Engel et al. use hydrochemistry to assess what the most important runoff-generation mechanisms in glacierized catchments are, and how they change temporally and spatially as a function of geology and climate. They do so by presenting an extensive field campaign performed over the course of two years (2014 and 2015) to sample stable isotopes, electrical conductivity, and element concentration. They interpret these results by leveraging additional streamflow and turbidity data as well as snow, and weather data from one ground-based station at high elevation.
Overall, the research presented here is relevant and in line with the audience and the interests of HESS. That said, I think that the framing, the presentation, and the discussion of these results would need some improvements before publication. Because of all the points I will detail below, I suggest the editor reconsider this manuscript after a minor-revision round. In fact, while the amount of comments is rather substantial, they mostly regard text clarity and do not regard the analyses, which I found quite robust.
A first point that I found quite confusing while reading this manuscript is the lack (at least to me) of an evident research hypothesis that could justify the research methods, frame the choice of the study area in light of the international literature, and thus make the contribution and the message of this manuscript specific enough to fit into one single scientific paper. In the introduction, authors say that “although the effect of catchment characteristics and environmental conditions on stream hydrochemistry at different spatial and temporal scales has well been studied in lowland and mid-land catchments (e.g. Wolock et al., 1997; McGuire et al. 2005; Tetzlaff et al., 2009), only few studies have focused on this aspect in glacierized or permafrost-dominated catchments”. In my understanding, this should be the main knowledge gap that Engel et al. have tried to fill, but at this stage it still reads quite broadly. Thus, they conclude the Introduction by formulating three research questions (see lines 121 – 125).
Of course, understanding the role of geology on the hydrochemical stream signatures over time is relevant and timely, as it is to clarify which are the most important nivo-meteorological indicators driving stream hydrochemistry during the melting period. However, each of these goals is so broad that could be the target of several stand-alone papers. The consequence of setting such diverse goals is that results tend (at least to me) to become a bit dispersive and general (see e.g. the first two lines of the conclusions). A second consequence is that readers (especially the international ones) lack the framework needed to understand, among others, why this extensive campaign in these specific catchments could contribute to global hydrology and hydrochemistry.
So my first major suggestion is that authors could (1) choose one of the three research domains currently introduced at lines 121-125; (2) state and comment the specific hypotheses they would like to test and thus the reasons that led to the choice of this study area; (3) reframe the introduction and the remainder of the paper according to the main key findings with regard to these specific hypotheses; (4) elaborate on the discussion section to expand the implications of this work for other regions of the world where similar studies would be beneficial.
I also found the language of the manuscript sometimes too qualitative. My (personal) opinion is that this is again due – at least partially – to the breadth of the research questions that the manuscript is trying to answer. A few examples, with my comments in square brackets:
“Results highlight the dominant impact of water enriched in solutes during baseflow conditions starting from late autumn to early spring prior to the onset of the melting period in May/June of both years [is there any way you can replace this “dominant role” with something more quantitative and specific? This also sounds more like a discussion sentence and I would expect results to focus on metrics that could quantify this impact rather than saying that “they highlight the dominant impact”]. Such an impact seemed to be highest [how did you measure this? Consider replacing “seemed” with something more definitive] in water from streams and tributaries reaching the most increased conductivity [I would explicitly mention numbers here rather than “the most increased”] at S6 during the study period compared to all sampled water types, ranging from 967 to 992 μS cm-1 in January to March 2015. During the same period of time, isotopic composition was slightly more enriched [how much?] and spatially more homogeneous [how much?] among the stream, tributaries, and springs than in the summer months.” (lines 294 – 300);
“In contrast, the Sulden River revealed relatively high EC [how much? Relatively to what benchmark?] at the highest upstream location (S6) and relatively low EC upstream [same as before] the confluence with the Trafoi River (S2) during baseflow conditions. The exponential decrease in EC (‘EC dilution gradient’) during this period of time was strongly linked to the catchment area [how did you measure this strong link?]” (lines 326 - 329) ;
“Furthermore, the interannual variability of meteorological conditions with respect to the occurrence of warm days [warm is relative, I would replace with numbers – maybe days with avg temperature greater than 0 degC or 5 degC as done for the snow-melt analysis?], storm events and snow cover of the contrasting years 2014 and 2015 is clearly visible [this is also relative, consider measuring with some metrics] and contributed to the hydrochemical dynamics (Fig.8 and Table 1). (…) In contrast, warmer days in 2014 were less pronounced and frequent [provide statistics] but accompanied by intense storms of up to 50 mm d-1. These meteorological conditions seem to contribute [I would replace this with something more definitive and informative] to the general hydrochemical patterns described above. (lines 403 – 417).
SPECIFIC COMMENTS (each of these comments start with the line number to which it refers)
- 60: what does “this objective” refer to here?
- 65-66: what do you mean with “topography with drainage network”?
- 74: maybe replace “address” with “quantify” or “clarify”?
- 77: to me, streamflow would (at least partially) correlate with air temperature even in other circumstances, e.g., Mediterranean catchments were temperature-driven ET is an important driver of water supply.
- 115: it was difficult to me to understand what is the specific “gap” that you aim to address here. If this is what is reported at lines 101-106, then the paragraph about permafrost makes the link misleading as it breaks the flow of information.
- 115: two year -> two-year
- 136: how does this glacier area in 2006 compare with more recent estimates from, e.g., the Randolph Glacier Inventory v6 released in July 2017?
- 140-143: it would be helpful to include more information on geological properties that are intuitive for a general audience, such as permeability or percentage of clay and silt in the soil layer (if at all available). These properties would help to relate these geological groups with expected infiltration patterns and thus runoff response.
- 158: maybe this was already discussed during the first round of revision, but could you comment on the expected representativeness of this station for both sub-basins?
- 178: could you quantify what do you mean with “very limited”? That is, could you provide statistics to make this more informative?
- 228ff: assuming snow-depth decreases as a proxy of snow melt is a simplified approach, but authors are clear on this point (see also the discussion section). As a side note, I would suggest authors convert the Delta SD data reported throughout the manuscript (e.g., see Figure 6) into snow-melt runoff, which can be estimated from Delta SD via an assumption on snow density. The advantage is that, in the snow-hydrology literature, snow-melt runoff is usually assumed positive as it is an input to the stream network (the larger, the more snow has melted). This would make result interpretation a bit easier to follow (e.g., snow-melt runoff would increase with radiation or air temperature in Fig. 6 as a diagonal reader would expect).
- 245: I may have missed how baseflow and melt periods were defined.
- 256: remove “the” before “this”
- Tables 4 and 5 are quite challenging to screen and understand, especially for diagonal readers. What about replacing them with something like a boxplot of VC where heavy metals and other elements are depicted with different colors, and move these tables to a supplement? Sounds like the main point here is the difference in chemical composition during snow-melt and baseflow, something that VC should easily measure (and indeed, VC is the main metric used to make this point in this section). I also found difficult to understand how “The observed geochemical patterns are confirmed by PCA results (Fig. 2) and the correlation matrix (Fig. 3)” – maybe a few words on this could be helpful.
- 363: passed -> exceeded
- Fig. 6 vs. 7: in Fig. 7, the range of snow-depth differences spans -150 cm and -50, but in Fig. 6 the minimum difference is about -80 cm. Am I missing something here?
- Fig. 8: the color of the line for turbidity is not clear to me
- 402: what flood are you referring to here?
- 410: what is the reference for this lapse rate?
- 426: many studies on rain-on-snow events set a minimum snow-depth threshold to define a rain event as a rain-on-snow event, especially because snow tends to be patchy when it is shallow. Could you comment on this in the manuscript?
- 431: again, replace “was more variable” and “slightly increased” with some quantitative statements.
- 451: how did you quantitatively conclude that the EC-discharge relationship was “the strongest”?
- 484: replace “probably needs to be excluded” with “was excluded” if results support this.
- 493: replace this with the actual concentrations. Same at line 494 (how much more enriched?)
- 520: replace “it is more likely” with something more informative (or just remove it)
- 565 – 566: could some of these factors be addressed just based on a DEM and some assumptions on radiation distribution, as often done in hydrology to distribute radiation across the landscape?
- 585 – 588: could you be more specific here with regard to how “Tracer dynamics of EC and stable isotopes associated with monthly discharge variations generally followed the conceptual model of the seasonal evolution of streamflow contributions, as described for catchments with a glacierized area of 17 % (Penna et al. 2017) and 30 % (Schmieder et al. 2017”? in other words, could you replace this interpretative statement with some quantitative results that could allow one to understand “how” and “how much” observed dynamics followed the conceptual models? Also, could you quantify what you mean with “isotopic dynamics were generally less pronounced compared to these studies”?
- 646-648: this sentence seems recursive to me
- 682ff: I think no field work will be ever able to capture all potential variability of hydrologic processes. If the representativeness of this campaign is something that should be discussed, that this should be done in greater details and probably earlier in the manuscript. |