Observations of capillary barriers and preferential flow in layered snow during cold laboratory experiments

Avanzi, Francesco; Hirashima, Hiroyuki; Yamaguchi, Satoru; Katsushima, Takafumi; De Michele, Carlo

doi:https://doi.org/10.5194/tc-10-2013-2016

Articles | Volume 10, issue 5

https://doi.org/10.5194/tc-10-2013-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/tc-10-2013-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 10, issue 5

Research article

|

09 Sep 2016

Research article |

| 09 Sep 2016

Observations of capillary barriers and preferential flow in layered snow during cold laboratory experiments

Francesco Avanzi, Hiroyuki Hirashima, Satoru Yamaguchi, Takafumi Katsushima, and Carlo De Michele

Download

Final revised paper (published on 09 Sep 2016)
Preprint (discussion started on 03 Dec 2015)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC C2881: 'Review of "Laboratory-based observations of capillary barriers and preferential flow in layered snow" by F. Avanzi, H. Hirashima, S. Yamaguchi, T. Katsushima, and C. De Michele', Anonymous Referee #1, 27 Jan 2016
- AC C3321: 'Response to Referee #1', Francesco Avanzi, 25 Mar 2016
RC C2938: 'Review on Laboratory-based observations of capillary barriers and preferential flow in layered snow', Christoph Mitterer, 03 Feb 2016
- AC C3323: 'Response to Dr. Mitterer - Referee #2', Francesco Avanzi, 25 Mar 2016

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Francesco Avanzi on behalf of the Authors (28 May 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (07 Jun 2016) by Guillaume Chambon

RR by Anonymous Referee #1 (09 Jun 2016)

Suggestions for revision or reasons for rejection

The authors did a good job in revising the manuscript. Almost all of my comments are incorporated, or where rebutted in the response to reviewer document. I think the manuscript now provides a much clearer contribution to the literature about liquid water flow in snow and I can recommend publication after a minor revision, as I think there are still a few issues where the manuscript can and should be improved. But this should be a minor effort of the authors to deal with my comments.

I have two "major" concerns:

1) The results are probably too much interpreted in terms of water entry suction only. In the implementation of Richards equation in the SNOWPACK model, the water entry suction is not considered, yet SNOWPACK is able to reproduce water ponding on microstructural transitions. So other factors that play a role are (i) that even in an equilibrium situation (no flow), the pressure head decreases with height above the ground only. Then, in fine grained snow, a particular pressure head will be associated with a higher LWC than in the coarse grained snow below, which one could interpret as ponding. This effect is also present during non-equilibrium flow. Note that it implies that interpreting a higher LWC in one layer as ponding due to the pressure head being below the water entry suction of the layer below is therefore also not justified. This should be interpreted in terms of pressure head. (ii) the capillary barrier will also introduce a jump in hydraulic conductivity, particularly if the snow below the barrier is dry. That means that a water accumulation can also arise when water is transported to the capillary barrier from above faster than can be transported away below the barrier. So I'm not so sure that all the ponding observed in the experiments can be uniquely explained in terms of water entry suction. Or can the authors argue that the water entry pressure is a kind of net description of the other processes I mentioned (i.e., difference in LWC for same pressure head and gradients in hydraulic conductivity)?

As the authors are very knowledgeable on this topic, they can maybe provide clarity to the snow science community at this point. I hope they can either rebut this point, or that they go once more through the manuscript to make sure that explanations provided are interpreted as representing the correct process and an appropriate discussion of causes for ponding on capillary barriers is provided.

2) It is somewhat inconvenient that the model simulations are presented differently from the previous version of the manuscript and that the discussion paper will remain public. I'm aware that the other reviewer was asking to repeat the simulations using the latest version of SNOWPACK and I'm glad to see that the latest SNOWPACK version is giving results where the authors are happy with. Actually, it looks like the new SNOWPACK version is doing a better job for the MC samples (Figures 5g, h and i). Do the authors agree that the results seem better than with the old version of SNOWPACK? It is only a bit strange that the authors now found it necessary to plot 2 profiles, and only show a part of them, whereas in the discussion paper, full profiles are shown at one point in time. I think it is important to show full LWC profiles, even though it is well argued why they choose the 2 profiles from different moments of time in the simulations. At least I feel it is as necessary that the authors include a sentence or two of how the new and old SNOWPACK simulations compare, as the old SNOWPACK simulations are now public in the discussion paper.

Some minor and technical issues:
P1L4: "to a constant supply"
P3/4L94: Maybe it is a good idea to state the research goals a bit more clear in the introduction. As the text is now, it all doesn't sound that ambitious. I therefore suggest that the authors amend that they want to obtain quantitative information about the liquid water flow over a capillary barrier. And instead of just writing that you compare the results with SNOWPACK models, I suggest to word this as something like (P4L102): "All laboratory experiments ... SNOWPACK model, in order to investigate how well 1D snowpack models representing liquid water flow are able to capture the behaviour of liquid water flow over capillary barriers."
P4L106: This is not so well argued. On P 5L141, you write that you used dry snow. That means that in your experiments you also will have to deal with a mixture of dry and wet snow. Best is to reformulate P4L106.
P4L127-129: In my original review, I was thinking that it is relatively easy to address this issue, as the saturated hydraulic conductivity of snow is generally very high, well above natural water fluxes in snow, and well above the water influx rates you use in your experiments. It sounds a bit weak that you did not consider any instability criterion. For me, you can just write that due to the high saturated hydraulic conductivity of snow compared to the water influx rate in your experiments, most instability criterions will predict unstable flow, which is confirmed by Katsushima et al. 2013. But this is up to the authors.
P6L179: "by a operation model". What do you mean? "by an operational model"? Also not clear what defines an operational model. Do you intend to say that it is a model that incorporates most processes related to snow and can be applied to natural snow covers?
P6L185: Probably better to cite Coleou and Lesaffre here, which defines the bucket scheme used in SNOWPACK. [Coléou, C. and Lesaffre, B.: Irreducible water saturation in snow: experimental results in a cold laboratory, Ann. Glaciol., 26, 64–68, 1998.]
P6L172: "Clearly...software.". I'm not sure if the authors included this based on my first review, but the sentence feels a bit too obvious, out of place and may be removed.
P8L261: "spatially restricted" well, actually it is almost no water spreading...
P8L267-269: This part is really hard to follow now. I think "Conversely" should be "In contrast". Then you can write: "In contrast, psi in fine snow for 5% and 10% LWC is 0.22 and 0.21, respectively and psi in medium snow for 5% and 10% LWC is 0.09 and 0.08, respectively. This implies that for typical low saturation values in snow, the difference in suction pressure in the finer snow with the water entry pressure of the courser snow is larger for fine snow." Or something similar.
P9301L: "other hints" -> I suggest "Other indications for the spatial variability are the observed spatial variability ..."
P9L303: "as this is a signature of LWC" -> "which can be linked to/attributed to/interpreted as differences in LWC"
P10L316: "say" -> "maximally"
P10L338-339: Actually, I think for a good investigation of the relationship of f with W, also larger containers need to be used, as now, with the 5cm rings, preferential flow tend to form at the boundaries of the sample, so there seems to be a boundary effect here (see Figure 1). When the authors agree, they can mention this here. This point may also be mentioned in the comparison section with SNOWPACK, where discrepancies are discussed. Boundary effects may also explain discrepancies between model and laboratory experiments.
P11L348: "less expected" -> I had the impression that the authors were arguing at some point that capillary barriers may influence the travel time, as the progress of the melt water front is temporarily blocked at the capillary barrier. Why is it then "less expected"?
P14L457: "basing" -> "based"
Conclusions: Some people only read conclusions, so maybe replace FC with fine-over-coarse, etc here.
P14L470: "point conditions" is a bit confusing. It may also refer to a point in 3D, but I guess the authors mean here the 1D representation in the SNOWPACK model?
P14L469-470: I generally don't like a sentence like this in the conclusions. It is the conclusion section after all. So please state the outcome (so, the conclusion!) of the discussion; a sentence like: "The comparison of observed and simulated LWC revealed this, this and this."

In response to the reply to reviewers:
The value of 1.3% found by Waldner et al. (2004) was originally used as a comparison with the laboratory experiments. The authors originally interpreted the value as 13% and made a comparison with their experiments. Now that it is clear that the value is 1.3%, the discussion is dropped by the authors. It is not clear now how this measurement relate to the work of the authors. I think it can be attributed to differences in the characteristics of the snow used in the experiments as well as the applied water flux, and the experiment duration? This measurement by Waldner et al. (2004) may still be important to be discussed in the manuscript, given the limited experimental data on capillary barriers.

Hide

RR by Christoph Mitterer (21 Jul 2016)

ED: Publish subject to minor revisions (Editor review) (23 Jul 2016) by Guillaume Chambon

AR by Francesco Avanzi on behalf of the Authors (02 Aug 2016) Author's response Manuscript

ED: Publish subject to technical corrections (10 Aug 2016) by Guillaume Chambon

AR by Francesco Avanzi on behalf of the Authors (14 Aug 2016) Manuscript

Short summary

We investigate capillary barriers and preferential flow in layered snow during nine cold laboratory experiments. The dynamics of each sample were replicated solving Richards equation within the 1-D multi-layer physically based SNOWPACK model. Results show that both processes affect the speed of water infiltration in stratified snow and are marked by a high degree of spatial variability at cm scale and complex 3-D patterns.