1. Introduction

Ice crystal growth under fixed temperature and humidity is well described by the habit diagram (Libbrecht, 2017). Plate-like crystals form between $-10^\circ\,\text{C}$ and $-22^\circ\,\text{C}$, evolving from simple plates to dendrites with increasing ice supersaturation, while columnar habits dominate at warmer temperatures ($-4^\circ\,\text{C}$ to $-10^\circ\,\text{C}$), ranging from solid columns to needles at very high supersaturation.

In the atmosphere, sedimenting ice particles experience changing growth regimes. While the transition from columnar to plate-like growth (capped columns) is well documented, the reverse transition remains poorly understood. No studies have examined dendrites entering the needle growth regime, despite the importance of dendrites for aggregation and secondary ice production (e.g., von Terzi et al. (2022)).

Here, we investigate the evolution of dendritic crystals exposed to needle growth conditions using experiments conducted in the cold chamber at the Johannes Gutenberg University of Mainz.

2. Experimental setup

Experiments were conducted over one week in the cold chamber at the University of Mainz. Ice crystals were grown above an aquarium containing $\sim 1\, \text{cm}$ of liquid water, maintained slightly above $0^\circ\,\text{C}$ using a heating mat. Evaporation from the water surface provided the humidity required for crystal growth.

A styrofoam lid with a small central opening was placed over the aquarium to channel the rising vapor, creating a localized growth region where ice crystals formed on a 100 μm nylon string. Temperature and relative humidity were continuously measured at the height of the string. A schematic of the setup is shown in Figure 1.

3. Experiments

3.1. Placing a dendritic particle into the needle growth regime

Thirteen experiments were performed in which dendritic ice crystals were grown on a nylon string at $-12^\circ\,\text{C}$ to $-17^\circ\,\text{C}$ and then gradually warmed into the needle growth regime ($-7^\circ\,\text{C}$). Crystal growth was recorded and followed by microscopic analysis (videos: http://leonie.von-terzi.de/research/fun_snow/). Experiments 4 and 8 are shown as representative cases.

In both experiments, dendrites grown for ~20 min became noticeably denser and more compact after 20 min in the needle regime. Distinct dendritic branches blurred, and columnar growth developed on existing branches, filling inter-branch space and producing a clumped morphology (Fig. 3 and 4).

3.2. Sequential Growth at $-15^\circ\,\text{C}$, $-7^\circ\,\text{C}$, and $-15^\circ\,\text{C}$

In Experiment 14, a dendritic crystal was grown at $-15^\circ\,\text{C}$, then warmed to $-7^\circ\,\text{C}$ to induce needle growth, and finally cooled back to $-15^\circ\,\text{C}$ for further growth. The particle was transferred to tweezers for close-up imaging during the final phase.

As shown in Figure 5, the dendritic structure remained largely intact after the second $-15^\circ\,\text{C}$ phase, with needle-like extensions exhibiting slightly broadened tips. Close-up images (Fig. 6) reveal that needles formed along the dendritic branches at $-7^\circ\,\text{C}$ and subsequently developed plate-like caps upon cooling, resulting in capped needles.

4. Conclusions

The experiments show that when a dendrite enters the needle growth regime, columnar growth develops on each dendritic branch, progressively filling the space between arms. This process likely increases particle density and, in extreme cases, alters the dendritic appearance, which may explain the rarity of pristine dendrites observed at the ground. These findings raise the question of whether a similar fill-in mechanism occurs for aggregates and whether it could explain density discrepancies between aggregation models (e.g., Leinonen and Moisseev (2015)) and observations. Further studies with improved control of temperature and humidity are needed.

All growth videos are available at: http://leonie.von-terzi.de/research/fun_snow/.