Review of “Brief communication: A framework to classify glaciers for water resource evaluation and management in the Southern Andes” by Shaffer & MacDonnell

SUMMARY

This brief communication proposes a new classification of the glacier landforms present in the Andes. The classification focuses on the sensitivity of the landforms to climate change and their hydrological impacts rather than purely on their geomorphological traits. The proposed classification is suggested to contribute to the discussion on the development of glacier protection legislation in both Chile and Argentina, which up to now have been relatively unsuccessful.

GENERAL COMMENTS

In general, the manuscript is well written and the message the authors intend to convey is clear. However, I have to say I am slightly confused about the intention of this communication. On the one hand, I do see the benefit of publishing this work in The Cryosphere, as this discussion may also exist in other parts of the world and a consensus in identifying glacier sensitivity from a policy standpoint could be beneficial. To serve this purpose, I do think the manuscript in its present form is (too) much focused on the Andean case. On the other hand, I wonder whether (the message of) this manuscript wouldn’t be a better fit for a journal or other medium that allows direct targeting of the intended audience, i.e. policy makers, nature conservatists and water resource managers in the respective countries. I am not saying I do not see the benefit for TC and a “general” audience, but a more general focus would better support that.

I am also wondering how relevant it really is to identify the different landform types from a legislation perspective. Apparently, the political discourse has not yet been fruitful with respect to the GPL, even when just considering them as a single entity. Wouldn’t introducing a system of sensitivities complicate things even further? In my opinion, the current manuscript does not express clearly enough how the introduction of the proposed classification would benefit the discourse around glacier protection, how it would benefit drafting related legislation, and how water resource management will be improved as a result.

I agree that the (quite minor) redefinition in classes defined by the authors with respect to traditional geomorphological categories of clean-ice, debris-covered and rock glaciers could improve assessment in terms of sensitivity and hydrological impacts up to a certain extent. However, in essence, the classification is still just based on a simple interpretation of the surface morphology, which is an oversimplification. This results in the straightforward and broad classes “sensitive” vs “insensitive”, which may be too much of a black and white approach to be really useful in practice. High heterogeneity and variability exist among glaciers in their sensitivity and hydrological response, and this is for a considerable part irrespective of glacier surface type. It may be due to other geomorphological specifics of a glacier that are not considered in the proposed classification (e.g. slope, elevation, bed lithology, aspect etc.), but also due to differences in local climate, local anthropogenic disturbances, and possible feedbacks therein. Could some of these components be included somehow? Wouldn’t an (even simple) modelling approach allow for a more informative estimation of the actual sensitivity of the glaciers? I would suggest the authors to at least elaborate on the limitations of such a simple classification and place it into a context of other, more developed approaches such as regional and/or individual glacier modelling. “Advanced” approaches would also be better to identify potential tipping points and transient effects, which could be very important arguments in policymaker discussions and conveying the urgency of expected changes in hydrology.

I do not really understand the difference between landform and glacier used in the manuscript. A glacier seems to me as single entity, especially since it is hydrologically connected, but here it is suggested that a glacier is actually a landform that can consist of multiple glacier types. I would suggest using a better description of and distinction between these terms. Also, taking the most sensitive part from a geomorphological perspective in a “hydrologically-connected” case to represent the sensitivity of the entire glacier/landform is not necessarily valid. The system should rather be classified as a whole. This goes back to my previous point: will this simple classification adequately represent sensitivity of the existing wide range of glaciers and glacier systems? For all glaciers, but particularly for a multi-type ones, sensitivity very much depends on the type of external forcing that causes a potential disturbance. If, for instance, the lower part of a glacier system is heavily debris-covered, it could be relatively insensitive to climate warming due to the insulation the debris provides in the ablation zone, but could be highly sensitive to processes that affect accumulation zone albedo such as the snowfall frequency or black carbon deposits.

SPECIFIC COMMENTS

L159-166. This is a good point, but it further reveals the complications with the sensitivity classification. I agree that a protection plan should evolve over time, but it should ideally already account for these temporal processes and effects from the get go. Is there any way temporal evolution could be included in the sensitivity classification approach? How would this affect the discourse and development of GPL?

L261-263. I am not sure why it is necessary or even desirable that non-experts can determine the sensitivity of a glacier. A well-developed database of glacier sensitivities created by experts using thorough analysis will deliver a much more insightful indication of the sensitivity of glaciers in a region or catchment and will serve policymakers better.

General comments:

1) The manuscript is well written and presents a sound classification schema of glaciers based on their sensitivity to environmental change. The methodology has solid support in the literature, and the consideration of this new classification is likely to be a valuable contribution to the management of glaciers, especially concerning their hydrological services.

2) The work is presented as a contribution towards the development of GPL, particularly as a solution to inadequate definitions in the Chilean GPL projects. In that sense, there is no clear articulation between the proposed classification and the GPLs. Both Chilean and Argentinean GPLs avoid conflict and ambiguities by protecting all glaciers equally, regardless of their type, size, location or debris cover. In that context, it is hard to understand how this classification schema can help the design of a law proposal of consensus and without the "legal issues" mentioned for the Argentinean GPL. If the authors propose a type-dependent level of protection (as stated in lines 224-225, 261), that should be clearly stated and followed with well-elaborated reasoning to support that proposal. Arguably, a type-dependent level of protection will only complicate things, especially given that the classification is sometimes ambiguous (lines 115-116) and changes with time (lines 159-164). In some sections, it even seems that the authors suggest a case-by-case assignation of the level of protection (lines 274-278). Much emphasis was put on the usefulness of the proposed classification for glacier management. However, the Argentinean GPL and Chilean GPL proposals aim to protect glaciers, not to manage them.

If the authors are mainly suggesting a type-dependent monitoring program or the addition of this classification to national inventory fields (as stated in line 271), this should be clearly stated from the start. In such a case, they should also include a more detailed explanation of how this classification will help water resources management and a throughout motivation of the methodology. For example, why a classification is better than a "sensitivity index" or case-by-case modelling.

3)Following the facts detailed in lines 31-32, it seems inaccurate to refer to glaciers/landforms as sensitive/insensitive. The differences seem to be related only to the timescales of their response to environmental changes. Maybe fast/slow response might be better terminology.

4) The main proxy to assess glacier change is mass balance, which depends on accumulation and ablation. However, this works seems to focus entirely on the ablation part of the equation.

In avalanche-fed glaciers, which is often the case for categories 2 or 3 (semi-sensitive and insulated), there could be a high climatic sensitivity associated with the snow accumulation on surrounding slopes that are not even part of the glacier. While such glaciers would "melt away slowly" due to their debris cover, their mass gain mechanism might have a very high sensitivity to environmental changes. In these cases, their water storage capability at inter-annual timescales would also have a high sensitivity to environmental change.

5) The use of the term "landform" makes the manuscript very confusing. While it can refer to anything (a glacier, a ridge, a mountain), it is often used to refer to a glacier, where the direct use of the term "glacier" would make the text much clearer. In some cases, for the same glacier the text says that it is a landform composed of multiple glacier types, and that it is a glacier composed of multiple landform types (line 118: "Where a landform is made up of multiple glacier types (Fig. 1a [Tapado Glacier])", lines 125-126 "Tapado Glacier [Fig. 1a] is made up of the three distinct landform types..."). Other sections use the concept of "glacier morphology" (line 161). More consistent use of the terminology is necessary: "Glacier" and "surface-type" could be better concepts to use (instead of randomly interchange either of those by "landform").

6) In the context of GPL and glacier inventories. It seems that the authors propose the use of their methodology nationwide or throughout the Andes. However, the examples presented in figures and Table 1 are biased to the semi-arid Andes; the same is true regarding the accuracy check proposed in line 234. All examples are within four degrees of latitude. It must be clear what is the geographical area for which this methodology has been designed. If the application area is the whole of the Andes, the authors should address the different challenges posed by tropical and Patagonian glaciers.

Specific comments (numbers refer to manuscript version 2) :

7: In the context of this paragraph and in particular the GPLs, "landform types" have a very different and more specific meaning than used in the rest of the text, as the most controversial definitions that have hindered consensus of the Chilean GPL are the definitions of Glacier, Periglacial, and permafrost. However, "landform types" in the manuscript refers interchangeably to glaciers or parts of a glacier with a distinct surface type (based on debris cover). This difference gives the impression to the reader that this work offers a direct solution to the definitions controversy that, has been in part, the cause of the lack of consensus, which is wrong.

21-22: Given that the authors seem to be opening the discussion over the idea of not protecting all glaciers equally but depending on their hydrological behaviour. It seems very important to elaborate on what legal issues have hindered the application of the Argentinean GPL, or at least give a reference for that affirmation.

23-24: This requires further elaboration. It is unclear how distinguishing between glacier types can reduce the legal ambiguity. In general, one would think that the current approach of Chilean and Argentinean GPLs (protecting all glaciers regardless of type) is less ambiguous than differential protection based on a glacier classification schema.

39: The switch between the "glacier" terminology and the use of "landform" should be explained here. Otherwise, simply keep using "glacier."

77-80: It seems against the objectives of this work to base the threshold of debris thickness on a single glacier. Arguably, debris type can have a significant influence, as well as the partitioning of the different melt processes affecting a glacier. In areas where sublimation is the primary melt process, a thin layer of debris might be enough to reduce melting significantly. In other cases, such as the temperate glaciers of New Zealand and Patagonia, a large amount of the melting is due to rain, and perhaps a much thicker debris cover is required to reduce melt rates. Pirámide glacier might be representative only of glaciers where shortwave radiation is the dominant melting process.

121-123: Again, it seems against the objectives of this work to include ambiguous criteria like this (what is "very minor"?).

Figure 2: Please include coordinates or some ID (either in the figure or caption) for all unnamed glaciers (b-f). Alternatively, add to the caption a reference to the additional information available in Table 1.

144-149: It is confusing to use the term "landform" when you mean "glacier". Unless the authors want to refer to different sections of a glacier but with different surface types, however, if that is the case, it does not make sense to say that the insulated part of Tapado Glacier is insensitive to environmental change while its accumulation area is a sensitive "landform".

159: "It is likely" seems a euphemism for something that unquestionably will happen.

Table 1: What is the point of comparing this article classification with DGA/IANIGLA classification? Each of these is classifying completely different attributes of the glacier: Glacier sensitivity to environmental change in this article, glacier shape/main characterizing feature for DGA, and glacier debris cover for IANIGLA.

214: Which are the distinct hydrological roles? The authors only point to differences in the timescales and the degree to which these glacier types play a role as water reservoirs.

227-229: While that might be more objective, it seems a nightmare from a legal point of view. One can picture a development project affecting a sensitive glacier because a logistic regression happens to assign it to the wrong category.

256-257: As for line 214, it seems that "role" is not the best word to distinguish between the hydrological effects of different types of glaciers.

Review of "Brief communication: A framework to classify glaciers for water resource evaluation and management in the Southern Andes".

Schaffer and MacDonell present a new classification of glaciers in terms of their hydrological importance. The new classification is proposed in the context of the ongoing discussion of the Glacier Protection Law (GPL) in Chile and how this classification could be useful in defining the level of protection. The classification is determined in terms of its sensitivity to environmental changes based on the percentage of debris-cover area and surface characteristics. Based on this, the authors mention that this classification could allow different degrees of protection depending on their sensitivity which, is closely related to the hydrological role. The classification includes debris-free glaciers (highly sensitive and greater contribution to streamflow), debris-cover glaciers, including thermokarst features and zones of exposed ice in the surface (semi-sensitive) and rock glaciers, classified as insulated from the environment and hence with a lower hydrological role. Some examples are given for the semiarid environment in Chile and Argentina. The manuscript is well written and could be an important contribution to the discussion of GPL but also to understand the hydrological role of glaciers and its differences.

My main concern, however, is related to the classification and the criteria used and how useful could be to GPL discussion. I understand this as a highly complex topic and probably there could be several classifications criteria depending on the researchers. So please, take my comments on this topic as recommendations. Having said this, I recommend clarifying some points in the manuscript in order to understand the proposed classification.

General comments

Classification

My general view is that the classification is simple and do not captures the diverse nature of glaciers in the Andes. In L115-116 it is mentioned that the guidelines need to be evaluated on a case-by-case basis. If this is the case, and I agree with previous reviewers, things seems to be more complicate following this classification. The hydrological role of a glacier, it is probably a concept that most of non-experts understand as a key role of the glaciers. In this sense, an explanation of the differences in time responses it is more adequate to introduce in a context of a GPL discussion in order to fully protect all glaciers. However, I agree that this classification can be useful to water resource management (L66-68).

Concepts

The concept of hydrological role is not well defined in the manuscript. Following the explanation in L144-157, it seems that it is related mainly to the contribution of each type of glacier to streamflow at an annual scale. I suggest a clearer definition of what the authors mean by “hydrological role” including temporal and spatial scales and also the potential contribution to groundwater. The authors mentioned (L30-33 and L155-157) that insulated glaciers (rock glaciers) storage and delay the runoff. This is an important point as the hydrological role and importance of these glaciers have a different time scale in comparison to debris-free glaciers and must be included to define the hydrological role.

“Sensitivity to environmental changes” I understand that probably the use of “environmental changes” is used to include the atmospheric drivers of the melt as well as the feedback (positive or negative) that the debris-cover and glaciers surface characteristics exert on melt rates. However, the concept of “environmental changes” is wide and includes several other factors. I suggest clarifying what exactly means “sensitivity to environmental changes”. Maybe, constrain this concept will allow a clearer link between the classification and the hydrological role. I think in a concept like “sensitivity to melt drivers” or probably something better.

Clarification on the level of protection

In order to avoid confusion, I suggest including, explicit, the order of the level of protection for each type of the classification i.e. the type that needs more protection according to your classification.

Specific comments:

L20 “Sendado” to “Senado” (also in the reference list).

L274-278: Although is not the focus of the paper, I think that the other values of glaciers mentioned here must be included in the Introduction. This manuscript is concentrated on the meltwater contribution to runoff, which of course is important, but as mentioned, glaciers also play other key roles.