Mesoscale Model Simulation of the Sublimation and Transport of Water from the North Polar Residual Ice Cap

Introduction: The North Pole Residual Ice Cap (NPRC) is probably the most important source of atmospheric water vapor on Mars. Shortly after summer solstice the NPRC is fully exposed to the atmosphere; and, during the seasonal period between L s ~105 o and L s ~120 o , the largest global and annual water column abundances are seen in the polar atmosphere. This is also the season during which the NPRC temperatures reach their maximum; ice temperature is the single most important factor in determining water ice subli-mation rates. A prominent feature of the water cycle during the present epoch is the atmospheric transport of a significant amount of this water to the south. General Circulation Models (GCMs) have been used in recent years to better understand the dynamics of the water cycle, e.g., [1], [2] and [3]. Results from these studies have agreed quite well with observations (specifically how the zonal mean distribution of water vapor varies with latitude and solar longitude), strong evidence that the large scale dynamics in these models are realistically representing that of the Martian atmosphere itself. New observations [4], along with a re-analysis of the TES water data, have revealed that the water cycle is actually much drier than was believed (only 2 3 as wet). Also, the role of the polar regolith in the water cycle is difficult to constrain, and it may be significant, e.g., [1], [2], [4] and [5]. The NASA Phoenix Scout mission is very timely in this regard, as " ground-truth " observations of water ice in the polar regolith and the humidity of the Martian air are being collected. The NPRC is a rather complex structure, both in topography and due to the sharp gradients in albedo and thermal inertia. It is also relatively small, and when considered in the context of typical GCM model domains (and their typical spatial resolutions) it is poorly resolved. Because of this, GCMs tend to overestimate the size of the NPRC [6], allowing more water to be released into the polar atmosphere, yielding a wetter water cycle than might otherwise be simulated. Moreover, the " pole problem " inherent with traditional GCM grids (globally uniform grid spacing in latitude and longitude), and the filtering that is required to maintain computational stability, raise questions regarding the reliability of the very high-latitude dynamics in these models. Such problems can only be fully