the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Wind tunnel experiments to quantify the effect of aeolian snow transport on the surface snow microstructure
Hagen Weigel
Sonja Wahl
Henning Löwe
Abstract. The evolution of the surface snow microstructure under the influence of wind is hardly understood but crucial for polar and alpine snowpacks. Available statistical models are solely parameterized from field data where conditions are difficult to control. Controlled experiments which exemplify the physical processes underlying the evolution of density or specific surface area (SSA) of surface snow under wind are virtually non-existing. As a remedy, we conducted experiments in a cold laboratory using a ring-shaped wind tunnel with an infinite fetch to systematically investigate wind-induced microstructure modifications under controlled atmospheric, flow and snow conditions. Airborne snow particles are characterized by high-speed imaging, while deposited snow is characterized by density and SSA measurements. We used a single snow type (dendritic fresh snow), cover wind speeds from 3 ms−1 to 7 ms−1 (for fixed temperature) and vary temperatures from -24 °C to -2 °C (for fixed wind speed). The measured airborne impact trajectories confirm the consistency of our coefficient of restitution with large scale saltation, rendering the setup suitable to realistically study interactions between airborne and deposited snow. Our measured densification rates in the deposit as a function of wind speed show clear deviations from existing statistical models, but can be re-parameterized through our data. The most drastic changes in densification and SSA rates of deposited snow are observed close to the melting point. This study, as a first of its kind, exemplifies a rich non-linear interplay between airborne and deposited snow particles which is discussed in view of a multitude of involved processes, i.e. airborne metamorphism, cohesion, particle separation and fragmentation.
- Preprint
(5103 KB) - Metadata XML
- BibTeX
- EndNote
Benjamin Walter et al.
Status: final response (author comments only)
-
RC1: 'Comment on tc-2023-112', Anonymous Referee #1, 15 Sep 2023
The authors present a very exciting and compelling experiment focused on the effect of wind snow surface microstructure. For how simple the question is, this is an incredibly hard problem to work on. We have limited tools at our disposal to make concise measurements of snow microstructure, and it is incredibly difficult to run such an experiment in the field. This group at SLF has succeeded at combining their expertise in snow microstructure and wind tunnel experiments to provide new insights into this intriguing aspect of snow metamorphism.
My only concern with this research is the significantly unphysical conditions under which snow is being transported. As it stands, I do not see a reason why the rate of change of any snow characteristics in their experiment should be related to any measurements of natural snow undergoing natural transport. It is unclear if the snow particles actually come in contact with the propeller driving their RWT. A schematic that shows this mechanism would be very helpful. what the ,
More importantly, the authors acknowledge that a large portion of snow particles are transported along the outer wall due to centrifugal forces and, among other effects, this causes a measurable impact on density. This is well outside the realm of normal saltation and suspension. Given that v_x is so much larger than v_z, this impact force may be considerably higher than in nature. As well, repeat impacts caused by snow working its way around a corner may cause orders of magnitude more fragmentation.
Given these concerns, could you please address the question of transport around the curves (impact velocities, momentum balance, fragmentation rate, restitution coeff, how many more impacts per second? etc.), or modify the manuscript in such a way that the reader knows while you may have novel measurements of a physical process, this physical process has little relation to what one may expect to find in nature? As it stands, I think the quantitative information provided needs to be qualified or better justified.
There are a few grammatical things that could be improved:
L8- Cover wind speeds?
L11-In the deposit?
L20- Is rolling different from creep?
L19-Chemical species?
L60: Do you mean necessary or inevitable?
L161: To make contact to previous studies?
Other comments
L53-54- At what height are these wind velocities?
L76-77: Very cool
L80-82: Do the particles not come in contact with the propeller?
Figure 2: How did you conclude the jump in RH was from snow particle sublimation? What’s the RH of the cold room?
L258-259: Very cool
L267-268: Again, how can you decouple this from the effect of particles smashing into walls that are necessarily there in nature?
Citation: https://doi.org/10.5194/tc-2023-112-RC1 -
RC2: 'Comment on tc-2023-112', Anonymous Referee #2, 15 Sep 2023
Review of the paper “Wind tunnel experiments to quantify the effect of aeolian snow transport on the surface snow microstructure” by Walter et al. submitted to The Cryosphere.
This paper presents an innovative set of measurements to investigate the effect of wind-induced snow transport on the physical properties of surface snow (density, SSA). These measurements were collected in a ring-shaped wind tunnel (RWT) that reproduces the main characteristics of aeolian snow transport. The authors quantified the changes in density and SSA during events with different wind speed and air temperature. Their analysis confirmed the increase in surface snow density with increasing wind speed that has been observed in the field and highlighted a slight decrease in SSA with increasing wind speed. The author also compared the densification rates measured in the RWT with parameterizations used in snowpack schemes.
The subject of this paper is very interesting for the snow community and presents a set of original measurements to quantify the effects of wind on the physical properties of surface snow. So far, these quantifications have mainly been obtained from field measurements (mainly for surface snow density) that are influenced by other physical processes, making it challenging to disentangle the effect of the wind from the other processes. These measurements can serve to develop more-physically based parameterizations of the impact of wind on the physical properties of the snow cover in multi-layer snowpack schemes such as Crocus and SNOWPACK. Therefore, this paper should be published in The Cryosphere. However, prior to publication, the author must carefully define in which context they are working (blowing snow event with concurrent snowfall) and revise accordingly which existing parameterizations they are evaluating in the context of the study. These two general comments are followed by more specific and technical comments.
General comments
1. In this paper, the authors study the effects of aeolian snow transport on the properties of surface snow using a RWT where snow is continuously added to mimic snow precipitation until the end of the different experiments. In this respect, the authors are reproducing in their experiments what happened during blowing snow events with concurrent snowfall. In such conditions, most of the snow transported by the wind is made of precipitating snow particles that fall continuously on the snow surface and are then transported by the wind as illustrated nicely by Figure 3 in the submitted paper. Snow particles initially present at the snow surface may also be transported if the wind speed is sufficient. This situation differs from blowing snow events without snowfall when the transported snow is only made of snow initially present at the snow surface without the constant supply of new snow particles from snowfall. Even if the physical processes involved are the same in terms of particle impacts and ejections, we can expect different densification rates and changes in SSA due the influence of the constant supply of fresh dendritic snow during blowing snow events with concurrent snowfall. Therefore, I strongly recommend to the authors to define well the type of blowing events that they are simulating in the wind tunnel and to discuss how the constant supply of fresh snow influences the results presented in their study. Without such clear discussion, their conclusions may be applied erroneously to different situations that were not captured yet in the RWT experiments.
2. The authors have evaluated two parameterizations used in snowpack schemes for snow densification that account for the effects of wind. This part of the paper is very valuable for snowpack schemes but it should be strongly revised to reflect well how the effects of wind on surface snow properties are included in snowpack schemes and then to make sure that the correct parameterizations are tested in this study.
The effect of wind in snowpack schemes such as Crocus, SNOWPACK and SNOWMODEL consist in a two-step process: (i) the falling snow density generally includes a dependency on wind speed (Pahaut, 1975; Lehning et al., 2002; Liston et al., 2007). It is computed at each model time step based on the meteorological forcing (wind speed, air temperature, …) and new snow is added at the top of the snowpack, (ii) the models then account for wind-driven compaction by including a wind compaction term when calculating compaction in the near surface snow layers (Brun et al., 1997; Liston et al., 2007; Amory et al., 2021; Wever et al., 2022). The wind compaction rate depends on the intensity of aeolian snow transport. This second component allows the models to simulate the increase in surface density during blowing snow events without snowfall (cf my first general comment). So far, in their paper, the authors are evaluating two parameterizations for the density of new snow and are ignoring the second component of wind-driven compaction in numerical snow models. I recommend them to better justify why they are only evaluating the first component. There is a clear grey zone in between these two model components with parameterizations that may overlap and may treat twice the same physical process. Ultimately, we could imagine a clearer separation in snowpack schemes where (i) the snowfall density only depends on temperature and density (to indirectly represent the variability in falling hydrometeors) and (ii) all the wind-related process are all treated by a wind-compaction routine. Under this assumption, the measurements presented in this paper would serve to adjust such wind-compaction routines.
One of my concerns with the current evaluation shown on Figure 8 is that the authors derive densification rates from parameterizations of snowfall density that does not include any temporal component. These parameterizations only provide a value of snowfall density under given meteorological conditions, but they do not specify which time is needed to reach this value (especially for the wind contribution). It would be very valuable to go back to the original datasets that were used to derive these parameterizations and to better understand the temporal aspect. For example, if these values were derived from measurements of snow taken on snow board, are they representative of 1, 3, 6 or 12, 24 hours snow accumulation? Maybe, it could help to understand the large differences in snow density resulting from these parameterizations. In addition, in Fig. 8, which value is taken when computing the initial ice volume fraction for the parameterizations?
Overall, there is no doubt that the data collected in this study will inform the development of improved compaction routines due to aeolian snow transport.
Specific Comments
P1 L9: Add at which height were taken the wind speed measurements?
P1 L10-12: as mentioned above, the main contribution of this paper is to provide a set of very rich measurements to better understand the impact of the wind on the properties of surface snow. I recommend the authors to highlight first in the abstract the main conclusions derived from these measurements before mentioning the comparison with existing parameterizations.
P1 L 18-22: at this stage, the introduction lacks clarity. I recommend the authors to make a better distinction between blowing snow events with and without concurrent snowfall and to describe the types of particles that are transported in these two situations.
P2 L 45-46: the author can refer here to Royer et al. (2021), Wever et al (2022) and Amory et al (2021) to illustrate how the parameterizations of the increase of surface snow density due to wind can be adjusted to better represent the properties of surface snow in the Arctic and in Antarctica.
P3 L 85-90: Section 2.2 explains the detail of the different experiments. Before jumping straight into the detailed description of the experiments, it would be good for the reader to give an overview of what is tested with these 12 experiments.
P 5 P 111-112: at which heights are measured the air temperature and relative humidity in the RWT.
P 6 L 137: Eq. 1 is not described in Lehning et al. (2002). It seems that the ZWART equation has been developed later. Can the authors add a reference? The equation is also described in this supplementary material (Section 4 of https://tc.copernicus.org/articles/17/519/2023/tc-17-519-2023-supplement.pdf).
P 6 L 146-148: if the authors manage to correctly justify why they are evaluating parameterization of falling snow density, the parameterization of Pahaut (1975) implemented in Crocus (Vionnet et al., 2012) could be tested as well. Indeed, it only depends on temperature and wind speed.
P 7 L 168-170: In this paragraph, the authors measure if the particle impact characteristics in the RWT are consistent with natural conditions. They compare their results with the measurements from Sugiura et al . (2000). However, these measurements were also collected in a wind tunnel. Can the authors elaborate on the definition of natural conditions?
P 10 L 198-200: could the authors test the statistical significance of the regression lines shown on Fig. 5a, 5b and 5c?
P 15 L 305-308: it would make sense to propose a fit that respects the physical grounds and tends to zero for very long times.
P 18 L 370-371: it would be interesting to mention that the effect of ambient relative humidity should be tested as well due to its large impact on blowing snow sublimation.
P 20 L 419: would it be possible to write Eq (7) in terms of SSA rate as the previous equations?
Technical Comments
Figures
Figure 3: can the authors add on the three photos the corresponding time stamps as well as a vertical and horizontal scale?
Figure 6: it would be interesting to have the same range of values for the y-axis of Fig 6c and 6d. Otherwise, it seems stronger SSA decreased are measured with the mico-CT.
Figure 6: If one micro_CT measurement has been collected for each experiment, what does represent the error bars shown on Fig 6b and d?
Tables
Table 1: mention in the caption if relative humidity is measured with respect to ice.
Table 1: it would be interesting to know on this table for which experiments micro-CT measurements have been carried out.
References (used in this review and not present in the initial manuscript)
Amory, C., Kittel, C., Le Toumelin, L., Agosta, C., Delhasse, A., Favier, V., & Fettweis, X. (2021). Performance of MAR (v3. 11) in simulating the drifting-snow climate and surface mass balance of Adélie Land, East Antarctica. Geoscientific Model Development, 14(6), 3487-3510.
Pahaut, E.: La métamorphose des cristaux de neige (Snow crystal metamorphosis), Monographies de la Météorologie Nationale, Vol. 96, Météo France, 1975.
Royer, A., Picard, G., Vargel, C., Langlois, A., Gouttevin, I., & Dumont, M. (2021). Improved simulation of arctic circumpolar land area snow properties and soil temperatures. Frontiers in Earth Science, 9, 685140.
Wever, N., Keenan, E., Amory, C., Lehning, M., Sigmund, A., Huwald, H., & Lenaerts, J. T. (2023). Observations and simulations of new snow density in the drifting snow-dominated environment of Antarctica. Journal of Glaciology, 69(276), 823-840.
Citation: https://doi.org/10.5194/tc-2023-112-RC2
Benjamin Walter et al.
Benjamin Walter et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 198 | 63 | 12 | 273 | 6 | 6 |
- HTML: 198
- PDF: 63
- XML: 12
- Total: 273
- BibTeX: 6
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1