Modeling the microphysics of mesospheric ice particles: Assessment of current capabilities and basic sensitivities

Considerable progress has been made over the past years concerning the experimental capabilities to observe mesospheric ice particles from space, from the ground, and in situ. Despite this progress regarding the observational data base, a quantitative description of related physical and chemical processes is still a challenging task due to uncertainties of several microphysical aspects concerning ice evolution in the harsh environment of the polar summer mesopause region. In the current paper, we review our current knowledge of the microphysics of mesospheric ice particles including issues like ice particle nucleation, the water vapor saturation pressure at mesopause temperatures, particle sedimentation, the equilibrium temperature of mesospheric ice particles, and particle coagulation. In addition, we consider the effect of variations of the atmospheric forcing variables like temperature, humidity, and turbulent transport. The sensitivity of ice particle properties towards these microphysical uncertainties and external forcings is assessed using the community aerosol and radiation model for atmospheres (CARMA). Simulated ice particle size distributions are analyzed applying Mie scattering calculations. Defining a hierarchy of uncertainties, our simulations suggest that the nucleation rate and number density of ice nuclei are most important, followed by the water vapor saturation pressure, and the accommodation coefficient affecting the particle temperature and sedimentation speed, and coagulation processes. Our study of the cloud sensitivity to changes of the forcing variables further reveals that close to the prevailing conditions in the polar summer mesopause region the cloud properties most strongly depend on a variation of water vapor, followed by temperature and eddy diffusion. Interestingly, our calculations suggest that the cloud brightness under the observing conditions of the SBUV/SBUV-2 suite of instruments is much more strongly controlled by variations of water vapor than temperature. Finally, we find that modeled ice particle size distributions are closely described by a Gaussian distribution. In contrast, the use of a lognormal distribution leads to a severe overestimate of the abundance of large particles.