the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Sep 2023
Climatology of Snow Depth and Water Equivalent measurements in the Italian Alps (1967–2020)
Roberto Ranzi
Giorgio Galeati
Abstract. A climatology of SWE based on data collected at 299 gauging sites was performed for the Italian Alps over the 1967–2020 period, when the Italian National Electric board conducted routinely and with homogeneous methods snow depth and density measurements. Six hydrological sub-regions were investigated spanning from the eastern Alps to the western Alps at altitudes ranging from 1000 m to 3000 m asl. Measures were conducted at fixed dates at the beginning of each month from 1 February to 1 June and on 15 April. To our knowledge this is the most comprehensive and homogeneous dataset of measured snow depth and density for the Italian Alps. Significant decreasing trends over the years at fixed dates and elevation classes were identified for both snow depth (−0.12 m decade−1 on average) and snow water equivalent (−37 mm decade−1 on average) in most of the six investigated areas. The analysis of bulk snow density data showed a temporal evolution along the snow accumulation and melt season, but no altitudinal trends were found. A Moving Average and Running Trend Analysis (MARTA triangles), combined with a Pettitt’s test change-point detection, highlighted a decreasing change of snow climatology occurring around the end of the 1980s. Correlation with climatic indexes indicate significant negative values of Pearson correlation coefficient with winter North Atlantic Oscillation (NAO) and positive values with winter West Mediterrean oscillation (WeMO) for some areas and elevation classes. Results of this climatology are synthesized in a temporal polynomial model useful for climatological studies and water resources management in mountain areas.
- Preprint
(2894 KB) - Metadata XML
-
Supplement
(2289 KB) - BibTeX
- EndNote
Roberto Ranzi et al.
Status: final response (author comments only)
-
RC1: 'Comment on hess-2023-223', Xavier Fettweis, 18 Oct 2023
This manuscript presents a new unique data set of snow height/density measurements in Italian Alps and discusses the recent changes with robust statistical methodologies. This paper fits well with HESS and deserves to be published. However, I have some remarks before publication:
- The average coordinates (lat, lon, elevation) as well as the elevation range of the 6 studied basins should be listed in a table. Adding the mean winter temperature as well as precipitation will be useful also.
- To compute SWE, two separate periods are used to fit the parameters. How does using one of them over the whole time series or fitting the parameters over the whole time series impact on the presented results? Some figures using other estimates of SWE could be added in supplementary material. Or uncertainties should be added in estimates of SWE.
- The discussion and comparison with NAO and WeMO is not useful for me as correlations are listed only and discussed. But nothing is said on how changes in NAO or WeMO could have impacted the presented time series.
- Section 3.5 should be in Section 2.6 where the model of SWE is presented.
- A table listing the correlation of each basin with the other ones by elevation classes will be very useful. Are the temporal changes correlated between the different basins ? What is the reasons to have chosen 6 basins and not 5 or 7 ? or only 3 basins? For example, Basins 2, 3, 4, 5 could be aggregated together. Moreover, the red lines in Fig 1 are confusing as they concern the hydrological basins while numbers concern the area of measurements (which are mountain ranges in fact and not basins). I suggest to add circles surrounding the measurements sites chosen for the 6 basins in addition to the red lines.
Citation: https://doi.org/10.5194/hess-2023-223-RC1 -
RC2: 'Comment on hess-2023-223', Anonymous Referee #2, 30 Oct 2023
The Authors present an analysis of variability of snow depth and snow water equivalent from 1967 to 2020 in the Italian alps based on a large dataset of manual measurements from 299 stations collected by the Italian national electric board. The data set present the peculiarity of been conducted at fixed date during the snow season (1 February, 1 March, 1 April, 15 April 1 May and 1 June).
The variability of snow depth and estimated snow water equivalent is presented, for six macro basins aggregation, in term of seasonality, elevation and non-stationarity (trends and changing points). The authors conclude on a significant decrease in both snow depth and snow water equivalent starting from the late 1980, with higher decrease at lower altitude, mainly attributed to increase 2m air temperature has seen by the HISTALP dataset.
Finally, the authors propose a simple mean SWE climatology regression model for one selected macro-basin (Oglio-Chiese-Sarca) as function of seasonality (day of the year) and elevation, for the sub period 1994-2020.
The paper is of interest for the scientific community working on the snow climatology, well written and in line with the HESS aims and scope, using a unique dataset collected by the Italian national electric board.
Nevertheless, the overall structure is confusing in both the objectives and the use of regressive approach vs direct observations, and in my opinion the conclusion on the role of climate driver and teleconnection should be better presented/supported. In general, I think the Authors should better highlight and clarify the main contributions of the proposed contribution, which in my opinion are:
#1-confirmation of a negative trend of observed snow depth in the Italian alps using a large independent dataset of manual observation with identified changing point.
#2- confirmation of a major variability of snowpack bulk density along season rather than elevation using a large independent dataset.
#3- confirmation of a negative trend in snow water equivalent using observed snow depth and observed /estimated(?) bulk density (based on a recalibrated regression on seasonality?), with identified changing point.
#4- Identification of the major role played by temperature over precipitation by comparison of changing point obtained with the HISTALP product, and discussing the major decrease of both snow depth and SWE at lower elevation.
Point #1 to #3 are results, while point #4 may be consider as discussion.
In the following, some general comments hoping it may help the authors to increase the overall readability:
-They are too many figures (11) in the current manuscript. The Authors should consider to focus on the main results, and move some figures to the supplementary to improve readability. Moreover, some of these figures are presented for a sub selection of the macro basins (figure 4, 5, 7, 13 and 14). In such case, the other macro basins should be reported in supplementary. I would suggest to limit the figures in the main text to figure 1 (maybe including figure 2), figure 6 (supporting point #1 with table 1 and 2), figure 7 (supporting point #2, improved using all macro-basins, and maybe including figure 8), figure 9 (supporting point #3, with table 4 and 5), figure 11 if better adapted (see below) to support the discussion of section 3.4 and point #4 (figure 12 can be a table or moved to supplementary) and a merge of figure 13 and 14 to present the mean snow climatology model.
- The mean snow climatology model could be slightly developed, improving its added value of the paper, by comparing the 1994-2020 period with the previous period. Moreover, here the choice of the sub period is based on the Mann – Whitney U test run over the dependency of snow depth versus elevation (line 288-89), while a huge effort has been done by the Authors to identify the changing point in the non-stationarity of snow depth and SWE. The authors may consider those changing points (e.g. 1988) to separate the two mean snow climatology (Note that in the current figure 13 and 14, the limitation to a single macro basin should be state in the caption)
- In general the MARTA triangles figures are not discussed in the text, excepted for the location of the changing point (also reported in table): Figure 5 is discussed line 255-256; figure 10 is discussed line 355-258 to highlight the lack of linear trend in precipitation; figure 11 is basically not discussed (while the interesting discussion on early/late season temperature trend in not supported by a figure, cf line 359).
- The conclusions on point #4 (role of temperature and eventually role of teleconnection) need to be better supported. For exemple, the Authors may further discuss the elevation dependence of changing point (supporting the role of temperature).
- Section 2.2 line 147-154 create a lot of confusion (to me). Here the Authors state here that bulk density is reconstructed when missing based on mean values of corresponding macro basins and elevation range, while in section 3.2 a regressive model is proposed (and recalibrated) and in 3.3 the SWE non stationarity is discussed based on a “SWE estimates” (line 325) which is still unclear for me (my guess is observed snow depth and estimated bulk density as presented in section 3.2?). This should be clarified.
- The Authors may reconsider their abstract and conclusion on the light of the actual contribution (main results).
In the following some minor comments:
- line 15 please be more precise than “most of the six investigated areas”
- line 25-31: The highlighted role of glacier in the introduction is somehow misleading as nothing is said about the interaction between snow and glacier in the paper.
- section 2.1: It is hard for the reader to have an opinion on the choice of the measurement spatial aggregation.
- in section 2.3 The average temperature over each macro-basin did non consider elevation dependence. This should be discussed.
- in section 2.4. The Authors cite Ranzi et al.2021 in evaluating the role of teleconnection, but it has also been done by others.
-in section 2.5. The Authors may better choose between parametric and non-parametric approach for linear trend estimation and associated significance. The t-test is fine for the least-square regression, while the Mann Kendall should better be associated to the Sen’s slope.
-in section 2.5. The role of teleconnection is investigated only in term of correlation over the whole period (this non considering eventual non stationarity in the response of snow climate to teleconnection indexes). The Authors may consider to use other statistical instruments (e.g. wavelet coherence or others), or to test the correlation on different base line (e.g. using identified changing point).
-section 2.6 (as well as section 3.5) may be renamed to avoid confusion with SWE variability analysis (3.3). "Mean snow climatology model" could be a proposition. Maybe the section could be rephrased to better introduce that such model is used for “mean” consideration.
-section 3.1 line 300-303: The Authors state that the (snow depth) model “show less reliable” results for 3/6 macro basins. Based on table 3, I would say that the model is not working at all for Toce and Adda (r2 ranging from 0 to 0.26) while results of Adige seems rather similar to the other (with r2 ranging from 0.63 to 0.85).
-section 3.4 The authors should consider the results of Colombo et al 2023 (DOI 10.1088/1748-9326/acdb88) on the role of temperature / precipitation.
-Supplementary: The current second table in the supplementary should be better shared as a data source, as it is hardly readable in the present form.
Citation: https://doi.org/10.5194/hess-2023-223-RC2 -
CC1: 'Review on hess-2023-223', Marijn van Dijk, 03 Nov 2023
DISCLAIMER: This review was prepared as part of graduate program course work at Wageningen University, and has been produced under supervision of Ryan Teuling. The review has been posted because of its good quality, and likely usefulness to the authors and editor. This review was not solicited by the journal.
The paper considers a large dataset of Snow Water Equivalent (SWE) for the titled time period. The stated objective of the paper is to understand snow climatology and characteristics better, verify previous studies with less dense datasets and create a simple polynomial model that can predict average snow depth and SWE for different elevation levels during the year. This is done by using an Italian dataset for which snow samples were taken the first of every month from February until June and additionally on the 15th of April. The research considers 6 major basins aggregated for their own elevation and climatological characteristics. It becomes apparent that elevation and resulting average temperatures influence the snow cover most, with regard to the different basins, that there is a negative trend over the years and that there is a change point around 1988.
The study uses the data for a Moving Average and Running Trend Analysis, which presents the data in a clear and fitting way. Additionally, they plot the data to see temporal changes from two time periods ('67-'93 and '94-'20). For the basins with the most datapoints, Piave-Brenta and Oglio-Chiese-Sarca, there are significant changes in average SWE, on all altitude levels, especially in the early melt season (April 15th & May 1st). From this, and plotting the data with linear regression, they infer that snow depth decreases with -0.12 m per decade and SWE decreases with -37mm per decade for the Oglio-Chiese-Sarca basin. Their results are summarized in the creation of the polynomial model, which can be used in a climatological way to simulate snow depth and SWE for elevation level and the day of the year. The declining trend confirms previous studies and is similar to other mountain ranges. Next to that, other studies often had no or limited access to bulk density measurements, which are needed to calculate SWE. The availability of these data makes this study novel as well. To my judgement, the study can be accepted with major improvements in the framing and delivery of the text. The statistical analysis is sound, but the framing of the text is too generic, which makes the objectives unclear to the reader. Also, some disclaimers should be added regarding data aggregation and the discussion of the results. The main points of my criticism will be elaborated in the next section of this review.
Issue 1: Data Aggregation
My first issue with the research is the data aggregation done for the different basins. It is logical that for this research, a spatial average is needed for the different basins, so I agree with the idea of aggregation. If one consults figure 1 however, it becomes clear that for some basins, the gauging stations were really clustered in a certain area of the basin. Especially basin 3 (Adda) and 5 (Adige) on the map in figure 1 have this issue, as Adda has all the measurements done central north and Adige mainly on 1 mountain ridge. These measurements may cover only a fraction of the entire macro basin that is considered, whereas for basin 4 and 6 the spread of the gauging stations is much better. This makes the reader to question the representativeness of the data for a whole macro basin. Namely, research on precipitation uncertainty in Austria by O. & Foelsche (2019) found that dense and regular distribution of gauges is necessary to obtain accurate rainfall estimates, which will also be comparable to snow estimates. Similarly, Volkmann et al (2010) found that precision and accuracy strongly with network density for catchment hydrology purposes, which could be compared to this study’s basin aggregation. To add to this, many of the concluding remarks are based on the well-measured Oglio-Chiese-Sarca basin, which is the only basin that is significantly affected by a Glacier. This glacier might have an influence on the snow dynamics in the surrounding region. This issue concerns section 2.1 of the methods and section 3.1 & 3.3 of the results & discussion.
To solve this issue, the text has to specifically state that the measurement spread and density differs per basin and that the researchers choose Oglio-Chiese-Sarca (and maybe Piave-Brenta) for the most representative results, as these have good measurement density and spread. Then add that the results of the other basins are more uncertain due to the lower spread of the data. This is also seen in the limited correlation for some of the mentioned basins in table 3, so it would be only fair to address this in the text. Moreover, add that Oglio-Chiese-Sarca is affected by glacier dynamics, so that it is uncertain if the results are representative for other basins or the larger Italian Alps. Also, the abstract states that -0.12 per decade is found on average, but this is again for the specific Oglio-Chiese-Sarca basin. That must be mentioned in the abstract, because otherwise it seems like the whole study finds this decrease, unless that is the case, but that is not explicitly mentioned in the results. In general, be more transparent on the data aggregation and admit that there is data shortage in some basin areas or revise the basins to be a smaller area of the landscape, so that the data is more representative. With that addition the data aggregation procedure is clearer for the reader and the researchers are more transparent.
Issue 2: unclear aims and objectives
Secondly, the introduction is not clear on the objectives and the aim of this research. The most concrete statement made by the researchers is that the dataset is important to understand snow variability and climate change impacts. This, however, is a rather vague and broad objective and it is unclear to the reader what the specific knowledge gap is and how this study will try to find novel knowledge or what the different hypotheses and questions will be evaluated. This undermines the rest of the paper, as a proper introduction would be greatly beneficial to introduce the objectives, aims and point out the novelty and relevance of the later results that are actually well-presented in the different figures and results section. This issue concerns paragraph 1 and 2 on page 3, the end of the introduction section, which should be expanded to be more concrete and research specific.
To improve on this the research objectives and aims should be clearly stated in the introduction. These are partially to verify earlier research that was done with modelling or with less available data, such as from Colombo et al. (2022) and Steirou et al. (2017). That objective would be to use the dataset to show and verify temporal trends in snow depth and SWE decline. Another objective would be to make a general polynomial model for climatological estimation of SWE and snow depth for different days of the year and elevation levels. The third objective could then be to state that NAO and WeMO indices will be correlated with the snow data, as other studies found a relationship and that should be verified by checking if this dataset shows a similar correlation. The aim of the research is next, as it should state what will be the contribution of these conclusions. The conclusions could for example help to better understand climate change effects on snow cover and are essential for future Ski tourism policy and for hydraulic power companies that are affected by changed snowmelt. With additions such as these, the paper has more societal significance and for the reader it will be clear what different subjects will be addressed in the paper and why.
Issue 3: vague section on climatological oscillations
Thirdly, the part on the correlation between snow depth and SWE and the two meteorological oscillation indices could be better. The overview of the Pearson’s correlation coefficients in figure 12 should visualize every coefficient, instead of just the significant ones. I would suggest the researchers highlight the significant values, but show the other values as well, for transparency. Mainly for the WeMO correlation, it is now unclear if there might have been negative values for the other basins and elevation classes. If the values are transparently shown, this will benefit the transparency and conclusions on a positive correlation with WeMO and a negative one with NAO. This issue concerns figure 12, section 2.4 of the methods, the last paragraph of section 3.4 and possibly the conclusion.
Next to that, the description in the manuscript on these correlations is somewhat blunt. The correlation is described and links are made with other literature that prove these correlations, but there is no description of the effect or substantiality that this correlation has. There is no impact of this result described, only that the correlation is present. I would suggest you add a concise description of what this experimental correlation would do for i.e., SWE prediction, SWE temporal variability or analysis of older records. The conclusion adds that further research is needed into this, but some general hypothesis on the effects of these climatological indices could definitely be mentioned, which could be a short adaptation of the study by Osborn (2011) and an additional source on precipitation. Otherwise, this part of the research feels like a detached addition to the whole of the paper and which could be remedied by integrating it into a more comprehensive section of the aims and objectives, which was already mentioned in issue 2.
Minor arguments
- The title could reflect more in depth what the research is actually related to. Climatology is now an overly broad term. Other ideas could be long-term snow cover trends or snow variability in a warming climate or anything else that is more specific for the research.
- Write in the abstract, line 13-15, that these results were found for basin Oglio-Chiese-Sarca, not for most of the investigated areas. Or write that a similar decline was found for most areas, now it seems like all areas have this average, which is not the case.
- In section 3.1 the text first handles different results on decline per decade and later discusses decline between the first and second half of the measurement period. This could maybe be emphasized a bit more, as it can be confusing that the decline is suddenly much bigger, because it concerns several decades.
- In section 3.2, line 311-313, I would add a reference about snow bulk density increase over the season, as it might not be common knowledge for most readers.
- In section 3.5, line 399-401, the sentence suddenly stops, as if there was more that the writer wanted to say, but he forgot to write it down. Make sure this is fixed.
- Figure 8, and section 3.2 are used to conclude that bulk density does not vary significantly with elevation, which confirms earlier studies. This is, however, only shown for the basin Piave-Brenta. It might be considered to state that the other basins showed comparable results, but Piave-Brente was visualized because of the data abundance.
Minor issues
- Line 20, p1, Mediterranean is spelled wrong (like Mediterrean)
- Line 175, p6, same here, Mediterranean is spelled wrong again.
- Line 183-184, p6, the reference of Rosso and Kottegoda is not included in the bibliography.
- Line 205, p7, maybe add a reference for the Pearson’s coefficient, as you did add a reference for the other statistical tests (t-test, MK-test, Pettitt)
- Figure 3, p22, consider making the limits of the Y-axis of the plots narrower so that the individual lines can be viewed better. Now it is hard for the reader to see all the differences.
- Figure 4, p23, maybe only add the slope of the trendlines, as the intercept value has no physical meaning here.
- Figure 13, p32, consider adding lines for the confidence interval and prediction interval of the polynomial fit, to show the (un)certainty of the fit.
- Figure 14, p33, you could consider using colours to represent these plots, which would add to the visibility and clarity of the figure.
- Table 1 to 5, p34-p38, the captions of the tables should be above the tables, not below the tables.
Citation: https://doi.org/10.5194/hess-2023-223-CC1 -
CC2: 'Comment on hess-2023-223', Rick Haanschoten, 07 Nov 2023
DISCLAIMER: This review was prepared as part of graduate program course work at Wageningen University, and has been produced under supervision of Ryan Teuling. The review has been posted because of its good quality, and likely usefulness to the authors and editor. This review was not solicited by the journal.
Review of Climatology of Snow Depth and Water Equivalent measurements in the Italian Alps (1967 - 2020)
The paper investigates with a large data set the snow dept and SWE between 1967 and 2020 in six subregions and for four elevation classes in the Italian alps. It uses multiple statistical techniques to investigate the data and its changes over time. The data is then also correlated with the NOA and WeMO indexes. The paper then also proceeds to propose a SWE model that depends on elevation and DOY.
The language is clear and of good quality. The work seems to be done in a good and scientific way, with no major flaws or errors. The paper fits in my opinion very well within the scope of HESS. The extensiveness of the data set used in the paper is furthermore unique, but the work done with the data is in my opinion not. The paper also fails to present a clear contribution and lacks novelty. I do think however that after revisions the work could be published.Major arguments
The first major concern is the lack of novelty, or argumentation for why it is novel. The paper cites a lot of papers that have done similar research for different time periods and/or geographic locations. This is good, since it shows that large parts of the methods, since they are similar, are widely accepted and commonly used. However, Colombo et al. (2022) is also focused on the Italian alps and has a time frame from 1960 to 2020, which is the same except it includes 7 more years. The only research not also done in Colombo et al. (2022) is the proposed SWE model. This model is improved from Guyennon et al. (2019). The paper does however not argue or state why their coefficients are better. This results in a paper that seems to combine and repeat the work of two recent papers.
It can however probably be solved by stating better what the differences are between this paper and the two mentioned before. For Guyennon et al. (2019) it is enough to show why your parameters are better than the ones they presented. I assume it is because you have more data, but it would be nice to get some proof or explanation. Especially, since the south-west of the Italian alps does not seem to be included in the research in the paper, while Guyennon et al. (2019) calibrated with this area included. For Colombo et al. it is probably most important that they only use 19 stations that are spatially distributed, while this paper has 299, which could result in much more significant results. The statistical tests are a difference that also could help stress the novelty of the paper. For the NOA and WeMO indexes, it could be nice to stress that the WeMO has never been correlated with snow dept or SWE.The second major argument is the failure to present a clear contribution. In the introduction a few statements are made that could be seen as a contribution, but all of these have flaws. It starts off by stating that it is important to know the snow extent for the 10-billion-dollar ski tourism industry. The paper however only looks at snow dept and SWE and therefore does not contribute to more knowledge about snow extent. The paper then continues to state it is important to know SWE for water managers, agriculture and hydropower plants and this is true and would seem like a good contribution. These parties do however need almost real time value to make their day-to-day decisions on either storing or discharging water. A long-term trend can of course give inside in typical values and a model would be a great contribution. The paper only already proves that the model needs significantly different parameters for different time periods, thereby creating the question if the model would be valid and useful for the upcoming years. The model also takes the average over multiple years and does therefore not take single year variability into account. This is however very important for water managers, agriculture and hydropower plants (HPP), since they need accurate, current values and not averages. It could result in massive over or underestimations of the amount of water stored in the mountains if these parties would use this model and see it as true values. This could then have severe influences resulting in droughts and water shortages or floods.It would be good to make a good and clear statement in the introduction on what the paper will contribute. I think the paper could still be very useful for water management, agriculture and HPP. The WeMo and NOA show some correlation and could therefore be used to estimate the amount of snow, especially when the model would be able to give some uncertainty or spread of possible values. Multiple scenarios for wetter and dryer conditions or warmer and colder conditions could also help. In the conclusion then revere back to the statement in the introduction and argue why the paper succeeded in, partially, solving the issue raised in the introduction.
Proksch et al. (2016) raise the issue that the method used to measure the density was not accurate for ice layers. Ice layers also have a density that is above the 0.75 g cm-3 (Watts et al., 2016) threshold that the paper used as errors in the data. The paper however completely fails to mention these layers and it is therefore not clear what was done with them.
The SWE could be underestimated, if they were ignored and deleted, since their density was too high. The paper either needs to explain how these ice layers were considered or add a brief explanation and discussion on this topic. It would be preferable if an estimation could be given of the potential underestimation if the ice layers were neglected.Minor arguments
It would be nice to show an overview of how the 299 measurements were distributed over the different basins and elevation classes. It might also be good to show the elevation range of a basin. Both would be nice, since it could help understanding why there is no data in certain basins at certain elevations and/or why data is not significant.For the correlation of NOA, WeMO and snow dept it is explained why winter NOA is used. It is however not explained why this is also done for WeMO. It is also not discussed why April snow dept values are used and why not for example march, may and/or June. It would be good to shortly argue this.
It would be nice to shortly explain why certain order polynomials are used and why they differ for different parameters. This is explained in Guyennon et al. (2019), which the paper references, but a brief explanation in the paper itself would be good.
P1, line 28: “Modifications of the Greater Alpine Region climate have been confirmed by the analysis of the HISTALP dataset, with significant trends in temperature, twice as the global average, precipitation and relative humidity (Auer et al., 2007; Brunetti et al., 2009).” Is not completely true since Brunetti et al. (2009) concludes: “If only the low-level areas are considered, relative humidity also has a clear long-term trend, with a decrease of about 5% per century. Such a decrease is, however, not shown by the record representing the high-level locations.” The paper is largely about high elevation locations, so this statement is therefore not completely true and misleading. Please adjust accordingly.
P2, line 33: “Snow cover regulates the surface energy balance, affecting circulation patterns and atmospheric flow regimes (Ge and Gong, 2009).” Ge and Gong (2009) do state in the introduction: “Anomalous snow cover can influence the surface energy balance and temperature over a broad land surface region, and in turn affect atmospheric flow regimes, circulation patterns, and hemispheric climate.” They however do not provide any data or sources to prove this statement, other than stating that literature has focused on this. The paper itself is furthermore about snow dept and its influences instead of snow cover and/or snow extent. It is therefore probably better to cite other literature, such as for example: Groisman et al. (1994), Gong et al. (2004) and/or Fletcher et al. (2009).
P2, line 58: The sentence should be: “… Lejeune et al. (2019) used a snow dataset of 57 years ...”.P7, line 222: It is not clear which slope is meant with m (slope) in this sentence. Up until that point slope has been used in reference to the soil/ground below the snow. The m here however depends on DOY, which leads me to think it is the slope of the snow in this case. It would be good to make this clearer, by adding a sentence on what slope is exactly meant in this context.
P13, line 401: The sentence ends with “and”. Either a sentence is missing here, or the sentence should end with: “… day of the year.”
P20, Fig 1: There is a red line in the figure, which is barely visible especially for colorblind people. It would be good to explain what the lines represent, I assume the division of the basins, and make them better visible, by for example changing the color.
P26, Fig 7: I assume “mag” and “giu” in the legend should be “May” and “Jun”.
P31, Fig 12: In general elevation is presented on the vertical axis instead of the horizontal. I would therefore change the figure to show elevation on the vertical and the basins on the horizontal axis.
P34, Tables 1,2,4 and 5 show a lot of “ND” and “—” values which makes it hard to read the data that is actually in there. It might be good to present this data in another way that focuses more on the values that are obtained. It would also be good to explain the difference between “ND” and “—" a bit better.
References
•Brunetti, M., Lentini, G., Maugeri, M., Nanni, T., Auer, I., Böhm, R. and Schöner, W.: Climate variability and change in the Greater Alpine Region over the last two centuries based on multi-variable analysis, International Journal of Climatology, 29: 2197-2225, 2009. https://doi.org/10.1002/joc.1857
•Colombo, N., Valt, M., Romano, E., Salerno, F., Godone, D., Cianfarra, P., Freppaz, M., Maugeri, M., and Guyennon, N.: Long-term trend of snow water equivalent in the Italian Alps, Journal of Hydrology, 614, 128532, 2022.
•Fletcher, C. G., Kushner, P. J., Hall, A., and Qu, X.: Circulation responses to snow albedo feedback in climate change, Geophys. Res. Lett., 36, L09702, 2009. doi:10.1029/2009GL038011
•Ge, Y., and Gong, G.: North American snow depth and climate teleconnection patterns, Journal of Climate, 22(2), 217-233, 2009. https://doi.org/10.1175/2008jcli2124.1
•Gong, G., Entekhabi, D., Cohen, J., and Robinson, D.: Sensitivity of atmospheric response to modeled snow anomaly characteristics, J. Geophys. Res., 109, D06107, 2004. doi:10.1029/2003JD004160
•Groisman, P. Y., Karl, T. R., Knight, R. W., & Stenchikov, G. L.: Changes of Snow Cover, Temperature, and Radiative Heat Balance over the Northern Hemisphere. Journal of Climate, 7(11), 1633-1656, 1994. https://doi.org/10.1175/1520-0442(1994)007<1633:COSCTA>2.0.CO;2
•Guyennon, N., Valt, M., Salerno, F., Petrangeli, A. B., and Romano, E.: Estimating the snow water equivalent from snow depth measurements in the Italian Alps, Cold Regions Science and Technology, 167, 102859, 2019.
•Proksch, M., Rutter, N., Fierz, C., and Schneebeli, M.: Intercomparison of snow density measurements: bias, precision, and vertical resolution, The Cryosphere, 10, 371–384, 2016. https://doi.org/10.5194/tc-10-371-2016
•Watts, T., Rutter, N., Toose, P., Derksen, C., Sandells, M., and Woodward, J.: Brief communication: Improved measurement of ice layer density in seasonal snowpacks, The Cryosphere, 10, 2069–2074, 2016. https://doi.org/10.5194/tc-10-2069-2016Citation: https://doi.org/10.5194/hess-2023-223-CC2 -
CC3: 'Comment on hess-2023-223', Thijs Merkx, 07 Nov 2023
DISCLAIMER: This review was prepared as part of graduate program course work at Wageningen University, and has been produced under supervision of Ryan Teuling. The review has been posted because of its good quality, and likely usefulness to the authors and editor. This review was not solicited by the journal.
This study looks at how the Italian Alps' snow depth and snow water equivalent have changed between 1967 and 2020. The study shows long-term trends and variability in snow cover using in situ measurements. In six macro-basins of the Italian Alps, the study examined snow depth and snow water equivalent (SWE). The findings show a considerable downward trend in SWE and snow depth. The average reduction of snow depth was 33%; the variations were greater at lower elevations (62% at 1000–1500 m) and smaller at higher elevations (30% at 1500–2000 m). SWE dropped on average by 32%, falling more precipitously at lower elevations (52% at 1000–1500 m) and less precipitously (28–29% at higher altitudes). According to the analysis, the data collected strongly suggest that the melting is impacted by global warming, which in turn has an impact on flood events, hydrological regimes, and water supply. Additionally, a change-point around 1988 was found by the study, which indicated decreasing snow depth and SWE in the decades that followed. The study produced a basic elevation- and time-based SWE model, which is a helpful tool for calculating the evolution of SWE in specific Italian Alps areas.
When looking at the overall relevance of the paper, the social relevance really stands out. As described in the introduction, knowledge about the snow depth and SWE is important for winter sports season prediction and can thus play an important role in the economic stabilization of the area during the increase in temperatures and thus snow melt.
The estimation of SWE based on snow depth is not a new concept M. Sturm (2010); however, it has not been done in this area. The unique parameters of the study area can give more accurate results for this unique set of snow depths. In addition, the reliability of the study is high because of the strong statistical analyses and background information supporting the claims and results found.
All in all, I think that the research is really valuable and well conducted. This paper is clearly worth being published after a few revisions on the following issues:
Major issue’s:
To start with issue (1): In the research, a model is made to determine the SWE (snow water equivalent) based on altitude and DOY (day of year). Although this model works quite well for some of the basins, for example, Oglio-Chiese-Sarca, this could not be the case for all other basins. As the paper already states, the R-squared value for this basin is the largest for Oglio-Chiese-Sarca. Since the article claims this model works well to estimate the SWE, not only the best case should be used but also the basins that will probably give a worse estimate. Also, the paper doesn’t touch on the potential estimation error of the SWE. The size of this error is an important factor in determining how well the model actually works and if it’s worth using for future research and measurements. My suggestion would be to look more into the model errors and statistical significance of the model to really be able to tell how well this model works in estimating the SWE. It would help calculate the mean squared error and the confidence interval.
Major issue 2: The aim of the study is unclear. In the introduction, multiple issues are stated as well as reasons for the usefulness of the study. This, however, is not translated to the aim of the study nor to research questions. Moreover, the words: question or aim are not even used in the paper. This generates a lack of study direction throughout the entire paper, making the conclusion feel open-ended. My recommendation would be to add a research question in the form: technical problem statement, result of the problem, research question. This would give the paper a clearer outline that can again be followed in the conclusion. I also recommend using a signal word like aim to really make it clear what the aim is and to prevent confusion.
Minor issue’s:
Minor issue 1: Figure 1 of the study area is not really helpful in understanding the study area, since it doesn’t give a clear representation of the study; it is hard to see what basins are aggregated together. While it is clearly described in the text what basins are aggregated, it would be good to also show this in Figure 1. Furthermore, the different basins are visualized by the red lines (I think this is not specified in the caption nor text), however, I can only distinguish 11 basins, which is less than the 15 basins described in the text. Besides these points, the measurement points are quite hard to distinguish due to the contrast at a certain location with the elevation map. I recommend adding a different colored line to outline the aggregated basins, together with an explanation of the red line in the legenda or caption. Adding the basins (red lines) that are still missing, and finally changing the contrast of the elevation map to show the measurement locations more clearly.
Minor issue 2: In the conclusion, the recommendations for future studies are missing. This would be relevant to show the importance of the research when looking at potential innovation in this research area.
Minor issue 3: Line 401 stops mid-sentence, which keeps the paper with an incomplete result of the SWE model.
Minor issue 4: The paper claims that T. Grünewald (2014) found a positive correlation between snow depth/SWE and elevation. This is partly correct; T. Grünewald (2014) found a positive correlation until a certain elevation. This is stated further down in the paper, but it might be good to also mention it with the earlier citation, since now it looks like a partly false statement.
Minor issue 5: In figure 5, the moving average triangle of the O-C-S and Toca have a different color scale. For comparing the different basin graphs, it would be helpful to make them the same.
Minor issue 6: Naming of locations is not constant. In the text, the full name of the locations is used. However, in Figure 12, abbreviations are used, for example: OC for Oglio-Chiese-Sarca (this abbreviation is explained in the caption). However, in tables 1, 2, 4, and 5, O-C-S is used without explanation. For easy understanding, I suggest making all abbreviations the same.
Minor issue 7: In the paper, multiple spelling and grammar errors can be found. For example, Line 54: 54-years period. I suggest reading it a few more times to get rid of these errors or using a grammar checker.
References
Grünewald, T., Bühler, Y., and Lehning, M. (2014). Elevation dependency of mountain snow depth, The Cryosphere, 8(6), 2381-2394.
Sturm, M., Taras, B., Liston, G. E., Derksen, C., Jonas, T., & Lea, J. (2010). Estimating Snow Water Equivalent Using Snow Depth Data and Climate Classes. Journal of Hydrometeorology, 11(6), 1380-1394.
Citation: https://doi.org/10.5194/hess-2023-223-CC3
Roberto Ranzi et al.
Roberto Ranzi et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 404 | 91 | 16 | 511 | 22 | 2 | 4 |
- HTML: 404
- PDF: 91
- XML: 16
- Total: 511
- Supplement: 22
- BibTeX: 2
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1