Mrams Today – One Example of Current Mars Mesoscale Modeling Capabilities

Introduction: Currently there exist a number of Mars mesoscale atmospheric models (e.g., [1], [2], [3]), with many more in development and/or testing around the world. In light of this burgeoning worldwide use of such models, it may be useful to describe some of the capabilities that a current Mars mesoscale model may possess. This will be accomplished here by reviewing notable changes and additions to the nonhy-drostatic Mars Regional Atmospheric Modeling System (MRAMS) since its introduction in 2001 [1]. Numerical Techniques: The basis of any numerical model, the " numerics " may also contain subtle in-stabilities and questionable assumptions. Fully compressible dynamics. In 2001, MRAMS used a quasi-Boussinesq approximation (inherited from the terrestrial RAMS model [4] on which MRAMS was originally based) that allowed for some aspects of fluid compressibility, but not all. In particular, a change in potential temperature would not directly affect the air pressure. For many terrestrial applications this approximation has only relatively minor consequences , although it has been recognized as a limitation in some cases (e.g., [5]). For the atmosphere of Mars, however, the use of this approximation resulted in significant mass loss (on the order of 0.1 hPa/sol) from the model domain. The drastic difference between the model's behavior on Earth versus Mars is thought to be primarily due to the significantly larger atmospheric heating/cooling rates (chiefly due to radiative transfer) on Mars, particularly those near the surface. The model's use of an altitude-based vertical coordinate (vs. a pressure-based one) may also have played a role. To remedy this issue, MRAMS was modified to treat the atmosphere as being fully compressible, following [5]. Improved Cartesian horizontal gradients. Many atmospheric diffusion schemes require the calculation and/or interpolation/extrapolation of Cartesian horizontal gradients of temperature and other fields. Particularly in regions of complex topography overlain by thermally-stratified air, simpler mathematical treatments to determine these gradients from the model state variables (which are on the native terrain-following coordinate system) often produce significant errors. Such errors typically manifest most strikingly in thermal fields, with temperatures in valleys far too cold (e.g., 50K), and temperatures above ridges far too warm (e.g., 400K). A new method to carefully calcu-late/interpolate/extrapolate these gradients with much