Secondary Ice Production in Simulated Deep Convective Clouds: A Sensitivity Study

• Published in: Journal of the atmospheric sciences Vol. 81; no. 5; pp. 903 - 921
• Main Authors: Han, Cunbo, Hoose, Corinna, Dürlich, Viktoria
• Format: Journal Article
• Language: English
• Published: Boston American Meteorological Society 01.05.2024
• Subjects: Climate models; Cloud microphysics; Clouds; Collisions; Convective clouds; Crystals; Drizzle; Fractions; Fragmentation; Freezing; Ice; Ice crystals; Ice formation; Ice particles; Laboratories; Mathematical models; Microphysics; Mixing ratio; Numerical models; Parameterization; Phase distribution; Phase transitions; Precipitation; Precipitation formation; Rain; Remote sensing; Rime; Sensitivity; Simulation; Temperature; Temperature dependence; Arctic region
• ISSN: 0022-4928; 1520-0469
• DOI: 10.1175/JAS-D-23-0156.1

Multiple mechanisms have been proposed to explain secondary ice production (SIP), and SIP has been recognized to play a vital role in forming cloud ice crystals. However, most weather and climate models do not consider SIP in their cloud microphysical schemes. In this study, in addition to the default rime splintering (RS) process, two SIP processes, namely, shattering/fragmentation during freezing of supercooled rain/drizzle drops (DS) and breakup upon ice–ice collisions (BR), were implemented into a two-moment cloud microphysics scheme. Besides, two different parameterization schemes for BR were introduced. A series of sensitivity experiments were performed to investigate how SIP impacts cloud microphysics and cloud phase distributions in warm-based deep convective clouds developed in the central part of Europe. Simulation results revealed that cloud microphysical properties were significantly influenced by the SIP processes. Ice crystal number concentrations (ICNCs) increased up to more than 20 times and surface precipitation was reduced by up to 20% with the consideration of SIP processes. Interestingly, BR was found to dominate SIP, and the BR process rate was larger than the RS and DS process rates by four and three orders of magnitude, respectively. Liquid pixel number fractions inside clouds and at the cloud top decreased when implementing all three SIP processes, but the decrease depended on the BR scheme. Peak values of ice enhancement factors (IEFs) in the simulated deep convective clouds were 10 2 –10 4 and located at −24°C with the consideration of all three SIP processes, while the temperature dependency of IEF was sensitive to the BR scheme. However, if only RS or RS and DS processes were included, the IEFs were comparable, with peak values of about 6, located at −7°C. Moreover, switching off the cascade effect led to a remarkable reduction in ICNCs and ice crystal mass mixing ratios.