An evaluation of ice formation in large‐eddy simulations of supercooled Arctic stratocumulus using ground‐based lidar and cloud radar

Abstract

In order to evaluate possible ice formation processes in mixed phase Arctic stratocumulus, we compare measurements of radar reflectivity and Doppler velocity and lidar backscatter coefficient and circular depolarization ratio with the corresponding quantities computed from large‐eddy simulations (LES). The measurements are taken from the Millimeter Cloud Radar and the Arctic High Spectral Resolution Lidar during the Mixed‐Phase Arctic Cloud Experiment in October 2004. Lidar depolarization, computed from LES results including well‐known ice formation and multiplication processes and measured ambient ice nuclei (IN), is near 0%, indicating negligible ice formation, whereas measured median depolarization is 84%, indicating strong ice formation, consistent with in situ aircraft measurements. Reducing ice particle fall speeds, increasing IN concentrations, or introducing a surface source of IN does not sufficiently increase the simulated depolarization values and/or results in poor agreement of other simulated parameters with the measurements. Introducing additional hypothetical ice formation processes (i.e., formation of IN from evaporating drops, freezing of evaporation drops, or droplet freezing rates per unit surface or volume) generally brings all investigated parameters into relatively close agreement with the radar and lidar measurements. These results provide additional evidence for the existence of one or more unestablished ice formation processes in Arctic stratocumulus.