Paleodust variability since the Last Glacial Maximum and implications for iron inputs to the ocean

Abstract

Changing climate conditions affect dust emissions and the global dust cycle, which in turn affects climate and biogeochemistry. In this study we use observationally constrained model reconstructions of the global dust cycle since the Last Glacial Maximum, combined with different simplified assumptions of atmospheric and sea ice processing of dust‐borne iron, to provide estimates of soluble iron deposition to the oceans. For different climate conditions, we discuss uncertainties in model‐based estimates of atmospheric processing and dust deposition to key oceanic regions, highlighting the large degree of uncertainty of this important variable for ocean biogeochemistry and the global carbon cycle. We also show the role of sea ice acting as a time buffer and processing agent, which results in a delayed and pulse‐like soluble iron release into the ocean during the melting season, with monthly peaks up to ~17 Gg/month released into the Southern Oceans during the Last Glacial Maximum (LGM).

The PMIP4 contribution to CMIP6 – Part 4: Scientific objectives and experimental design of the PMIP4-CMIP6 Last Glacial Maximum experiments and PMIP4 sensitivity experiments

“…Models that include interactive dust, for example, should allow interactive emissions at the LGM. For this purpose, two alternative reconstructions of LGM dust emission regions are provided for models without dynamic vegetation (Hopcroft et al, 2015;Albani et al, 2016: see the PMIP4 website) and modelling groups are free to choose either one of these. If dust-enabled models do not include dynamical vegetation, then vegetation should be changed in the LGM dust emission regions so that dust emission can occur (e.g.…”

“…These are described by the "erodibility map" from the Albani et al (2016) data set and a bare soil map for the Hopcroft et al (2015) data (Fig. 6a and b, respectively).…”

“…Ginoux et al, 2001) so that dust emissions are allowed. For models that compute the dust radiative forcing from atmospheric dust mass loading, two data sets are available for the LGM: Albani et al (2014Albani et al ( , 2016 and Hopcroft et al (2015). The prescribed LGM mass loading should be implemented as perturbations of the piControl loading, i.e.…”

“…The prescribed LGM mass loading should be implemented as perturbations of the piControl loading, i.e. by either adding an anomaly to these piControl loads or by multiplying them by a ratio, the anomaly, or the ratio being computed from the Albani et al (2014Albani et al ( , 2016 or Hopcroft et al (2015) data sets. Alternatively, modelling groups can compute their own atmospheric dust mass loads offline and use them as prescribed fluxes in their coupled simulations.…”

The devil's in the disequilibrium: multi-component analysis of dissolved carbon and oxygen changes under a broad range of forcings in a general circulation model

“…Iron fertilization simulations calculate the input flux of dissolved iron to the ocean surface assuming a constant solubility and using the glacial atmospheric dust field of Mahowald et al (2006) as modified by Nickelsen and Oschlies (2015) instead of the standard pre-industrial dust field; note that this is not entirely in agreement with more modern reconstructions, which could potentially have an influence on the induced biological blooms, both in magnitude and geographically (e.g., Albani et al, 2012Albani et al, , 2016. Four iron fertilization experiments were run with the lowest radiative forcing with LGM ice sheets, as well as one model run similar to the control run.…”