DISPERSAL OF ASH IN THE MARTIAN ATMOSPHERE

Extensive fine-grained, layered deposits on Mars have been interpreted by some workers as ash fall deposits. However, in some cases, the source volcanoes for these deposits are ambiguous or unknown. To investigate a pyroclastic origin for these deposits, settling velocities and advection distances can be calculated for a range of ash particles, subject to detailed aerodynamic interactions with the atmosphere. Sedimentation rates of volcanic ash depend on particle properties (size, density, shape) and the characteristics of the ambient environment (atmospheric structure, wind field, gravity). On Earth, at almost all relevant altitudes, the atmosphere is well described by continuum mechanics, and the conventional Newtonian description of particle motion allows settling velocities to be related to particle characteristics via a drag coefficient. However, under the rarefied atmospheric conditions of Mars, there is a range of altitudes for which non-continuum effects become important for ash-sized particles. In this case, the mean free path of the atmospheric gas is large relative to the particle size, and the resistance to particle motion is determined by the sum of molecular collisions with the ash particle. We have developed an equation of motion based on statistical mechanics as the appropriate approach to calculating particle motion under conditions of very low atmospheric pressure. For any given start height, the variation of settling velocity with altitude is computed, switching to a continuum approach when lower (denser) atmospheric altitudes are reached. When coupled with an appropriate treatment of the atmospheric wind field, this approach allows for calculation of particle trajectories and settling times, and development of inferences for deposit extent. We find that settling velocities are substantially greater in rarefied atmospheric conditions than are calculated with a continuum approach. However, irregular particle shapes and particle aggregation are competing influences on settling rates. Overall, we infer that deposit dispersal is likely to be substantially less than has been calculated previously, which requires a reassessment of the potential source volcanoes of fine-grained layered deposits and/or a reinterpretation of the origins of the deposits themselves.