To identify and quantify the principal processes that control the partitioning of carbon among oceanic reservoirs and between the ocean and atmosphere on focal and regional scales, with a view towards synthesis and prediction on a global scale, is a specific goal of the U.S. JGOFS Synthesis and Modeling Project. As a contribution towards achieving this goal, Drs. Barber, Peng, Chai, Dugdale and Wilkerson will develop an ecosystem model for the equatorial Pacific Ocean, with a focus on how silicate and iron affect new and export productivity and the partitioning of carbon between the atmosphere, surface ocean and deep ocean. The study will use an ecosystem model embedded in a state-of-the-art general circulation model for the equatorial Pacific Ocean to investigate how new and export productivity responds to changing physical and chemical forcing. The domain of the model is between 30S and 30N, 120E and 70W, with real geometry and topography, but analysis will focus on the equatorial region from 5N to 5S. The recent upgrade of supercomputers at North Carolina Supercomputing Center (NCSC) (CrayT90) and Arctic Region Supercomputing Center (ARSC) (Cray-YMP) and the award of several hundred hours of CPU time to Peng, Chai and Barber make it possible to embed an ecosystem model with modest complexity in a high resolution, three dimensional prognostic ocean model, and to conduct numerous experiments on the ecosystem model structure and parameters in a timely and efficient manner.

Phase 1 of the project will modify an existing five-compartment ecosystem model by adding three more compartments (silicate, diatoms and mesozooplanktonic grazers) following the approach of Dugdale et al. The preliminary objective of this three-dimensional Si/N/light model is to reproduce High Nitrate-Low Silicate-Low Chlorophyll (HNLSLC) conditions. With size-dependent growth rate responses in small phytoplankton and diatoms and varying grazing vulnerability, the role of new diatom production regulating on Si and Fe can be thoroughly investigated. Also in Phase 1, TCO2 and total alkalinity (ALK) will be added in order to calculate pCO2. The pre-industrial atmospheric CO2 (280 ppm) will be used to hindcast air-sea flux of CO2 in the equatorial Pacific. New production regulating on silicate should provide a more accurate calculation of CO2 compared to using nitrate as a regulating nutrient.

In Phase 2 the effect of iron is added to the model making a, the initial slope of the photosynthesis vs. irradiance curve, a function of iron. The values of a are based on equatorial observations of natural and experimental iron additions. Independently, Ks for Si(OH4) is made a function of iron, an effect that involves only diatoms. The `balance to bloom` transition will be simulated with the two iron effects to reproduce the IronEx 1 and 2 phytoplankton responses to a transient iron addition. This modeling study will provide estimates of new and export productivity, and a formal description of Si and Fe as regulating mechanisms in the equatorial Pacific Ocean. When new and export productivity is modeled accurately and validated with JGOFS studies, it will possible to predict with increased confidence how climate change may alter, via biogenic export, maintenance of the air-sea dpCO2 and hence the ocean's uptake and release of CO2.