the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 03 Nov 2023
Assessing the glacier projection uncertainties in the Patagonian Andes (40–56° S) from a catchment perspective
Abstract. Glaciers are retreating globally and are projected to continue to lose mass in the coming decades, directly affecting downstream ecosystems through changes in glacier runoff. Estimating the future evolution of glacier runoff involves several sources of uncertainty in the modelling chain, which to date have not been comprehensively assessed on a regional scale. In this study, we used the Open Global Glacier Model (OGGM) to estimate the glacier evolution of each glacier (area > 1 km2) in the Patagonian Andes (40–56° S), which together represent 82 % of the glacier area of the Andes. We used different glacier inventories (n = 2), ice thickness datasets (n = 2), historical climate datasets (n = 4), general circulation models (GCMs; n = 10), emission scenarios (SSPs; n = 4), and bias correction methods (BCMs; n = 3) to generate 1,920 possible scenarios over the period 1980–2099. For each scenario and catchment, glacier runoff and melt on glacier time series were characterized by ten glacio-hydrological signatures (i.e., metrics). We used the permutation feature importance of random forest regression models to assess the relative importance of each source on the signatures of each catchment. Considering all scenarios, 61 % ± 14 % of the catchment area (30 % ± 13 % of glacier area) has already peaked in terms of glacier melt (year 2020), and 43 % ± 8 % of the catchment area (18 % ± 7 % of glacier area) will lose more than 80 % of its volume this century. Considering the melt on glacier signatures, the future sources of uncertainty (GCMs, SSPs and BCMs) were the main source in only 18 % ± 21 % of the total catchment area. In contrast, the reference climate was the most important source in 78 % ± 21 % of the catchment area, highlighting the importance of the choices we make in the calibration procedure. The results provide a basis for prioritizing future efforts (e.g., improve reference climate characterization) to reduce glacio-hydrological modelling gaps in poorly instrumented regions, such as the Patagonian Andes.
- Preprint
(32764 KB) - Metadata XML
-
Supplement
(27949 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2325', Anonymous Referee #1, 10 Dec 2023
In the manuscript “Assessing the glacier projection uncertainties in the Patagonian Andes (4-56 S) from a catchment perspective”, a large set of glacier model simulations obtained with the OGGM model is used to assess the uncertainty in glacier runoff and glacier melt projections. The study looks at the effect of different glacier outlines, different glacier volume estimates, various historical climate datasets, different GCM, different emission scenarios and different bias correction methods. Each of these different datasets are discussed and it is shown how they vary, mainly focusing on their spatial patterns. Using a random forest regression method, the relative importance of each of these “model choice” uncertainties is examined. The study concludes that the reference climate is the most important source of uncertainty for a range of glacio-hydrological signatures, even for signatures that represent a signal beyond the reference period.
Overall, I think that the study presents an impressive amount of model run comparisons at a large regional scale and shows clear insights into spatial differences of glacier volume and runoff changes in the Patagonian Andes. Moreover, the effect of each source of uncertainty on the glacio-hydrological signatures nicely illustrates their variable importance for different aspects of the change in glacier runoff. However, I feel that some parts of the manuscript could be improved, such as the use of “catchments”, and the discussion section, which should describe more the implications and possible hypotheses of the findings, rather than only a summary of the findings and comparison with other studies. Please find below a list of more detailed comments, also explaining these two examples.
Abstract:
- The introduction of “catchments” starts here in the abstract. Without stating what “catchment” represents in this study, it is quite hard to follow. The reader doesn’t know if the study is about a few catchments, or many, and how they are delineated/defined. Accordingly, all the statements with “xxx% of the catchment area” are difficult to interpret. On a more general note, it did not became clear why catchments are used in the study? Most of the results are rather described on the basis of the hydrological zones. In the study area description, it is noted “847 catchments were selected” without any further information. On the basis of what were these catchments defined? Since the study is not using any downstream information or non-glacierized catchment info (“only using glacierized grid cells”), I think the “catchment” part should not be part of this study. Results could then be just presented as xx% of glaciers show this and that, while also keeping the aggregation level of the hydrological zones.
Introduction:
- L44 undergoing “a” shift – shouldn’t shift not be an increase, as otherwise there is no increase in risk? Or does it relate to larger volumes that are stored before bursting out?
- L65 “have had to be used” – have been used? Also check the “but” which is more an additional problem than a contrast?
- I would suggest to move table 1 to the SI. The abbreviations in the first column are not yet clear and in general the aim/message of the table did not become clear in the introduction
- L93 “to project the evolution of the glacier area” – the model is not only used for projecting the area?
Study area:
- L104-105 “crucial”, “essential” in the same sentence, maybe one is enough?
Methods:
- Step 1 of the calibration procedure – I wondered if it should be discussed what the effect is of only calibrating the melt factor/temperature sensitivity using geodetic mass balances. If the precipitation is off (e.g. too little) then the temperature sensitivity might be smaller, to compensate?
- Step 2 – the description is a little bit unclear, is the gradient adjusted? Or what is meant with “a mass balance residual to add”?
- Step 3 – maybe the equation could be given here? I think it is similarly known as equation 1, but helps the reader understanding what is needed for the inversion and at what scales it is applied
- Step 4 – what is a “constant mass balance run”?
- Figure 2 – I think it would be helpful to add the number of the steps in the figure. And what is the arrow from reference climate to bed inversion? It may refer to one step before that?
- L259 “according to the location of the glacier terminus” – I am not sure to understand this addition? Maybe some more explanation of glaciers that are crossing catchment/hydrological zone borders could be added here
- L291 here the text could benefit from some explanation/ careful discussion how streamflow metrics can be used to apply on aggregated glacier runoff data (i.e. glacier runoff is not the same as downstream streamflow and so their effect on the aquatic ecosystem is not 1:1 comparable)
- Table 2 misses information about which period was used to calculate the signatures, apart from the ones that explicitly state “ref and future period”
- Header of 3.5 – maybe choose the titles of section 3.3 and 3.5 in such a way that it is more clear what their different content is
- L299 – where comes 329 from?
- In the calculation of RMSE, what is the baseline? i.e. how is RMSE calculated?
Results:
- Figure 2: possibly remove the blue/gray background so the results are better visible
- L335 – why was the glacier thickness divided by the catchment area, while all other variables are focused on the glaciers/glacier grid cells?
- L374 – maybe I missed it, but what are “the main catchments”?
- L408 “historical sources” – climate?
- L435-L440 Does isolating the melt from glacier runoff result in more or less uncertainty? There is a hint that melt only has an effect of temperature, but it also states that precipitation can compensate for the change in melt? The 60% suggests that 40% of the glacier runoff is generated from off-glacier melt, or liquid precipitation on+off glacier?
Discussion:
- L484 “acquisition dates” – this sentence could benefit from more explanation about what the difference is in both RGI outlines with respect to acquisition dates.
- L487 “this threshold” – do you mean a 10% threshold for volume?
- L494 – Isn’t this paragraph suggestion a contradiction? It would help starting the paragraph like that.
- L510 – a space needs to be inserted before “The selected….”
- L512 “the ten selected….of the assessed warming – not sure how this sentence fits in?
- L526 Why are the “older” estimates named after the newer estimate and referred to as “recently”?
- L539 “of its current volume” – although it is clear that the study only discusses glaciers, the confusing use of “catchments” means that here there “glacier” needs to be added
- For the discussion in general, I found it sometimes hard to follow what was described in the different parts. The first part discusses the uncertainties, but the other two parts as well. Maybe by less repetition of the results, and more focusing on the implications of the results would help in restructuring the discussion. I think the comparison with other studies is mostly well described, but would also benefit from an additional thought on what are the implications of that it agrees well or not.
- I was wondering if there should be maybe a discussion on the way such glacier models are calibrated? If all model runs are calibrated equally well (RGI areas and glacier volumes are used for calibrating), then how come the results are so different, especially regarding reference climates? These parameters propogate in all other simulations, right? This must then come from processes that are not captured when looking at annual and long-term metrics only when calibrating model parameters? Or maybe there are other processes that are relevant to discuss for improving glacier modelling?
- L594 – possibly add that Mackay et al. deals with glacio-hydrological modelling in Iceland
Conclusions
- L610 – “differences”, “different”, “varied” in one sentence – consider rephrasing.
- L613 isn’t it a separate point, the one on reference climate being most important for uncertainty?
- L622 “tended to converge towards an overall decrease” – decrease of what?
- L625-628 double?
- L629 – what is meant with “local”?
- L634 “relative contribution of non-glacial water sources” – shouldn’t future studies not focus on the dynamics of these sources, rather than the relative contribution? And what should future studies do with these non-glacial water sources? The follow up sentence does not directly fit here, i.e. there is some gap between knowing a relative contribution and understanding other catchment stores. Maybe a general sentence on extending the scope from glaciers to downstream hydrology fits better?
Citation: https://doi.org/10.5194/egusphere-2023-2325-RC1 - AC1: 'Reply on RC1', Rodrigo Aguayo, 07 Feb 2024
-
RC2: 'Comment on egusphere-2023-2325', Anonymous Referee #2, 19 Dec 2023
- AC2: 'Reply on RC2', Rodrigo Aguayo, 07 Feb 2024
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 533 | 240 | 27 | 800 | 53 | 26 | 18 |
- HTML: 533
- PDF: 240
- XML: 27
- Total: 800
- Supplement: 53
- BibTeX: 26
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1