1. Environmental & Climate Sciences Department Seminar

    "Invariant and insensitive: climate model microphysics as a scaling problem"

    Presented by Mikael Witte, National Center for Atmospheric Research

    Thursday, July 19, 2018, 11 am
    Conference Room Bldg 815E

    Hosted by: Yangang Liu

    Clouds are inherently multiscale phenomena: the particles that make up clouds are typically microns to millimeters, while the large-scale circulations that drive cloud systems can be hundreds of kilometers across. Limited computational power and the need to accurately represent the large-scale circulations in numerical simulations of the atmosphere make explicit inclusion of cloud microphysics a practical impossibility. In the last 20 years there has been a shift toward representing microphysics as scale-aware processes. Despite this shift, many unanswered questions remain regarding the scaling characteristics of microphysical fields and how best to incorporate that information into parameterizations. In this talk, I will present results from analysis of high frequency in situ aircraft measurements of marine stratocumulus taken over the southeastern Pacific Ocean aboard the NCAR/NSF C-130 during VOCALS-REx. First, I will show that cloud and rain water have distinct scaling properties, indicating that there is a statistically and potentially physically significant difference in the spatial structure of the two fields. Covariance of cloud and rain is a strong function of length/grid scale and this information can easily be incorporated in large-scale model parameterizations. Next I will show results from multifractal analysis of cloud and rain water to understand the spatial structure of these fields, the results of which provide a framework for development of a scale-insensitive microphysics parameterization. Finally, I compare observed microphysical scaling properties with those inferred from large eddy simulations of drizzling stratocumulus, applying the same analyses as applied to the aircraft observations. We find that simulated cloud water agrees well with the observations but the drizzle field is substantially smoother than observed, which has implications for the ability of limited-area models to adequately reproduce the spatial structure o