This paper investigates the impact of a dry and warm winter on the snow dynamics in a discontinuous boreal forest in northeast Canada. Comparing observations of snow dynamics in a low-snow winter with a winter close to normal conditions can give insights about expected future changes. In boreal forests, the snow dynamics differ between open gaps and under the canopy. The study uses observations at three nearby locations (under the canopy, small gap, large gap) in a small catchment in northeastern Canada. Measurements of snow physical properties, its thermal regime, and soil measurements were taken over two consecutive winters, which represent low-snow and normal conditions. Results show less snow accumulation and an earlier melt, which was slower due to lower radiative forcing, in the low-snow year. The topsoil layers were cooler and under the canopy soil freezing was enhanced in the warmer year. The spring freshet occurred earlier but was less intense, due to slower melt.

Generally, this is a well-written paper, which presents a lot of interesting observational data on various aspects of snow dynamics accompanied by relevant figures. The paper is well structured. However, some discussion about the limitations of the approach to give insights into future changes are missing and the second objective is not addressed in-depth.

Specific Comments: Major

Exceptionally dry year

At your study site, the winter 20/21 was exceptionally warm AND dry. You mention that it is “plausibly representative of future winters” (l.23). In the introduction you described the expected warming levels in boreal forests. However, I missed an introduction to how future precipitation is projected to change for boreal forests and eastern Canada. From the statement in l. 39 I assume, that annual precipitation is projected to increase. Is winter precipitation also expected to increase?

If future winter precipitation is projected to increase, the winter 20/21 is not representative of the projected future. I would have expected a discussion on this and how this impacts the conclusions you can draw from your observational study for future changes in snow dynamics and runoff.  How do your results differ from what you would expect with climate change? In l. 54 you state that more frequent and intense winter rainfalls are expected with climate change. Such increased winter rainfall could lead to more rain on snow events, especially at the beginning and end of the winter, which likely influences the discharge. Could you please elaborate on the above aspects in your discussion?

Evaluation of the effect of snow dynamics on spring runoff

In the abstract you mention a research gap: “Although the effects of warmer winters on snow-related processes are well documented, their interactions to influence the spring runoff in evergreen forest remain poorly understood.” (l.17-19). It sounds like this is one of the two research gaps you would like to address in your study, which is made clear in the introduction: “The second one is to evaluate how these factors interact together to modulate spring runoff.” (l.71). From the sentence in the abstract and the objective I would expect that you look at the interactions of several processes to distinguish their individual influence on spring runoff. However, in the results, you show discharge measurements only in relation to air temperature and SWE changes. You do not consider, how individual factors influence the SWE changes and the discharge, such as the increase in snow permeability and the soil freezing. You do mention that the decrease in available energy in the melt period probably decreases the magnitude of the spring freshet, but the relationship between, e.g. the effect of the earlier onset of the melt season in relation to the infiltration vs. surface runoff of snow melt remains unclear. To estimate the effect of increased snow permeability and soil freezing on the spring freshet, you would need information about the partitioning between infiltration and the surface runoff.

I would expect a more in-depth discussion about the limitations of achieving objective 2, e.g. why you did not measure infiltration and surface runoff.

Moreover, regarding my first comment on the exceptionally dry winter, I would expect a discussion about how spring runoff is affected if winter precipitation increases and rain on snow events increase.

The second objective in general is addressed in much less detail in your study than the first one, which is addressed extensively. For example, in the methodology, it is not clearly introduced which methods are used to achieve objective 2. Also, the title of the paper only encompasses the first objective. Therefore, I suggest regarding the evaluation of the effect of snow dynamics on spring runoff not as a second objective, but rather as a further analysis and frame the paper accordingly.

Introduction: clearly identify the research gap

In general, I think the paper is well-written and well-structured. However, I struggle a bit with the introduction, which could be more concise and better structured, I think. The introduction about soil freezing is very long in comparison to the introduction of other background information and processes. I also had difficulties in identifying the exact research gap you would like to address based on the introduction. It remains unclear to me whether changes in snow dynamics in relation to warmer winters are known in boreal forests or in other biomes in general. The statement in line 39-40 contradicts the statement in l. 18 (“effects of warmer winters on snow-related processes are well documented”).  This makes it difficult for the reader to understand what similar relevant research has been done and to identify the existing research gaps you aim to address with this study. The last part of the introduction is very well written (l.69-82).

Snow Stratigraphy results:

• l. 335: “At all sites, there was a greater proportion of faceted crystals (FC) and depth hoar (DH) in the snowpack during the low-snow year than during the reference year”. However, I see from Figure 10 that the portion of light blue and dark blue colour (FC and DH) is smaller in the low-snow year than the reference year for the canopy.
• l. 336: “In contrast, we observed fewer rounded grains (RG) in W20–21 than in W21–22.” This does not seem to be the case for the canopy looking at Figure 10.
• l. 336-337: “In both years, FC and DH layers were proportionally thicker under the canopy than in the gaps.” From Figure 10, I see that the light blue color (FC) covers a smaller proportion of the snow height in the canopy than in the gaps, which contradicts your statement. Maybe you mean that FC and DH layers combined were proportionally thicker?

 Specific Comments: Minor

• l. 27: warm year instead of warmest year
• l. 27: “Overall, we observe that the spring streamflow discharge was significantly reduced in the warmest year due to a slower melt and low precipitation in April and May.” I would argue that it is mostly reduced due to less snow accumulation in winter and thus less snow melt that can contribute to the spring freshet. Why did you not mention this aspect?
• l. 37 “rather dry regions” and l. 39 “humid boreal forest”: What is rather dry and humid in the context of a boreal forest? Can you provide a definition? Do you expect different behaviors?
• l. 100: “The stations were located in the vicinity” Please include the distance to the flux tower
• l. 130: lowest -> lower, highest -> higher, otherwise confusing it if is really only to probes
• l. 166: the subscript should be “i”, I think, but “l” is used
• l. 233, 234: It would be helpful for the reader if you can also give the precipitation anomaly in percent in comparison to long-term mean.
• l. Line 265, 315: how do you define the onset of snowmelt?
• l.  230, 255, 283, 293, 309, 321, 331, 369: You always use the same sentence structure: “Figure X shows …”. These sentences basically repeat what can be seen in the figure caption. Stating the same in the main text is not necessary. To make the text more concise, I suggest removing these sentences and referring to the figures after the first statement about the results shown in the figure, e.g: W20–21 was the driest winter of the 1982–2022 period, with 199 mm recorded from January to April (JFMA), including 167 mm of solid precipitation (Fig.2, Table 4).

• l. 362: in “in April to June” instead of “April to May”
• l. 370-371: Could you give the runoff in mm/d or m3/s instead of total m3? Normally, in hydrology we use either mm/d or m3/s as units for runoff, as these are easier to grasp.
• l. 476: “precipitation” change to “liquid precipitation”
• l. 389: “Their relative size shows the importance of the process between the gaps and the subcanopy locations.” What do you mean by importance? Do you mean the magnitude? Please make this clear.
• l. 390: “Large black arrows are applied all three locations.” This is unclear. I think you mean that this analysis is not made for the three sites but rather for the larger catchment (as data from NEIGE station and discharge gauge at outlet is used).
• l. 423: “with canopy closure”. It is unclear to me what you mean with this, please rephrase to make it clear.
• l. 450: For the other discussion sections you used statements as titles which makes it easy for the reader to grasp the main point. Can you also do this here?
• l. 523-525: Can you elaborate on this statement. Was this expected? I would expect that in most years the precipitation and temperature conditions drive how much snow can accumulate and when it melts and this drives the spring freshet: Or are there examples where soil refreezing drives the spring freshet? Moreover, to me, it is not clear what you consider under “weather conditions” and “snow characteristic” in this context. I would think the amount of snow (SWE) belongs to snow characteristics, however, it depends on the weather conditions and influences the spring freshet.

Figures

• Figure 1c): It seems like a fish-eye perspective, but could you put a scale bar here, so the size is clear?
• Figure 1a) DEM color is not so color-blind friendly
• Figure 2a): cumulative precipitation plots maybe better to show what you want?
• Figure 5: It is quite difficult to compare the two years to each other and see which values are larger, especially for c) and d). Plotting both years in a single plot would make the comparison easier for the reader.
• Figure 6: very nice plot, very easy to grasp!
• Figure 7: Also here it is quite difficult to compare the two years and see the differences between the two years. You can plot both years in the same plot, by using different line styles (solid, dashed).
• Figure 11cd: y axis label should be mm/d I guess.
• Figure 11: You compare the liquid precipitation over the whole catchment to the SWE averaged over the stations. Can you elaborate on whether the SWE averaged over the stations is representative of the whole catchment? Is the experimental site located at a representative elevation for the catchment?
• Figure 12: Very nice to have an overview figure of all results. A legend of what the colors mean is missing. The black arrows at temp, snowfall, precip. and discharge can be easily confused with the other arrows that just show the relationships. Using another color (maybe grey) for the large black arrows would help. Also,

Paper # https://hess.copernicus.org/preprints/hess-2023-191/

How does a warm and low-snow winter impact the snow cover dynamics in a humid and discontinuous boreal forest? An observational study in eastern Canada

Bouchard et al.

This manuscript provides a data-rich description of two winters at Montmorency Forest (Québec, Canada) with contrasting meteorological conditions. The manuscript is interesting, easy to read and data are presented in thorough and clear manner. However, I have a number of comments that the authors may wish to consider to help guide the clarity and purpose of the messages in this paper, for the wider community.

Major comments:

The authors do a good job to realize their first objective – to quantify and compare the effect of snow under forest canopy and in canopy gaps on soil properties around the phase boundary and snow properties. This creates a thorough descriptive narrative, but which very largely reinforces what we already know, and struggles to justifiably generalize beyond the study site. Snow is well known to have a very important influence on insulating the relatively cold winter air temperatures from warmer soil. Very broadly, shallow snow means cooler soils and vice versa. Slater et al. 2017 (doi:10.5194/tc-11-989-2017) demonstrated that at effective mean snow depths of 50 cm the influence of the atmosphere on soil temperatures decouples. Hence shallow sub-canopy snow at Montmorency (< 50 cm) has a bigger influence on the variability of soil temperatures relative to deeper snow in gaps. The snow properties (effective conductivity) go a little way to mediating this influence, but are secondary in importance to the magnitude of the snow depth. In addition, earlier snowmelt meaning slower melt rates due to lower incoming shortwave just reinforces the point made by Musselman et al 2017 (as cited) and the lower SWE leading to lower stream discharge is intuitive.

While comparisons between snow and soil properties in sub-canopy and forest gaps are consistently made, the explanation of these differences is often missing and makes the discussion highly speculative. This appears in the discussion section where the language used often relies on ‘would imply’, ‘seemed to’, ‘may have’, ‘could be’ or ‘suggests that’ to dilute the strength of conclusions that can be drawn. An atmosphere-forest-snow-soil model would allow the quantified explanation of processes that govern the observed snow/soil properties outcomes. While I respect the authors right to control the narrative, which is currently clearly expressed that this is an observational case study, the lack of a modelling approach hugely limits the capacity to quantifiably explain key processes and help to generalize beyond this catchment and beyond the two winters presented. In particular, the capacity to explore the forest canopy impact on interacting energy and mass balance processes (second objective on ln 71) would be unlocked. You have presented a fantastic snow and soil dataset, you have good forest canopy structure data using HPEval, so by including some process modelling (e.g. CRHM for hydrology or Crocus with a canopy model for snow properties) you would be able to more adeptly justify your explanations. For example:

• ln 421-422 states earlier melt onset suggests net radiation is lower, but this is not shown and the impact of sensible heat fluxes are not considered? Even in the sub-canopy where turbulence is lower, when the air temperatures go above zero then sensible heat fluxes can have a significant effect.
• On ln 428-429 you state that the structure of the canopy must be considered in models. I fully agree, and here but you have the capacity to show this in a model and address your own statement.
• A modelling approach would go some way to explain why heat loss was sufficient to favor soil freezing under a canopy but not in gaps (ln 434-435) – I would expect net LW to be important here, but comparison of modeled fluxes would allow a more robust analysis.
• If you couldn’t manually assign an albedo class, which would be hard to do for purposes of spatial and temporal generalization, how would a modeled estimate of albedo affect the relative impact of energy and mass balance processes?

These are just a non-exhaustive number of examples where I feel inclusion of a simple modelling approach would allow the speculative areas of the discussion to be either removed or better justified.

Much is made of winter 20-21 being analogous to a warmer climate, but while we should expect warmer winters in a warming climate, it is much less understood as to the impact on winter precipitation. This would have an impact on the mass of snow (see earlier comment) and the potential for rain on snow. Consequently, this could benefit from a much more robust underpinning using future model projections of climate (e.g. NA-CORDEX) to show not just where 20-21 fits within past measurements (nicely shown in Table 4), but where it lies in the future. Some big statements are currently being made (e.g. ln 464-465) about how snow thickness could override impacts of increased air temperatures. This could benefit from a more solid foundation in future climate projections.

The second objective features much less in the manuscript and is much more speculative (in its current form without hydrological modelling). It may be that this could be sacrificed in a revised manuscript in order to improve the focus on objective one?

I appreciate the main thrust of my major comments (to include a modelling component) is non-trivial. Hence for now I will restrict the minor comments to a few obvious changes (also because the manuscript is well written and has very few obvious minor issues):

Ln 310-311: Are the soil profile characterizations (inc. porosity) shown anywhere? These may be important when considering soil hydraulics and thermal transfer.

Ln 52; delete ‘17-19 May’

Ln 168: ‘measurement’ rather than ‘measurements’.

Ln 272: April rather than Avril.

What evidence is there for the forest being humid? Particularly in the winter? No humidity measurements are presented or referred to.