The ocean currently absorbs about one-third of the CO₂ emitted by human activity. Understanding how ocean uptake of CO₂ will be modified in the upcoming century is an important contribution for reducing uncertainties in projections of future climate. The supply of soluble iron to the oceans is fundamental to oceanic primary production and CO₂ uptake. Iron deposition has the largest effect in high-nutrient low- chlorophyll (HNLC) areas that comprise approximately 30% of the world's oceans. The vast majority of the global ocean, including the HNLC areas, is dependent on wind-transported atmospheric dust as a source of iron. Ocean productivity depends specifically upon soluble or bioavailable iron. The fraction of soluble iron from soil particles is low but much higher solubility has been reported for atmospheric aerosols, suggesting aerosol modification downwind of the source. The controls on aerosol iron solubility include photochemistry and acidity, particularly during cloud processing, where aerosols can be coated with an acidic film of sulfates or nitrates. The fraction of atmospheric soluble iron also depends upon the mineral content of the source region along with the presence of other aerosol species. For example, iron from the combustion of fossil fuel, incinerator fires, and biomass burning can account for up to half of iron deposition in some regions. Emission of aerosol precursors (SO₂ and NO_X) and organic matter from natural sources and anthropogenic sources may also influence atmospheric iron cycling.

In this project, scientists at Columbia University and Cornell University will develop and implement an atmospheric iron lifecycle module into two widely used Community Earth System models (NCAR CESM and NASA GISS ModelE) to provide more realistic deposition fluxes of soluble iron into the ocean. Soluble iron can be introduced at only modest cost in terms of both personnel and computational expenses, because our project efficiently uses existing aspects of both models, including prognostic cycles of separate dust mineral types and combustion aerosols, along with heterogeneous uptake of sulfate and nitrate aerosols. Many Earth System models assume a constant fraction of soluble iron from either prognostic or prescribed desert dust aerosol deposition. This is in spite of the fact that the soluble iron fraction is generally observed to increase downwind of sources due to atmospheric processing. The project will provide more realistic fluxes of soluble iron that impact ocean productivity and the carbon cycle through uptake of atmospheric CO₂ . These fluxes can be used by the ocean biochemistry models, allowing the calculation of the sensitivity of the carbon cycle to more realistic atmosphere iron processing.

The models will include dust and combustion sources of iron and will be evaluated and constrained using a worldwide set of observations. Among other issues, the project will improve our understanding of the relative role of natural and anthropogenic sources in the deposition of soluble iron and of the different mechanisms involved in the atmospheric processing of iron. We will investigate whether increased mechanistic complexity leads to model improvements that are distinguished by the data or whether more efficient empirical schemes are sufficient. Regional variations of the soluble fraction of iron deposition for different climate periods and emission scenarios will be compared to the common modeling assumption that this fraction is uniform. Given the existing uncertainties in the iron cycle, an important innovation in this proposal is to provide estimates of soluble iron from two publicly available Earth System models and to calculate to what extent the soluble iron deposition is consistent among the models. This will allow us to look at the uncertainties in the soluble iron budget in a much more rigorous and complete way than is possible by using one model only. Improvements in the soluble iron distribution will be used in separate projects to drive the coupled-carbon-climate model in the newest versions of the CESM and ModelE.