Thick and thin models of the evolution of carbon dioxide on Mars

We present results from a new simulation code that accounts for the evolution of the reservoirs of carbon dioxide on Mars, from its early years to the present. We establish a baseline model parameter set that produces results compatible with the present (i.e., Patm?6.5 mbar with permanent CO₂ ice cap) for a wide range of initial inventories. We find that the initial inventory of CO₂ broadly determines the evolutionary course of the reservoirs of CO₂. The reservoirs include the atmosphere, ice cap, adsorbed CO₂ in the regolith, and carbonate rocks. We track the evolution of the free inventory: the atmosphere, ice cap and regolith. Simulations begin at 4.53 Gyr before present with a rapid loss of free inventory to space in the early Noachian. Models that assume a relatively small initial inventory (?5 bar) have pronounced minima in the free inventory of CO₂ toward the end of the Noachian. Under baseline parameters, initial inventories below ∼4.5 bar result in a catastrophic loss of the free inventory to space. The current free inventory would be then determined by the balance between outgassing, sputtering losses and chemical weathering following the end of the late bombardment. We call these “thin” models. They generically predict small current free inventories in line with expectations of a small present CO₂ ice cap. For “thick” models, with initial inventories ?5 bar, a surplus of 300-700 mbar of free CO₂ remains during the late-Noachian. The histories of free inventory in time for thick models tend to converge within the last 3.5 Gyr toward a present with an ice cap plus atmospheric inventory of about 100 mbar. For thick models, the convergence is largely due to the effects of chemical weathering, which draws down higher free inventories more rapidly than the low. Thus, thick models have ?450 mbar carbonate reservoirs, while thin models have ?200 mbar. Though both thick and thin scenarios can reproduce the current atmospheric pressure, the thick models imply a relatively large current CO₂ ice cap and thin models, little or none. While the sublimation of a massive cap at a high obliquity would create a climate swing of greenhouse warming for thick models, under the thin model, mean temperatures and pressures would be essentially unaffected by increases in obliquity.