SIMULATIONS OF THE GENERAL-CIRCULATION OF THE MARTIAN ATMOSPHERE .2. SEASONAL PRESSURE VARIATIONS

We have simulated the CO2 seasonal cycle of the Martian atmosphere and surface with a hybrid energy balance model that incorporates dynamica l and radiation information from a large number of general circulation model (GCM) runs. This information includes heating due to atmospheri c heat advection, the seasonally varying ratio of the surface pressure at the two Viking landing sites to the globally averaged pressure (r( k)), the rate of CO2 condensation in the atmosphere, and solar heating of the atmosphere and surface. The GCM runs collectively covered a fu ll set of seasonal dates and a large range of dust optical depths. We have compared the predictions of the energy balance model with the sea sonal pressure variations measured at the two Viking landing (VL) site s and the springtime retreat of the seasonal polar cap boundaries. Num erical experiments with the energy balance model indicate that the fol lowing quantities have a strong influence on the VL seasonal pressures : albedo A(is) of the seasonal CO2 ice deposits, emissivity e(is) of t his deposit, atmospheric heat advection, and the pressure ratio r(k). This last factor does not enter into the seasonal CO2 condensation/sub limation cycle in a significant way. The numerical experiments also in dicate that the following factors have only a minor effect on the VL p ressures: (1) the net radiative effects (solar plus thermal) of atmosp heric dust at the latitudes of the polar caps, and (2) the subsurface heat conduction. The significant influence of the pressure ratio r(k) on the VL, seasonal pressures is due to large seasonal variations in t he global distribution of surface pressure. At low and mid-latitudes, these ''weather'' variations are engendered by seasonal changes in the Hadley circulation and by seasonal changes in the atmospheric scale h eight close to the surface. Comparison of the VL1 and VL2 pressures wi th one another provide direct evidence for the presence of such a ''we ather component'' in the measured pressures. The differential weather component (VL2-VL1) derived from the data is reproduced approximately by the energy balance model. We find that the seasonal weather variati ons account for about 20% and 30% of the seasonal pressure variations measured at VL1 and VL2, respectively, that dynamical and scale height variations make comparable contributions to the weather component dur ing years without global dust storms, and that the dynamical contribut ion is the larger one during years with global dust storms. Interannua l variations in the weather component, rather than variations in CO2 c ondensation rates, are the dominant sources of the observed interannua l variations of pressure during the season of global dust storms. Opti mum fits to the Viking pressure measurements and the data on the polar cap boundaries are achieved with values of about 0.45 and 0.75 for A( is) and e(is), respectively. The former value is consistent with avail able photometric determinations of the albedo of the seasonal caps, wh ile the latter value, especially in light of infrared thermal mapper b rightness temperatures at high latitudes, may reflect, in part, the in fluence of the polar hood on the radiation balance of the winter polar regions.