Reconstructed glacier area and volume changes in the European Alps since the Little Ice Age

Reinthaler, Johannes; Paul, Frank

doi:https://doi.org/10.5194/egusphere-2024-989

Preprints

https://doi.org/10.5194/egusphere-2024-989

Preprints

03 May 2024

| 03 May 2024

Reconstructed glacier area and volume changes in the European Alps since the Little Ice Age

Johannes Reinthaler and Frank Paul

Abstract. Glaciers in the European Alps have experienced drastic area and volume loss since the end of the Little Ice Age (LIA) around the year 1850. How large these losses were is only poorly known as, published estimates of area loss are mostly based on simple up-scaling and alpine-wide reconstructions of LIA glacier surfaces are lacking. For this study, we compiled all digitally available LIA glacier extents for the Alps and added missing outlines for glaciers >0.1 km² by manual digitising. This was based on geomorphologic interpretation of moraines and trimlines on very high-resolution images in combination with historic topographic maps and modern glacier outlines. Glacier area changes are determined for all glaciers with LIA extents at a regional scale. Glacier surface reconstruction with a Geographic Information System (GIS) was applied to calculate (a) glacier volume changes for the entire region from the LIA until today and (b) total LIA glacier volume in combination with a reconstructed glacier bed. The glacier area shrunk by 2405 km² (-57 %) from 4211 km² at the LIA maximum to 1806 km² in 2015 and volume was reduced from about 281 km³ around 1850 to 100 km³ (‑65 %) in 2015, roughly in line with previous estimates. In the mean, glacier surfaces lowered by -43.3 m until 2015 (-0.25 m a^-1), which is three-times less than observed over the 2000 to 2015 period (-0.82 m a^-1). Strongest volume losses occurred around 1600 m and at least 1832 glaciers melted away completely. Many glaciers have now only remnants of their former coverage left, which led to deglaciation of entire catchments. The new datasets should support a wide range of studies related to the effects of climate change in the Alps.

Received: 04 Apr 2024 – Discussion started: 03 May 2024

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 1907 KB)

Supplement (3219 KB)

Download & links

Johannes Reinthaler and Frank Paul

Status: final response (author comments only)

CC1:
'Comment on egusphere-2024-989', Melaine Le Roy, 06 May 2024
Dear,
First, congrats to the authors for this long-awaited and well conducted work!
I have two comments on the manuscript:

1) Regarding Western Alps outlines
The Figure 2 shows Ecrins massif with 'new' outlines in 'blue'.
However, these LIA glacier outlines were already available (please see the figure attached here) in Marie Gardent’s 2014 PhD thesis (where the 'yellow' outlines you use originate I suppose)
So why have you redigitized them and presented them as new here (Fig. 2)?
By the way, the two Gardent’s references listed at the end are not actually quoted anywhere in the main text/figures of the manuscript.
So, why not quoting them in the caption of Figure 2, as it seems yellow outlines come from them?
If not, please provide a reference here for previously available outlines.

2) Regarding the timing of LIA maximum extent
It sounds to me somehow misleading and too simplistic to present differences in the timing of LIA maxima as depending only on the geographical location (e.g Western Alps/Italy in 1820 CE vs Austria in 1850 CE) as it was done here in the Introduction of the manuscript.
Better, the response time of glaciers should be mentioned as one of the (most likely) explanations for the differences in the chronology of LIA glacial maxima.
Indeed, we showed recently in two review papers dealing with the early LIA period (https://journals.sagepub.com/doi/full/10.1177/09596836221088247) and the Holocene (https://www.sciencedirect.com/science/article/abs/pii/B9780323997126000180?via%3Dihu) in the Alps, that (small) glaciers reached LIA maximum extent during any periods of LIA glacier maxima. For instance, many small glaciers, located from the Ecrins massif (to the west) up to the Tirol (to the east), reached absolute LIA maxima during the early 14^th century.
The LIA chronology is therefore biased towards large (and most slowly reacting) glaciers because majority of available dates come from these sites, where most work has been carrying out in the past.
Related comment for Line 42:
‘in contrast to other regions in the world, extent differences (e.g. between 1850, 1820 or 1600) are small in the Alps’
Please provide references to support your point here. This affirmation does not seem straightforward to me at all.
Mainly because:
Even if it’s relatively true at large Alpine glaciers, it is less at small glaciers.

But foremost, dates with temporal resolution high enough to assert this (archives, tree rings) are barely available outside the Alps

Kind regards
Melaine Le Roy
Citation: https://doi.org/10.5194/egusphere-2024-989-CC1
- AC2: 'Reply on CC1', Johannes Reinthaler, 25 Jun 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-989/egusphere-2024-989-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-989-AC2
CC2:
'Comment on egusphere-2024-989', Riccardo Scotti, 10 Jun 2024

Dear Editor, dear authors,

Considering the topic of the manuscript under discussion, I would like to flag to the authors prior studies that have focused on LIA glacier extent within the Italian region of Lombardy.

In lines 89-90 the authors state: “The largest regions without published LIA outlines were the Italian region of Lombardy and parts of the Dauphiné Alps (see 89 Table S3 for a list of regions with previously missing LIA glacier extents).”
In this context, relatively recent LIA glacier extent reconstructions have been conducted and published for sectors of the Lombardy region. Scotti et al. (2014) dealt with the Livigno Alps, Monte Disgrazia Massif and the Orobie Range. Scotti et al. (2017) and Scotti and Brardinoni (2018) focused on Dosdè and Val Viola. Finally, Hagg et al. (2017) have conducted LIA glacier mapping in the Suretta Sud glacier in Valle Spluga.
The mapping approach in the foregoing studies entailed visual interpretation of sequential historical aerial photos and a recent LiDAR-derived hillshade raster, which was complemented by inspection of historical ground-based oblique pictures and historical maps, as well as by fieldwork.

To properly acknowledge prior work on the reconstruction of LIA glacier extent in Lombardy, I’d kindly ask the authors that the foregoing studies be cited in lines 89-90, and/or in lines 32-38 of the original manuscript.
Best regards,

Riccardo Scotti

References:

HAGG, W., SCOTTI, R., VILLA, F., MAYER, E., HEILIG, A., MAYER, C., TAMM, W., HOCK, T. (2017): Evolution of two cirque glaciers in Lombardy and their relation to climatic factors (1962–2016). Geografiska Annaler: Series A, Physical Geography, 99(4), 371-386. https://doi.org/10.1080/04353676.2017.1368834
SCOTTI, R., BRARDINONI, F., CROSTA, G.B. (2014): Post-LIA glacier changes along a latitudinal transect in the Central Italian Alps. The Cryosphere, 8, 2235–2252. https://doi:10.5194/tc-8-2235-2014
SCOTTI, R., BRARDINONI, F., CROSTA, G.B., COLA, G., MAIR, V. (2017): Time constraints for post-LGM landscape response to deglaciation in Val Viola, Central Italian Alps. Quaternary Science Reviews, 177, 10-33. https://doi.org/10.1016/j.quascirev.2017.10.011
SCOTTI, R., BRARDINONI, F. (2018): Evaluating millennial to contemporary time scales of glacier change in Val Viola, Central Italian Alps. Geografiska Annaler: Series A, Physical Geography ,100 (4), pp.319-339. https://doi.org/10.1080/04353676.2018.1491312

Citation: https://doi.org/10.5194/egusphere-2024-989-CC2
- AC1: 'Reply on CC2', Johannes Reinthaler, 25 Jun 2024
  
  Thank you very much for pointing us to these studies which we can certainly include when discussing earlier work. Our statement in L89/90 referred to missing digital outlines (downloadable in shape file format from a given link) rather than publications about outlines. So thank you for providing us the LIA outlines for Val Viola via our separate request. Please send those for the other regions also to the GLIMS glacier database when they are ready.
  
  Citation: https://doi.org/10.5194/egusphere-2024-989-AC1
CC3:
'Comment on egusphere-2024-989', Renato R. Colucci, 18 Jun 2024

I read this work with interest, as it certainly represents a necessary contribution to the better understanding of glacial evolution in perhaps the most studied mountain sector in the world when dealing with mountain glaciers. Nonetheless, my view is that the uncertainties of assessment that a work of this type inevitably entails could have been minimized by wider cooperation. In a mountain chain as deeply studied as the Alps, the individual specificities of the various Alpine sectors are certainly clearer locally.
After this general comment, I will go into the specifics of the area of my major competence. Particularly, on line 163 and in several figures, I may disagree with the subdivision of area 13, as it would be more correct to indicate Dolomites and Julian Alps, instead of Carnic Alps. In fact, in the Carnic Alps there is only one documented glacier in Austria, the Eiskar Glacier.
I might suggest also reading and citing the following as it represents the first detailed inventory of the LIA Julian Alps glaciers with the first details on what in this preprint is discussed at paragraphs 3.1, and 4
Colucci R.R. (2016). Geomorphic influence on small glacier response to post Little Ice Age climate warming: Julian Alps, Europe. Earth Surface Processes and Landforms, 41: 1227-1240

http://onlinelibrary.wiley.com/doi/10.1002/esp.3908/abstract

Citation: https://doi.org/10.5194/egusphere-2024-989-CC3
- AC3: 'Reply on CC3', Johannes Reinthaler, 26 Jun 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-989/egusphere-2024-989-AC3-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-989-AC3
RC1:
'Comment on egusphere-2024-989', Anonymous Referee #1, 01 Jul 2024

The manuscript by Reinthaler and Paul describes a glacier surface and extent reconstruction for the Little Ice Age maximum (assumed to be in 1850) and calculates area, thickness and volume changes from the LIA maximum to 2000 and 2015. The study makes use of numerous existing datasets and newly determines LIA outlines and surface topography for 19% of all glaciers in the Alps (>0.1 km²) for which outlines were missing in previous inventories. I find that the study uses generally robust methods to generate novel and significant findings. The manuscript is well-structured and well-written, the presentation is clear, and the results, analysis and discussion provide useful perspectives on acceleration of changes as well as uncertainties. I suggest the manuscript can be published after minor revisions. My comments are detailed below. I have some doubts about the method that is used to calculate the total volume error estimate, which I would like to see addressed. Other (minor) comments are mostly suggestions for textual revisions, added discussion, clarification and additional references.
Comments:
Volume uncertainty: The current error estimate of the volume does not seem correctly calculated. Summing squared relative errors and taking the square root is normally applied to calculate the total error of a variable that is the product (or division) of other variables with known errors (see e.g. https://www.statisticshowto.com/statistics-basics/error-propagation/). The total volume is however not the product of glacier outlines, surface reconstruction, and bed topography, but rather the product of the related ice thickness and area. Therefore, a better way to calculate volume errors is to apply error propagation on the product of area and thickness where relative errors of glacier area and thickness are individually defined. I think this formula could be used; eps_V = V*sqrt((eps_H/H)^2 + (eps_A/A)^2) where eps_V, eps_A and eps_H are volume (V), area (A) and thickness (H) errors, respectively.
L19-20: "Many glaciers have now only remnants of their former coverage left, which led to deglaciation of entire catchments.". Please specify how many glaciers this applies to and which catchments are affected, because now it is a rather hollow statement.
L20: "The new datasets should support a wide range of studies related to the effects of climate change in the Alps.". This is also rather vague. It could be good to specify types of studies / study fields that would benefit from the dataset.
L36-37: References and links to GLIMS and RGI should be included here and/or in the Data Availability section.
L40-41: "However, in contrast to other regions in the world, extent differences (e.g. between 1850, 1820 or 1600) are small in the Alps.". Is there a reference for this?
L51-54: "Whereas reconstructions ... European Alps.". This could be merged with / moved to the paragraph ending at L38.
L76-77: "Due to differences in interpretation of glacier extents by different analysts for the two datasets, we will only present glacier changes at a regional scale rather than per glacier.". This helps in case the observer errors are random, but not when they are systematic. Are there any indications for potential systematic errors between the two outline inventories?
L82: "As a starting point for the LIA outline digitizing, we used outlines from 1967-1971 (for France) and the RGI v7.0 from 2003 for the other regions.". What is meant by "as a starting point" here? Is it just to get a rough idea where the glacier is? Or is more information extracted than that?
L95-96: "A few glaciers that melted away before 2003 would lower this number by a few decimals.". What data is this statement based on? Is it possible to give an estimate of the number of glaciers that may have melted entirely since the LIA maximum?
L102-103: "The method calculates a scaling factor by dividing the gradient by the LIA elevation change (from interpolating outline points only)." This is not very clear. Please give a short summary of this method (as described in Reinthaler and Paul, in review) and how it compares to other methods.
L105: Please add a reference for the Topo to Raster toolbox.
L118-119: "Similarly, we have used the year 1850 as the date of maximum LIA extent.". This is a (justifiable) assumption of the timing of LIA maximum extent. It would be good to add this information already in the abstract or introduction.
L128: "This probably resulted in positive elevation changes in several accumulation areas (Figure S6).". Does this also explain differences with Sommer et al. (2020)?
L144-145: See my previous comment about possibly incorrect quantification of the volume error.
L170: "15%". Is it coincidence that this is exactly the same value as the volume loss or is it inherent to the method that was used to reconstuct the surface shape back in time which preserves much of the modern glacier shape (i.e. area - thickness distribution)?
L171-172: "This is caused by differences in interpretation from different analysts, sensor resolutions (Landsat vs. Sentinel-2) and mapping conditions (snow, clouds and shadow) rather than by growing glaciers.". Is there any reference for this? I also wonder whether a similar overestimation in 2015 likely also applies to the other (larger) glaciers?
L179-180: "Generally, elevation changes were largest at an elevation of around 1600 m (dominated by Region 6) and decreasing towards higher elevations (Figure 4).". Can this elevation also be calculated for P2 (2000-2015)? With glaciers retreating I would expect this elevation to migrate upwards as well with time.
L181-182: "The largest elevation changes (-105 m) were found at 1650 m in the western (Figure 4a) and at 2250 m (-65 m) in the eastern Alps (Figure 4b)". This does not seem consistent with the 1600 m for the entire Alps (?).
Figure 3: In the caption please indicate which period it applies to (P1 or P3).
L191: "238 km³". Why is the 5% for the missing glaciers only added to the upper bound and not to this lower bound?
Figure 4: This figure could be improved. The resolution is currently rather poor, reducing readability. Furthermore, the right axis is missing in both panels.
L210-212: "Overall, ... (Figure 3c).". Since these glaciers are not experiencing frontal ablation, elevation change rates, and trends of those, could be compared with longterm surface mass balance trends. I am not well up to date regarding available long-term SMB datasets for the Alps, but when available it would be interesting to compare long-term SMB trends with the surface height change trends presented here.
L216-217: "the regional variability ... Hugonnet et al. (2021).". It could be good to indicate the average relative changes compared to P1 for these as well.
L217-218: "Further research is necessary to investigate what causes the differences among the available datasets". It could be considered to include a comparison with surface mass balance studies, e.g. Davaze et al. (2020; https://doi.org/10.3389/feart.2020.00149), in order to have some independent source to compare to. I would also like to see some more discussion on potential causes of the discrepancies between the available datasets.
L220: "The absolute volume change rates...". I suppose this is based on the use of Hugonnet et al. (2021) for P2? It could be good to clarify since in the previous paragraph different surface height change datasets were considered.
L265: "which is certainly not the case in 2015". I may be missing the point here. Why are 2015 conditions relevant for the Haeberli and Hoelze estimate (published in 1995)?
L277: "depending on the specific characteristics of a glacier". It could be interesting to discuss these characteristics a bit further (as it might give some directions for future work).
L300: "an approximate regional average of 1850 has been used.". Larger response times of large (thick) glaciers will likely cause many of those glaciers to have a later LIA maximum extent date than small glaciers. Although this may be hard to account for in the approach, at least some discussion on the size dependence would be good.
L310-313: "The observed change in median elevation of 143 m would translate to a temperature increase of 0.84 to 1.43 °C, depending on the atmospheric lapse rate applied (Haeberli et al., 2019; Kuhn, 1989; Rolland, 2003; Zemp et al., 2007). This is lower than the 1.5° and 1.6° temperature increase determined by Begert and Frei (2018) and Auer et al. (2007) for Switzerland and the Alps, respectively.". This needs some clarification. I assume the temperature increase in the first sentence is purely due to the elevation drop. But what is the 1.5-1.6 deg C change in the second sentence referring to? Does it include both anthropogenic warming and elevation-drop induced warming?
L317-327: Interesting discussion on peak runoff!
Conclusions: The conclusions section gives a good summary of the work. However, I do miss some recommendations for future work. Would there for example be room for inverse (ice flow) modelling methods to generate physically more robust LIA surface topography?
Textual revisions:
L10 (and elsewhere): digitising --> digitization

L17: In the mean --> On average

L18: 1600 m --> 1600 m above sea level (m a.s.l.)

L55: "first" --> "first complete"

L60: "Regional subdivision" --> "Study regions"

L99: "DEM" --> "Digital Elevation Model (DEM)"

L112: "the total glacier volume" --> "the contemporary total glacier volume"

L141: "Uncertainties of the bed topography impact" --> The impact of bed topography uncertainty"

L143: Add "volume" between "overall" and "uncertainty".

L160: "reconstructed to" --> "estimated at"

L170: "shrunk" --> "shrank"

L196: "altitude" --> "altitude range"

L201: "are" --> "is"

L203: "Increase in change rates". Please change to something more specific (indicating the variables it refers to).

L252: "if considering" --> "when including"

L261: "would" --> "could"

L269: "estimated" --> "estimate"

Figure 6b: Please mention that changes are negative.

L330: "coverage by area" --> "areal coverage"

L331: "these glaciers" --> "all glaciers" (otherwise it refers to the missing glaciers)

L356: "GILIMS" --> "GLIMS"

Citation: https://doi.org/10.5194/egusphere-2024-989-RC1
- AC4: 'Reply on RC1', Johannes Reinthaler, 16 Sep 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-989/egusphere-2024-989-AC4-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-989-AC4
RC2:
'Comment on egusphere-2024-989', Arindan Mandal, 01 Sep 2024

Dear authors and editor,
Please find the attached pdf file which includes my comments.

Best regards,
Arindan Mandal

Citation: https://doi.org/10.5194/egusphere-2024-989-RC2
- AC5: 'Reply on RC2', Johannes Reinthaler, 16 Sep 2024
  
  The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-989/egusphere-2024-989-AC5-supplement.pdf
  
  Citation: https://doi.org/10.5194/egusphere-2024-989-AC5

Johannes Reinthaler and Frank Paul

Supplement

https://doi.org/10.5194/egusphere-2024-989-supplement

Johannes Reinthaler and Frank Paul

Viewed

Total article views: 1,289 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
794	322	173	1,289	116	18	18

HTML: 794
PDF: 322
XML: 173
Total: 1,289
Supplement: 116
BibTeX: 18
EndNote: 18

Views and downloads (calculated since 03 May 2024)

Month	HTML	PDF	XML	Total
May 2024	328	155	34	517
Jun 2024	138	59	16	213
Jul 2024	96	34	8	138
Aug 2024	60	15	2	77
Sep 2024	94	27	20	141
Oct 2024	39	13	48	100
Nov 2024	31	12	41	84
Dec 2024	8	7	4	19

Cumulative views and downloads (calculated since 03 May 2024)

Month	HTML	PDF	XML	Total
May 2024	328	155	34	517
Jun 2024	138	59	16	213
Jul 2024	96	34	8	138
Aug 2024	60	15	2	77
Sep 2024	94	27	20	141
Oct 2024	39	13	48	100
Nov 2024	31	12	41	84
Dec 2024	8	7	4	19

Viewed (geographical distribution)

Total article views: 1,249 (including HTML, PDF, and XML) Thereof 1,249 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 11 Dec 2024

Short summary

Since the end of the Little Ice Age (LIA) around 1850, glaciers in the European Alps have melted considerably. We collected LIA glacier extents, calculated changes using geoinformatics, and found a 57 % decrease in area (4211 km² to 1806 km²) and a 65 % decrease in volume (281 km³ to 100 km³) by 2015. The average glacier surface lowering was 43 m. After 2000, elevation change rates tripled. Over 1800 glaciers melted away completely, impacting entire regions.


Total:	0
HTML:	0
PDF:	0
XML:	0