Numerical Simulations of the Formation and Evolution of Water Ice Clouds in the Martian Atmosphere

A one-dimensional aerosol model was used to study the microphysics of water ice cloud formation and evolution on Mars. The time-dependent nature of the model permits realistic simulations of diurnal cycles. The results are sensitive to the choice of the temperature profile, the eddy diffusion coefficient, the efficiency of nucleation of the condensation nuclei, and the condensation history of the cloud. Simulations having a large diurnal temperature variation yield condensed water total optical depths of a few tenths, in good agreement with the Viking Lander opacity data. Ice clouds are typically simulated at 25-30 km altitude with mean particle radii of 0.2-5.0 μm. We also obtain good agreement with the particle extinction observed in the Viking Orbiter limb data for a temperature profile constrained by the Infrared Thermal Mapper data. In order for repeated cloud formation to occur each day, we need to include a vertical and/or horizontal resupply of water and dust to the higher altitudes, where the condensation occurs, because of loss by sedimentation. This can take the form of a micrometeoritic influx of particles at the top of the atmosphere or a latitudinal transport of material. The efficiency of the nucleation process is determined by the contact angle. Changing this angle from 0.945 to 0.999 can have the effect of doubling the total ice optical depth of the cloud. The time evolution of the cloud is important for comparison with the observations if a diurnal temperature change exists. Under these conditions, the optical thickness of the cloud is dependent on the number of condensation cycles having previously occurred and the time of day. This time dependence and the need for a resupply of material into the condensing region show that the cloud formation process cannot be considered a steady-state phenomenon. We have shown that the eddy diffusion coefficient and particle radii cannot be constrained by the available data, even if a detailed model is used for the interpretation of the observations. In our simulations there is an extensive vertical redistribution of water by clouds. However, water is only lost to the surface when the temperature is at the frost point near the ground. Precipitation across subsaturated layers is not significant.