Are numerical models a reliable source of information on the Martian environment ?

One particular strength of using a GCM to compile such a database is that it provides a physically consistent estimate of the environmental conditions on Mars for seasons and dust loadings which are not covered by the observations. But is the climate database consistent with the available observations ?

First, several key parameters (pressure, dust scenario) of the General Circulation models have been tuned to match the observations. For instance, the GCM does a very good job in reproducing the seasonal pressure variations recorded by Viking Lander 1 and 2 because the total atmosphere mass and polar caps albedo were chosen for this purpose. On this basis, we are extremely confident in the ability of the database to predict the surface pressure elsewhere on the planet, something only a GCM can do !

Second, the MCD has been validated against most of the available observations. In most cases, it is found that the model is able to predict the observations with a very good accuracy. Figures 1, 2, 3, 4 and 5 show such comparison for recent observations from the Mars Global Surveyor and Mars Pathfinder mission. Further comparisons can be find in Forget et al. (2001b) and in Lewis et al. (1999) . This validation processes has also shown that some problems remain in some locations and seasons. For instance, the GCMs do not simulate the large temperature inversions which are sometime observed in the tropics in summer ; polar night profiles seems to be slightly colder in reality than in the database during southern winter ; there is no permanent southern CO$_2$ ice polar caps in the database...

Nevertheless, the agreement between the database and the observations should be good enough for most applications, and we believe that the database is the best tool available for most purpose.

Figure 2: Example of very good fits to the observations that can be obtained with the database MGS scenario at various seasons. The black solid lines show temperature profiles measured by radio-occultation with Mars Global Surveyor. The red dashed lines are the MCD predictions at the same locations and times. The model is usually able to simulate accurately the variations of the temperature profiles due to change in dust loading and insolation.

Figure 3: Mean meridional cross-sections of atmospheric temperatures retrieved from the MGS Thermal Emission Spectrometer observations (top, left pannel) and corresponding gradient winds (bottom, left pannel) at $L_s=270^o$ compared to time-mean zonal-mean plots from the Mars Climate database using the MGS scenario for the same season. One can notice that the simulated summer pole atmosphere at $L_s=270^o$ is slightly cooler than in the TES inversions, resulting in weaker or absent simulated easterly winds high in the summer hemisphere. This discrepency is probably due to the occurence of a regional dust storm between $L_s=260^o$ and $L_s=272^o$ at the edge of the south polar cap during the year observed by MGS. TES figure from Smith et al. (2001).

Figure: A comparison of the Mars Pathfinder surface measurements with the MCD. The small squares show the Mars Pathfinder measurements and the solid line is the mean of the observations taken over the first 30 days of the mission. The large circles connected by dashed lines show the MCD predictions interpolated to the Mars Pathfinder location. Red lines: MGS ``dust scenario'' (the low dust scenario gives very similar results). Green line: Viking scenario. Top pannel: The pressure diurnal cycle : the MGS scenario appears to underestimate the total amplitude of the surface pressure tide, whereas the Viking dust scenario appears to give a better fit. Middle panel: the temperature at the top of the 1 m meteorological mast (about 1.27 m above the surface). The dotted lines are the MCD surface temperature (top line at midday) and the MCD lowest atmospheric level (5 m) temperature (bottom line). Bottom panel: The wind direction axis, indicated by the compass point from which the wind is blowing from ($0^o$ is a northerly). The general sense of rotation is determined by the passage of the diurnal thermal tide; the details, such as the small rotation toward an easterly near dawn before the subsequent rotation to a westerly, are a consequence of the local topography.

Figure 5: A comparison of the density measured in-situ (red dots) by the Mars Global Surveyor accelerometer during aerobraking around 125 km during northern winter (Keating et al., 1998) with the densities predicted at about the same location by the Mars Climate Database (MGS scenario). The absolute value of the density is well predicted in spite of the extreme sensitivity of density to the entire atmosphere below. In addition, the MCD predicts longitudinal variations comprising wave-like wavenumber 1 and 2 structures which are quite similar to the observations.