A Review of Ice Particle Shapes in Cirrus formed In Situ and in Anvils

Abstract

Results from 22 airborne field campaigns, including more than 10 million high-resolution particle images collected in cirrus formed in situ and in convective anvils, are interpreted in terms of particle shapes and their potential impact on radiative transfer. Emphasis is placed on characterizing ice particle shapes in tropical maritime and midlatitude continental anvil cirrus, as well as in cirrus formed in situ in the upper troposphere, and subvisible cirrus in the upper tropical troposphere layer. There is a distinctive difference in cirrus ice particle shapes formed in situ compared to those in anvils that are generated in close proximity to convection. More than half the mass in cirrus formed in situ are rosette shapes (polycrystals and bullet rosettes). Cirrus formed from fresh convective anvils is mostly devoid of rosette-shaped particles. However, small frozen drops may experience regrowth downwind of an aged anvil in a regime with RH ice >~120% and then grow into rosette shapes. Identifiable particle shapes in tropical maritime anvils that have not been impacted by continental influences typically contain mostly single plate-like and columnar crystals and aggregates. Midlatitude continental anvils contain single-rimed particles, more and larger aggregates with riming, and chains of small ice particles when in a highly electrified environment. The particles in subvisible cirrus are <~100 μm and quasi-spherical with some plates and rare trigonal shapes. Percentages of particle shapes and power laws relating mean particle area and mass to dimension are provided to improve parameterization of remote retrievals and numerical simulations.

Global Statistics of Ice Microphysical and Optical Properties at Tops of Optically Thick Ice Clouds

“…This finding is largely consistent with theory and observations for temperatures above 233 K (Bailey & Hallett, 2009; Hashino & Tripoli, 2007), although the slight increase in column fraction around 260 K appears to be inconsistent with the expected increase in thin plates. Furthermore, at conditions colder than 233 K and at sufficiently high supersaturations, columnar habits are generally assumed to more prevalent than plate‐like crystals (Bailey & Hallett, 2004; Lawson et al, 2019). However, at such temperatures, crystals are generally polycrystalline and often classified as irregular crystals, especially in thick ice clouds, making firm conclusions on the dominating shapes of the crystal components from in situ images to be effectively impossible (Bailey & Hallett, 2009; Lawson et al, 2019).…”

“…Furthermore, at conditions colder than 233 K and at sufficiently high supersaturations, columnar habits are generally assumed to more prevalent than plate‐like crystals (Bailey & Hallett, 2004; Lawson et al, 2019). However, at such temperatures, crystals are generally polycrystalline and often classified as irregular crystals, especially in thick ice clouds, making firm conclusions on the dominating shapes of the crystal components from in situ images to be effectively impossible (Bailey & Hallett, 2009; Lawson et al, 2019). Note that, while our results indicate that plate‐like crystal components dominate, the average aspect ratio of about 0.5 is consistent with thick plates, which are often confused with short columns and may also be components of “budding rosettes” (Bailey & Hallett, 2004, 2009).…”

Process‐Based Simulation of Aerosol‐Cloud Interactions in a One‐Dimensional Cirrus Model

“…Quasi‐spherical particles together with hexagonal plates and columns are frequently observed habits of sub‐100‐ m ice crystals in many cirrus types (Lawson et al, ). Spheroids are therefore used as a first‐order characterization of the habits of model ice crystals for simplicity and generality.…”

“…The latest version of this method, Diffusion Surface Kinetics Ice Crystal Evolution (DiSKICE), is described in Harrington et al () and allows to incorporate crucial physical mechanisms and dependencies that are entirely absent in conventional uptake models. However, other shapes, most notably bullet rosettes, can be found dominating the habit distribution below maximum crystal dimensions 50–100 m in some types of cirrus (Lawson et al, ). In such cases, habits may at this point be diagnosed from single ice crystal mass‐dimension relationships.…”

“…This issue is strongly tied to the inherent complexity of the physics of vapor growth—the prediction of ice crystal morphology requires sound knowledge of a plethora of processes controlling H O surface attachment kinetics (Libbrecht, ). At the same time, in situ measurements reveal a rich suite of ice crystal habits in cirrus, depending on cirrus type and meteorological conditions in which these clouds form and mature (Lawson et al, ), calling for better understanding and explanation.…”