Simulation of Mars Dust Storms with a Limited Surface Dust Source and Self-determining Wind Stress Thresholds in the MarsWRF General Circulation Model

Observations of Mars reveal changes in the surface dust distribution during and after major storm events, and suggest that at least some storms may be curtailed by depletion of source regions. In general circulation models (GCMs), most regions of high surface wind stress lose more dust via lifting than they receive via sedimentation over multiple realistic Mars years (years with and without storms), suggesting that on Mars such regions may be fully depleted of dust over time, hence leaving them unable to act as primary lifting centers for storm events.

We hypothesized that regions exist which gain as much dust as they lose over multiple realistic Mars years, and are the true primary lifting centers for dust storms despite typically having much lower wind stresses. We tested this hypothesis using the MarsWRF GCM with parameterized wind stress and dust devil lifting, and including a finite total dust budget, with the amount of dust at each point on the surface tracked over time as dust removal (due to lifting) and deposition (due to sedimentation) occurred. Performing multi-decadal simulations, we intermittently increased the globally-uniform dust injection rate parameter, α_N, to maintain realistic atmospheric opacities as increasing numbers of source regions dropped out. We ran the model for >>100 Mars years for five choices of globally-uniform threshold wind stress τ_t (the minimum wind stress required for dust lifting to occur), but were unable to find an ‘equilibrium state' – i.e., one in which both (a) no further persistent increase or decrease in surface dust occurred at any grid point, and (b) dust storms continued to occur without further increases in α_N being required. We will present a subset of these results.

We then tested the Pankine and Ingersoll [2004] hypothesis that dust on Mars may be in a state of self-organized criticality, with τ_t in a given location allowed to increase [decrease] as the local surface dust cover is decreased [increased]. In this case, and for certain functions relating the variation in threshold to the change in surface dust, we found an ‘equilibrium' solution satisfying both (a) and (b), and we will present a subset of the storms and surface dust distributions produced.

Though somewhat plausible on the basis of ‘sheltering' arguments (e.g., that as dust is removed, the remaining dust is lower in the PBL and sheltered from the wind by larger, non-erodible elements, thus receives less wind stress, equivalent to the local threshold increasing), our parameterized variation of τ_t may not accurately represent the real physical process(es) occurring on Mars, but is rather intended to demonstrate the potential importance of finite surface dust and self-determining thresholds, and as a prelude to further studies.