The MarsWRF model is set up with fully interactive dust at 5° × 5° and 2° × 2 resolution. The latter allows for a better representation of topography and other surface properties. An infinite reservoir of surface dust is assumed for both resolutions. For 5° × 5°, surface dust lifting by wind stress takes place over broad areas, occurring in about 20% of the model’s grid cells. For 2° × 2°, it is more spatially restricted, occurring in less than 5% of the grid cells, and somewhat reminiscent of the corridors Acidalia‐Chryse, Utopia‐Isidis, and Arcadia‐West of Tharsis. The onset times of major dust storms ‐ large regional storms or global dust storm events (GDEs) ‐ do not exhibit much inter‐annual variability, typically occurring at around L_s 260°. However, their magnitude does show significant inter‐annual variability ‐ with only small regional storms in some years, large regional storms in others, and some years with GDEs ‐ owing to the interaction between major dust lifting regions at low latitudes. The latter is consistent with observed GDEs having several active dust lifting centers. The model’s dust distribution is found to better agree with observation‐based albedo and dust cover index maps for the 2° × 2° run. For the latter, there is also significant surface dust lifting by wind stress in the aphelion season that is largely confined to the Hellas basin. It has a recurring time pattern of 2‐7 sols, possibly resulting from the interaction between mid‐latitude baroclinic systems and local downslope flows.

Mars General Circulation Models (MGCMs) simulate the Mars climate and atmosphere for many Martian Years (MYs). A challenge is to configure such models to produce Mars global dust storm events (GDEs) in few but not all MYs. That is because GDEs are known to occur once every few MYs, on average. We set up the MarsWRF MGCM in this way using “interactive dust,” meaning the model freely lifts, transports, and deposits surface dust (assuming an inexhaustible amount of available surface dust). We use different horizontal model grid point resolutions and compare their results in terms of dust storm source regions and changes in surface dust loading. For the high‐resolution experiment, we find that GDEs are likely to develop if regional dust storm activity around two equatorial source regions on the planet, namely, south of Chryse Planitia and in the northern Hellas basin, combine with one another. The latter is consistent with knowing from observations that GDEs have several active dust lifting centers. Also, we find that the model's surface dust distribution in the high‐resolution experiment agrees potentially better with observation‐based dust cover maps.