Comparison of manual snow water equivalent measurements: questioning the reference for the true SWE value
- Département des sciences du bois et de la forêt, Université Laval, Quebec city (Qc), G1V 0A6, Canada
- Département des sciences du bois et de la forêt, Université Laval, Quebec city (Qc), G1V 0A6, Canada
Abstract. Manual measurement of snow water equivalent (SWE) is still important today for several applications such as hydrological model validation. This measurement can be performed with different types of snow tube sampler or by a snow pit. Although these methods have been performed for several decades, there is an apparent lack of information required to have a consensus regarding the best reference for “true” SWE. We define and estimate the uncertainty and measurement error of different methods of snow pits and snow samplers. Analysis was based upon measurements taken over five consecutive winters (2016–2020) from the same flat and open area. This study compares two snow pit methods and three snow samplers. In addition to including the Standard Federal sampler (SFS), this study documents the first use of two new large diameter samplers, the Hydro-Québec sampler (HQS) and Université Laval sampler (ULS). Large diameter samplers had lowest uncertainty (2.6 to 4.0 %). Snow pit methods had higher uncertainty due to instruments (7.1 to 11.4 %), close to that of the SFS (mean = 10.4 %). Given its larger collected snow volume for estimating SWE and its lower uncertainty, we posit that ULS represents the most appropriate method of reference for “true” SWE. By considering ULS as the reference in calculating mean bias error (MBE), different snow pit methods overestimated SWE by 16.6 to 26.2 %, which was much higher than SFS (8.4 %). This study suggests that large diameter samplers are the best method for estimating “true” SWE.
Maxime Beaudoin-Galaise and Sylvain Jutras
Status: final response (author comments only)
-
CC1: 'Comment on tc-2021-354', Martin Schneebeli, 27 Dec 2021
This study supports previous studies on the subject of "objective SWE". The large sampler they use is very similar to the ETH tube (as used in the extensive survey of Lopez-Moreno et al., 2020). In my opinion, the title of this paper is well-chosen, but I have difficulty recognizing novelty.
- AC3: 'Reply on CC1', Maxime Beaudoin-Galaise, 02 Feb 2022
-
RC1: 'Comment on tc-2021-354', Anonymous Referee #1, 31 Dec 2021
The study compares SWE measurements of three different snow samplers and what the authors call the snow pit method. The authors conclude that large diameter samplers are the best method for estimating “true” SWE.
While the topic of the study is relevant in the field of snow science, I strongly believe the study in the current form is not adequate for TC due to the following reasons:
- The main finding of the study is not new, already others (e.g. org/10.1002/hyp.13785) came to the same conclusion.
- The authors question, already in the title, the reference for the true SWE by assuming the true SWE is the one undertaken with what they call the snow pit method. The snow pit method as it was used in this study uses a density cutter of 250 cm3. As “prove” for their assumption they reference seven studies, of which only two also used similar-sized density cutters. The other studies did not specify a reference or the size of the cutter used or used a much larger cutter (up to 3500 cm3). The volume of the cutter and also its usage horizontally (per layer) or vertically with a plate plays an important role. If used horizontally, as in the current study, the application in a continuous manner is crucial. Sentences like “density measurements were made in each contrasting snow layer that was thicker than 5 cm” leave the impression that these measurements were performed in a subjective manner, which can cause large errors (doi.org/10.5194/tc-10-371-2016) and could explain the partly contrasting results to earlier studies.
- The authors use the same height (h) in their formula 1 und 2, which is definitely wrong as the height of the sampled core can always be different from the height of the snowpack. One reason that h in formula 1 & 2 is different is given in the study by the fact, that for the HQS “it is necessary to insert a plate in a slot at its base to prevent snow loss”, which implicates that not the entire height of the snow pack could be measured.
- Since the height of the snow pack, also in a perfect field like the one at NEIGE site, can spatially vary (due to radiation, wind or rain events) it is important to reference the measured density to a fixed snow height or to specify the uncertainty involved by the varying snow height.
- There is no information given about the type of snowpack (e.g. typical stratification, mean density) or about the distribution of the measured snow heights.
I’d recommend the authors to fully rewrite the study, to focus more on the new ULS snow sampler and to publish in another journal.
- AC1: 'Reply on RC1', Maxime Beaudoin-Galaise, 01 Feb 2022
-
RC2: 'Comment on tc-2021-354', Anonymous Referee #2, 10 Jan 2022
Manuscript presents comparison between three different snow samplers and a density cutter which are used for SWE measurements. Uncertainties and errors are presented and evaluated for the methods. The study results that the large samplers should be used as reference for SWE measurements.
Topic of the study is scientifically relevant, and results are supported by other existing studies. Data set is novel, comprehensive and collected with stable manner ensuring quality. The results are interesting for groups starting SWE measurements and choosing instruments for that, in addition to groups using already one of the instruments or similar ones, or groups, such as modelers, using data from the instruments. However, I would like to see more novelty and progress beyond current understanding in the study.
Snowpack structure and snow properties should be presented to be able to evaluate for which snow conditions the presented results could be applicable. The reference for the true SWE is highly related to experienced snow conditions. For example, large and short samplers are not suitable for very deep and hard snow. It could be stated already in the title that which type of the snowpack is the comparison made for.
As stated in lines 365-367, it is possible to study methods in several snow conditions with the data set. I would like to see these results, how samplers are working in accumulation period and melting period, and what could create additional errors in those conditions, such as increasing amount of ice layers in melting period or crust layers during melt-freeze events in accumulation period.
It would be interesting addition to density sampler comparison if results for different layer properties (e.g. hardness and grain shape classes) would be presented, similarly as for different density ranges. However, a problem with wedge cutter is that the sample includes more snow from the bottom of the sample location than top of it due to shape of the cutter. When samples are taken inside a layer, related error should be small, but layers are still naturally changing gradually. Related uncertainty could be checked. For thin layers, problem with using density from the same grain type is that it may also vary depending on snowpack structure. At least, analysis and description on how densities of the same grain types vary should be added. You could consider also average density from the closest measurements above and below instead of averaging the whole snowpack.
I think that novelty and impact of the study are not strong enough for publishing in The Cryosphere at present. However, I recommend improving the manuscript and publishing in another journal. In the case of significant improvements on the manuscript, resubmission to The Cryosphere could be considered.
Specific comments:
It would be nice to have map and figures from the sampling locations.
Lines 28-29, 32: The first documentations about SWE samplers and snow courses have been published a bit earlier in 1920’s in Europe, but those are quite not possible to find since written in German and not available online. I recommend rephrasing such as “On our knowledge, the first documentation in English…”
Line 29: add 1¾ inch also in cm
Lines 139 and 240: I would recommend using uniformly unit of mm when writing about SWE instead of cm, then it will not mix with snow depth that easily. Now, both units are used which is confusing.
Line 204: “...is the total thickness of all snow layers (other than ice layers) (cm)” Otherwise it might look like thickness also includes ice layers, even though it is written in the next sentence.
Line 379: Also “under similar snow conditions” would require better description earlier on what kind of snow conditions you had.
Line 404: “sections. Although”
Line 408: Chapter 4.2 could be simplified and main points better clarified
Line 455: “drier snow” - newly fallen snow can be also wet (defined by liquid water content). Replace with “This lighter snow”.
Line 520: Replace “in the methods section“ with “in the Section 2 Material and methods”
Lines 560-565: Mention that using large samplers as “true” SWE is also environment related, like in deep snow conditions it is more reasonable to use extendable samplers.
- AC2: 'Reply on RC2', Maxime Beaudoin-Galaise, 01 Feb 2022
-
CC2: 'Comment on tc-2021-354', Charles Fierz, 02 Feb 2022
I agree with the reviewers and the comment by M. Schneebeli that this is a valuable, carefully performed study but also that it mainly supports previous findings. I too cannot find novelty and basically new findings that would warrant title and publication.
In my view the study fails by comparing methods using core samplers with pit methods that are inadequate to determine the water equivalent of snow cover (SWE). Indeed, as mentioned by Anonymous Referee #1, it is of great importance to perform a continuous sampling of the snow cover irrespective of any stratigraphic, subjectively determined layers to obtain SWE. The pit methods presented in the paper are far from complying with this requirement. Even a nice uncertainty and error estimation exercise does not convince me of basically new findings here and, in fact, I am not surprised by the (negative) outcome. On the other side, the comparison of the core snow samplers do support previous findings indeed, thus no novelty there.
I understand from the authors' replies that adjustments to the paper will clarify some open questions, but this will not suffice to overcome the main flaw addressed above.
- AC4: 'Reply on CC2', Maxime Beaudoin-Galaise, 04 Feb 2022
-
CC3: 'Comment on tc-2021-354', Nicholas Kinar, 03 Feb 2022
This paper is an up-to-date and intelligent commentary on differences between manual SWE samplers in a regional context. Many papers on comparisons between manual SWE samples and pit methods have been published in journals associated with snow conferences and most of the comparisons have been made at high mountain locations. In this paper, the authors provide an analysis of comparisons that support other studies, and the data is collected at the Foret Montmorency site, an experimental forest in Eastern Canada. Most papers written in science are not completely novel since scientific research is based on the work of other researchers and there is always a need for validation and verification studies. This paper therefore helps to provide ancillary data to support other studies and provides an important test of different manual samplers at a forest site that is not situated in the mountains. The paper is important in a regional context.
I strongly recommend eventual publication after the authors have addressed review comments and added additional information. The paper compares novel snow samplers such as the Hydro-Québec sampler and the Université Laval sampler and is therefore important for characterizing these new devices, particularly at a forested continental site in North America. The title of the paper can be modified to better communicate this idea.
There can be some additional information added to the paper:
- Wavelet or fractal scaling mathematics can be added to the paper to quantify differences between devices. This would nicely complement the statistics presented by the authors.
- Snow pit sampling is always a subjective process since it is performed by human beings. The authors should indicate in a revision how this subjectivity influenced the study and how the height of snow measurements have uncertainty. Some additional information can also be provided on the regional characteristics of the snowpack. The paper therefore provides a test quantifying this subjectivity.
- Also, it is not possible to individually sample snow layers that have a width < 5 cm, and this should be clearly indicated in the paper, along with a discussion on the sampling practice. I believe that the snowpit was sampled in a continuous fashion, but the current draft of the paper does not clearly communicate this sampling practice.
- Additional information should be added to the paper on how the results are important in a regional context.
- Some additional clarification should be added related to sampling procedures and the use of formulae.
I believe that the study is novel in a regional context and therefore contributes to the literature. The paper is an excellent fit for The Cryosphere journal and without hesitation, I would cite a revised version of the paper in a future review paper written on snow and snowpack processes. I ask the editors to consider the eventual publication of this paper after the authors have addressed some comments. I think that the authors should submit a major revision, but the paper is a valuable contribution to the literature and provides novel data in a regional context.
Dr. Nicholas Kinar
University of Saskatchewan- AC5: 'Reply on CC3', Maxime Beaudoin-Galaise, 09 Feb 2022
Maxime Beaudoin-Galaise and Sylvain Jutras
Maxime Beaudoin-Galaise and Sylvain Jutras
Viewed
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
653 | 232 | 30 | 915 | 12 | 10 |
- HTML: 653
- PDF: 232
- XML: 30
- Total: 915
- BibTeX: 12
- EndNote: 10
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1