Hydrological response to warm and dry weather: do glaciers compensate?

compensate? Marit Van Tiel1, Anne F. Van Loon2, Jan Seibert3, and Kerstin Stahl1 1Environmental Hydrological Systems, Faculty of Environment and Natural Resources, University of Freiburg, Germany 2Institute for Environmental Studies, Vrije Universiteit Amsterdam, the Netherlands 3Department of Geography, University of Zurich, Switzerland Correspondence: Marit van Tiel (marit.van.tiel@hydrology.uni-freiburg.de)

-Introduction: I like the structure/storyline and content of the introduction, maybe it is a bit (too) long and could be shortened without losing important information and content? Sometimes I wondered whether the difference between "the buffering effect of glaciers" and "the compensation effect of glaciers" is always clear (or if there is any big difference at all…). Maybe you could check that point while going through the introduction (or the whole manuscript) again? Moreover, please find a number of specific comments which hopefully help to improve specific parts of the introduction below… -Data: I have, unfortunately, some concerns about how you deal with and use measured mass balance data in your study; you cannot just take a median of measured glacier mass balance data, you have to weight the measured data with glacier area, i.e. calculate area-weighted average mass balances, otherwise you might get wrong results. This is relevant when it comes to compare or correlate mass balance with levels of compensation C, I clearly think you need to check and clarify that; moreover, it does not really make sense to me to just use "country-wide" mass balance data for individual catchments you analyze, there are better ways to extrapolate measured mass balance data to individual catchments, especially in areas with a high density of measured mass balance data like Norway, Switzerland, Austria. Please see also specific comments thereupon below.
-Results: In section 4.2 you only refer to WD (or WWD) events, right (CD, and ND events are excluded)? Maybe make that clearer in the text (e.g. by changing the title of the section to "4.2 Glacier compensation during WD events in different catchments"); it was not very much clear to me why the chosen catchments and years for individual regions were selected for Figure 3 (just an example or do these three years and catchments reflect some of the "general observations, trends, and C values over the whole observation period"?)… -Results: Section 4.4 and figure 9: As I understood how you calculate catchment-wide mass balance or take into account measured mass balance data to correlate it with levels of compensation C, in my opinion this is not quite correct (see comments above and specific comments thereupon below), maybe by correctly calculate catchmentwide mass balance your resulting correlations of mass balance with levels of compensation would quite change and show another relation between these variables?!...
-Results: Still section 4.4 and figure 10: To be honest, I didn't really understand how you calculated correlations between changes in C and glacier changes over time (mostly retreat for your period of observation, but with intermittend phases of readvance, for instance for Norwegian glaciers in the 1990s!), this is neither very clear from the methods section nor very clear in the results section… what data sets did you take into account to check if glaciers in the analyzed catchments did change in area and volume over your observation period?this aspect, relating also to figure 10, is not very clear… -In the whole manuscript: I think it would make more sense to use "southwestern Norway" instead of "Norway" and "western Canada" instead of "Canada" in the entire manuscript (these countries are so large and you analyze only catchments in specific regions)…

Specific comments and technical corrections
To facilitate the author's correction of the manuscript I combined specific comments and technical corrections (including language or spelling and comprehensibility issues). Sometimes comments contain both specific comments and technical corrections, sometimes just one or the other. I would ask the authors to implement my comments and suggestions as far as possible.

Introduction
Ln 22: "…alter the input of hydrological systems…" "…alter the water input of hydrological systems…"?
Ln 30: "In high mountain regions, where snow and ice are present, these snow and ice storages…" repetition of "snow and ice", maybe rephrase for smoother readability?
Ln 31: "…, because of temperature-driven water supply." why not directly write what you refer to here? i.e. enhanced snow and ice melt due to high temperatures?...
Ln 34: "…these storages can also be themselves be depleted…" one "be" too much "…these storages can also be depleted themselves…" Lns 42f: "Hence, groundwater and snow storages might not always be a perfect buffer during warm and dry periods." Please bee a bit clearer and more precise here, I guess what you want to say here is something like "Hence, groundwater only has a limited buffering capacity (in terms of runoff, provides only baseflow), while the buffering capacity of snow is higher (in terms of additional runoff) but temporally limited (if all snow has melted, there is no buffer anymore).
Ln 55: "…from other streamflow contributions such as snowmelt." maybe add some additional important streamflow contributing processes (in high mountain areas) here "…from other streamflow contributions such as snowmelt, surface runoff, interflow, groundwater flow or melting permafrost."?!?! Ln 58: "…during such extreme warm and dry periods." why "such"?
Lns 60ff: "During warm and dry years, glaciers can provide more meltwater to streamflow, and during cold and wet years they generate less meltwater so that altogether the interannual streamflow variability is relatively low." do the studies you refer to here give a "lower threshold ratio of glacierization" for individual catchments (i.e. minimum percentage of glaciercovered area in a catchment) from which the dampening effect of glaciers on interannual streamflow variability applies? -would be interesting added value here… Ln 64: "The result of the balancing between melt and precipitation is assumed…" In order to be clearer, I'd rather write "Whether the amount of runoff from melting glaciers has a dampening effect on the interannual streamflow variability or not is assumed…" Lns 66f: "…but also other climate and catchment characteristics appear to influence the streamflow sensitivity to climatic anomalies…" other climate and catchment characteristics like what?can you give some examples mentioned in the studies you refer to here?
Ln 69: "…mountain glaciers have been retreating…" "…mountain glaciers around the globe have been retreating… Ln 72: "…provided an additional source of water in the summer compared to the seasonally delayed contribution." why "COMPARED TO the seasonally delayed contribution"? rephrase in order to be clearer… Ln 73: "However, this source will not be sustained" "However, additional meltwater from shrinking glaciers will not be sustained forever." Ln 74: "…some regions…" "…some glacierized mountain areas…" Ln 78: "…differs in different catchments…"
Lns 80ff: "However, this contribution may reach a maximum, as the higher glacierized catchments are generally located at higher elevations that receive more orographic precipitation amounts…" can you please check that statement? In my opinion, it is not primarily the increased amounts of precipitation that causes a maximum value of relative catchment glacier cover in terms of increased streamflow contribution by glacier melt, but rather air temperature and climatic conditions in highly glacierized catchments, isn't it?
Ln 83: "…is generally assumed to be highest in August and September, …" of course, this is only true for mountain glaciers in the northern hemisphere with a mass balance regime similar to the one in the Alps… I am sure you are aware of that but it might be good to be more precise here… (as your study is interesting to people from all over the world ;-)) Ln 85: "…relative contribution is, …" "…relative glacier melt contribution is, …" Ln 96: "This scale enables…" "The chosen time scale enables…"?
Ln 97: "Moreover, this scale…" "Moreover, this time scale…" Lns 99f: "…we analyzed observed hydrological responses to WD events for catchments with varying glacier cover in Norway, Canada, Switzerland and Austria." I recommend to add some information about the observation periods / temporal time frame of your study here.

Data and hydroclimatology of selected glacierized catchments
Ln 108: "…based on length of the time series (long records), …" "…based on the length of available data time series (long records), …" Ln 109: "…missing values…" "…data gaps…" Ln 109: "…(including low and high glacierized catchments)…" …"(including catchments of low to high relative glacier cover)…" Lns 109f: "A few of these catchments are nested." What does this exactly mean here? -Can you maybe briefly explain in key words in parentheses after the sentence?
Ln 115: "…for the different catchments differed, between 50 and 68 years." "…for the different catchments varied between 50 and 68 years." Ln 118: "…but the selection of events…" "…but the selection of WD events…"?
Ln 119: "…occurred mostly in the winter months." "…coincide mostly with the winter months." Lns 122f: "Therefore, regional average mass balance time series were calculated from all available mass balance observations per year per country (Austria, Switzerland, Norway and Canada) by taking the median." If I understood your approach right, then, as a glaciologist, I have some concerns here: 1) Taking just an average or median value from measured mass balance data of individual glaciers is not really correct, you need to calculate area-weighted average mass balance values (as for instance large valley glaciers have their ablation area at lower elevation (therefore more melt, therefore more negative mass balance) compared to smaller mountain glaciers having their terminus at higher elevation (therefore less melt, therefore less negative mass balance)); i.e. you have to multiply all mass balance data you take into account by glacier area, then calculate the sum of these values for all glaciers, and finally divide that sum by the summed up glacier areas 2) Even though for instance Switzerland and Austria are small countries, measured in-situ mass balance sometimes comes from glaciers with significantly different regional climatic conditions (

past and future glacier mass balance data is available (per glacier!) from the Global Glacier Evolution Model (GloGEM)). Whatever data you use or method you apply to extrapolate (measured) mass balance data to glaciers of your analyzed catchments, do not forget that you have to calculate area-weighted values again if you have to work with "catchment-wide" mass balance data for your analyses. If you want you can also contact me personally (mauro.fischer@giub.unibe.ch) and I would be happy to discuss that with you and try to help you there.
Ln 128: "…(GI4, 2015)…" I think you need to add the correct reference for the Austrian glacier inventory used here.

Ln 130: "…Randolph glacier inventory (RGI) for Canada…" maybe add the version of the RGI you worked with (as there were a lot of updates from the first to the latest RGI version)…
Ln 128: "…ranging in size from…" rather put the "in size" at the end of the sentence (after the numbers)
Lns 202ff: "It was assumed that in these high elevation catchments there is an immediate response to the WD events (or the other dry events) and streamflow data of the dates exactly corresponding to the events were selected." can you add some information on why (from a process understanding point of view) it is ok to work with that assumption? I argue that you're right with this assumption but giving some rationale here would be good I think.
Ln 216: "C of the events…" "Resulting C values of the events…"?
Ln 219: "…general trend of streamflow during the event." "…general trend of streamflow behavior during the event."? Lns 227f: "…the importance of different event drivers." as for example?

Ln 250: I would delete both commas there…
Ln 257: "…the importance of glacier cover…" "…the importance of relative glacier cover…" Ln 260: add a comma before "too." Ln 262: "…for the two lower gc catchments…" "…for the two catchments with lower gc…" Ln 263: I would delete the "pushed" Ln 265: "…compared to the regime…" "compared to the long-term daily streamflow regime…" Ln 270: "…part of the variance of C…" "…part of the variance in C…" Ln 275: "…below the normal regime…" "…below the long-term daily streamflow regime…" Ln 279: "The spread in responses…" "The variability in streamflow responses…" Ln 287: "The sreamflow responses of WWD and WD events…" "The streamflow responses to WWD and WD events…" Ln 291: "…were below the normal regime…" "…were below the normal streamflow regime…"?
Lns 295f: "Different rainfall amounts…"; "…high rain amounts in summer…"; "…low rain amounts in summer…" I guess you refer to the climatological statistics here (how much rain falls on average in one region in summer, cf. figure 2)… I would write that somehow in this sentence in order to be clear… Ln 297: I think you can delete the comma there… Ln 298: "Having less rainfall…" "Less rainfall…"?
Lns 300f: "Also, the relative glacier covers of the Canadian and Alps catchments are more complementary than comparable in this sample of catchments." I believe to see your point here, but doesn't make this statement figure 7 and your rationale/descriptions here about the influence of the amount of average summer rainfall on the levels of compensation a bit obsolete?...
Ln 304: "…show these sometimes rather large ranges. To explain these ranges,…" "…show this sometimes large variability in catchment-wide C. To explain this variability,…" Ln 306: "…variance of C…" "…variance in C…" Lns 323f: "…when MBsum is larger (more negative)." "…when MBsum is more negative." as I wrote above, try to avoid speaking of larger/higher and smaller/lower mass balance, this is always confusing (use "more negative" and "more positive")… Lns 328f: "Most significant trends were found in June (Canada) and September (European Alps), which were all negative. Norwegian catchments showed mostly positive trends, except in September Ln 534: "…with levels…" "…with compensation levels…" Ln 535: "…streamflow anomalies 30 days before the event…" "…streamflow anomalies during 30 days before the event…" Ln 538: "Understanding these different mechanisms…" "Understanding the different mechanisms…" Ln 541: Delete "these" Ln 543: I would add "in high mountain catchments" before "over the summer"    "…level of compensation (C)."; "Higher elevated glaciers:…" "Glaciers at higher elevation:…"; "Percentile of mean temperature in spring (MAM)" which percentile?; "Percentile of sum of precipitation in winter (DJF)" which percentile?; "The streamflow percentile of the 30 days before the WD event"

Tables
which percentile?; "Higher MBsum…" I would delete that because "higher" means more positive and here you mean more negative (it is always easier to talk about "more negative" or "more positive" glacier mass balances in order to avoid misunderstandings when using "higher" or "lower" mass balances!), moreover you use the abbreviations MBsum and MBwin but don't explain them (I can easily guess what it means but someone else not too familiar with glaciers might not at first glance); please see also my comments above about how you calculated "regional mass balances", I think if you calculate catchment-wide mass balance or mass balance anomalies in a more appropriate/more correct way you can use these numbers much better for a more realistic interpretation of "C" with the help of mass balance data… "Years with most events"; in my opinion it would be good to add the actual observation periods  in parentheses after "Alps", "Norway" and "Canada"! Table 4: Table caption: "The numbers in between brackets…" "The numbers in parentheses…"; in my opinion it would be good to add the actual observation periods  in parentheses after "Alps", "Norway" and "Canada"!