Using MC3E In-situ Data to Develop Stochastic Representations of Cloud Microphysics for Models

Description

The representation of ice microphysical processes in numerical models is challenged by the wide range of particle shapes, sizes, and densities observed in various environments. These properties are typically specified using empirical parameters (e.g., mass (m)-dimension (D) relation parameters and particle size distribution (PSD) parameters assuming a gamma function) derived from in-situ observations, and held constant or represented as simple functions of environmental parameters. While some studies have investigated the sensitivity of simulated bulk variables to the choice of coefficients, few studies have shown the natural variability of these parameters. A range of parameters considered equally realizable for similar environmental conditions is proposed here for use in a stochastic framework that represents variability in parameterizations. Data from a variety of environments sampled during the Mid-latitude Continental Convective Clouds Experiment (MC3E) are used to characterize the variability of m-D and particle size distribution (PSD) parameters. A range of a and b coefficients describing the m-D relation as m=a x D^b under similar environmental conditions, represented as a surface of equally realizable solutions in (a,b) phase space, is first considered using a technique that minimizes the 𝜒^2 difference between the total water content (TWC) and radar reflectivity (Z) derived from PSDs measured by cloud probes on the University of North Dakota (UND) Citation and that directly measured by a TWC probe and radar. The sensitivity of these surfaces as a function of spatiotemporal scale is explored for different environments and a means of implementation within a stochastic microphysics parameterization is discussed. A similar method that determines the range of N_0, μ, and λ coefficients describing the PSD as N(D) = N_0x D^μ x e^(-λD) is proposed to quantify the variability as an equally realizable ellipsoid. The dependence of the PSD ellipsoid on environmental conditions is also analyzed. Complementary modeling efforts suggest that fixed coefficients in numerical modeling schemes do not adequately represent the variability of cloud conditions and can change the cloud optical depth and radiative impact. Future work aims to further quantify how the autocorrelation scale impacts microphysical and radiative properties in numerical simulations.