Marine nitrogen fixers mediate a low latitude pathway for atmospheric CO2 drawdown: Biogeochemical and physical output.

The data included in this repository includes both the physical and biogeochemical fields that were generated by 89 simulations over 10,000 years with an Ocean General Circulation Model (OGCM) and attached ocean biogeochemistry: CSIRO Mk3L-COAL v1.0. The 89 simulations alter one of three major processes to understand the relationship between N2 fixation and CO2 drawdown from the atmosphere into the ocean. The three major processes are (1) different circulation states, (2) different representations of the marine nitrogen (N) cycle, and (3) different rates of aeolian iron (Fe) deposition to the ocean surface.

Four physical states were generated. Three were of preindustrial (PI; 1850 CE) climate and one was of the Last Glacial Maximum (22,000 BCE). The preindustrial states were generated by forcing the OGCM with the boundary conditions produced by piControl runs of CSIRO Mk3L v1.2, GFDL-ESM2G, and HadGEM2-CC. The Last Glacial Maximum state was generated by forcing the OGCM with the boundary conditions produced by a glacial run of CSIRO Mk3L v1.2 (Buchanan et al., 2016). These boundary conditions were sea surface temperature, salinity and the meridional and zonal components of surface wind stresses, and the final 10 years of these runs were averaged and regridded onto the CSIRO Mk3L-COAL grid space. Numerous alterations to the parameters governing the N cycle were undertaken.

These include:

Deactivating the N cycle and holding mean NO3 content stable

Increasing/decreasing the Fe half saturation coefficient for N2 fixers

Increasing/decreasing the PO4 half saturation coefficient for N2 fixers

Altering the C:P ratio of N2 fixers organic matter

Increasing/decreasing the base rate of sedimentary denitrification to alter mean ocean NO3

Iron deposition rates to the ocean surface vary between 25% and 2500% of the modern rate, which is taken to be the climatology produced by Mahowald et al. (2005). All deposition rates are made to be a factor of this modern field, expect for the two highest fields of 500% and 2500%, which represent the Last Glacial Maximum dust deposition field of Lambert et al.(2015) assuming 3.5% Fe content and 0.4% or 2% solubility.

These three major processes are altered in isolation and in combination to thoroughly elucidate the relationship between N2 fixation and the drawdown of CO2 into the ocean. Furthermore, some simulations were undertaken with an additional model development step, which involved allowing atmospheric CO2 to freely evolve as the ocean absorbed or release carbon. These are noted by having file names with "open CO2".

Climatologies of sea surface temperature, sea surface salinity, and x and y vectors of sea surface wind stresses were produced by both the PI and LGM coupled experiments and were used to force the ocean general circulation model. Additional climatologies of sea ice fractional cover, sea surface wind speeds, net incident short-wave radiation, and the aeolian deposition of iron and reactive nitrogen were important for forcing the biogeochemical model. These climatologies are made available here. Also available are the three-dimensional global annual averages of oceanic properties for all simulations at their steady-state solutions. These include the physical properties of temperature, salinity, ideal water mass age, velocities, overturning streamfunctions and the various fluxes of heat, salt and momentum at the surface. Three-dimensional (3D) fields of oxygen, apparent oxygen utilisation, dissolved inorganic carbon, total alkalinity, phosphate, nitrate, nitrate-15, and iron are provided by the biogeochemical model output.

The biogeochemical model also provides 3D fields of the sources and sinks of the N cycle (N2 fixation, Water column denitrification and sedimentary denitrification), as well as 2D fields of export production of organic and inorganic carbon, atmospheric Fe and NO3 deposition, the elemental stoichiometry of organic matter and the bexponent of the Martin Curve used to predict remineralisation rates.

A full description of the CSIRO Mk3L v1.2 climate system model can be found in both:

Phipps, S. J., et al. (2012) and Phipps, S. J., et al. (2011).

Chronological descriptions of the developing biogeochemical ocean model within the CSIRO Mk3L-COAL v1.0 Earth System Model that was used can be found in:

Buchanan, P. J., et al. (2018). , Buchanan, P. J., et al. (2019).  and Matear, R. J., and Lenton, A. (2014).

A full description of the variables is available in the readme file.

These simulations were run by Dr Pearse J Buchanan of the ARC Centre of Excellence for Climate Systems Sciences, as part of his PhD in quantitative marine science.