Characterizing spatially heterogeneous and homogeneous low-level mixed phase clouds over the Arctic using in situ measurements

Abstract. Polar mixed phase clouds (MPCs) are notoriously difficult to represent in models of varying scales; and can be difficult to characterize with varying observational platforms. Accurate measurements/characterizations of MPCs are therefore vital to improve our understanding of them. This study uses airborne, in situ measurements acquired during three field campaigns to examine cloud phase properties and the spatial heterogeneity of Arctic MPCs. Cloud phase is diagnosed using a novel methodology incorporating a Bayesian Gaussian Mixture Model. To determine how the proposed methodology compares with previously developed classification methods, cloud phase occurrence frequencies are compared between four different methodologies. Large differences exist based on methodological approach, warranting future intercomparisons. This study introduces three MPC types which classify consecutive 1 Hz in-cloud measurements as 1) liquid-dominated (containing only liquid and mixed phase samples), 2) ice-dominated (containing only ice and mixed phase samples) and 3) phase-segregated (containing liquid and ice phase samples), which are the least common. Liquid-dominated MPCs are more spatially heterogeneous than ice-dominated MPCs. Discretizing MPC measurements into these categories may help disentangle complex relationships between physical factors controlling MPCs. For example, we evaluate differences of spatially heterogeneous and homogenous MPCs by controlling for the frequency occurrence of different cloud phases within MPCs. Spatially heterogeneous liquid-dominated MPCs have higher (lower) concentrations of large droplets (ice crystals) having diameters ~25–60 µm (>~1000 µm) than spatially homogeneous MPCs. Minor differences are observed for turbulent kinetic energy; whereas heterogeneous MPCs preferentially occur from −10° to −5 °C.