Mesoscale and large-eddy simulation model studies of the Martian atmosphere in support of Phoenix

In late May of 2008, the NASA/JPL Phoenix spacecraft will touch down near its targeted landing site on Mars (68.2°N, 126.6°W). Entry, descent, and landing (EDL) occurs in the late afternoon (∼1630 hours local solar time (LST)) during late northern spring (Ls ∼ 78°). Using a mesoscale and a large‐eddy simulation (LES) model, we have investigated the range of conditions that might be encountered in the lower atmosphere during EDL. High‐resolution (∼18 km) results from the Oregon State University Mars MM5 (OSU MMM5) are used to understand the hazards from the transient circulations prominent during this season. Poleward of ∼80°N these storms produce strong winds (∼35 m s−1) near the ground; however, owing to the synoptic structure of these storms, and the deep convective mixed layer equatorward of the seasonal cap boundary during EDL, our modeling suggests the spacecraft would not be in winds stronger than ∼20 m s−1 at parachute separation. The storm‐driven variability is much weaker at Phoenix latitudes than it is poleward of the seasonal cap edge (result from an extensive sensitivity study). The OSU MLES model is used to explicitly simulate the hazards of convection and atmospheric turbulence at very high resolution (100 m). This modeling suggests that an upper bound for the maximum expected horizontal‐mean atmospheric turbulent kinetic energy (TKE) is ∼12 m2 s−2, seen ∼3 km above the ground at ∼1430 hours LST. TKE amplitudes are greatest when the horizontal mean wind is large (shear production) and/or the surface albedo is low (a lower albedo enhances buoyancy production, mimicking decreased atmospheric stability after a storm advects colder air into the region). LES simulations predict deep mixed layers (∼6–7 km), ∼1.5 km deeper than the mesoscale model (∼5 km). Mesoscale modeling suggests that the actual landing site differs meteorologically from other longitudes (larger‐amplitude diurnal wind cycle), a consequence of the strong thermal circulations that are excited by the very large regional topography. The OSU MLES model was modified for this work to utilize time‐ and height‐dependent geostrophic wind forcing (constructed from OSU MMM5 results). With this forcing, the OSU MLES provides a site‐specific simulation, where the time/height variability of the horizontal mean LES wind field is in good agreement with the OSU MMM5. On the basis of some statistical analysis, we have good confidence that the “full‐spectrum” wind field is within engineering guidel