oxygen data with satellite-derived production data and ii) use these estimates to better understand inter-annual changes of the phenology and magnitude of primary production. Our analyses showed that estimates of peak timing of GPP and NPP estimates based on in-situ dissolved oxygen and satellite-derived production, respectively, were similar, while estimates of annual production between the two varied. By using multiple data sources to estimate production, we were able to better evaluate uncertainties and correct for limitations. Our analyses with in-situ and satellite data help establish how measurements at various spatial and temporal scales are best integrated, and allow better predictions of primary production dynamics.