the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Mar 2026
Automated Avalanche Hazard Indication for Southeast Alaska
Abstract. Snow avalanches represent a significant yet poorly characterized natural hazard in Alaska, where risk assessment is limited by extreme data scarcity, rapidly changing climate conditions, and land-use policies that often inadequately account for environmental uncertainty. We address these challenges in Southeast Alaska by developing the region’s first systematic, large-scale avalanche hazard indication maps to support public safety, infrastructure planning, and land-use decision-making. To overcome sparse observational records, we developed a hybrid modeling framework that integrates downscaled reanalysis for historical baselines (1981–2010) with dynamically downscaled climate projections for mid-century conditions (2031–2060). More than 3.5 million avalanche simulations were performed using RAMMS::LSHIM driven by downscaled snow inputs. Forest landcover masks were incorporated to represent both suppression of avalanche release and vegetation-induced braking during runout, recognizing that these simplified effects remain sensitive to landcover classification accuracy and assumed release-area configurations. The resulting maps reveal a heterogeneous response of avalanche hazards to climate change. At lower elevations, hazard extents generally decrease as warming temperatures shift precipitation from snow to rain. In contrast, select high-elevation areas of northern Southeast Alaska are projected to experience increased runout, where persistently low temperatures, combined with enhanced atmospheric moisture, lead to greater maximum snowfall. Collectively, these results provide the first region-wide, climate-informed assessment of avalanche susceptibility in Southeast Alaska, establishing a critical foundation for hazard adaptation, infrastructure resilience, and future mitigation strategies across Alaska’s sub-Arctic landscapes.
Competing interests: Author Marc Christen is the founder of RAMMS, AG. Author Yves Bühler is an Editor at NHESS.
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 paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.- Preprint
(5006 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1345', Anonymous Referee #1, 22 Jun 2026
-
RC2: 'Comment on egusphere-2026-1345', Anonymous Referee #2, 01 Sep 2026
Recommendation: Major Revision
1. Does the paper address relevant scientific and/or technical questions within the scope of NHESS?
Assessment: Yes — clearly.
The manuscript fits well within the scope of Natural Hazards and Earth System Sciences. It deals directly with:
- snow-avalanche hazard;
- numerical modelling of mass movements;
- hazard-indication mapping;
- climate-change effects on natural hazards;
- implications for infrastructure and land-use planning;
- hazard assessment in a data-sparse environment.
The central scientific problem—how to generate regional avalanche hazard information where conventional observations and site-specific assessments are unavailable—is both scientifically relevant and technically important.
The focus on climate-related changes in avalanche hazard also aligns particularly well with the interdisciplinary scope of NHESS.
No major concern under this criterion.
2. Does the paper present new data and/or novel concepts, ideas, tools, methods or results?
Assessment: Yes, particularly in terms of application, scale and resulting datasets.
The fundamental components of the methodology are not themselves entirely novel. RAMMS: LSHIM, automated PRA delineation and regional hazard-indication mapping have been developed previously, especially by Bühler et al. (2022 – cum bibl.).
However, the manuscript presents substantial novelty in the following areas:
- application of the methodology to Southeast Alaska - a particularly data-sparse, maritime/sub-Arctic environment;
- integration of historical reanalysis and future dynamically downscaled climate information;
- regional-scale implementation involving more than 3.5 million avalanche simulations;
- development of a regional dataset where no comparable systematic hazard-indication product previously existed;
- comparison of modelled avalanche hazard extent between historical and future climatic periods.
This work shows moderate methodological novelty but high applied and regional novelty. The manuscript should articulate this distinction more explicitly.
3. Are these up to international standards?
Assessment: Partly; substantial improvement is required.
The general modelling framework is consistent with established international avalanche-modelling practice, particularly through the use of RAMMS::LSHIM and established PRA methodologies.
However, several aspects currently fall short of the level of methodological robustness normally expected for quantitative international hazard assessment:
- the treatment of 30- and 300-year scenarios is not probabilistically robust;
- uncertainty is discussed but not quantified;
- validation is predominantly qualitative;
- climate uncertainty is represented by only one model/scenario combination;
- vegetation thresholds are partly heuristic;
- effective information resolution is occasionally overstated;
- terms such as risk are used where the analysis actually represent hazard indication or susceptibility.
Thus, the study framework is internationally relevant, but the current implementation and interpretation require strengthening before the results can fully satisfy international standards of quantitative hazard assessment.
4. Are the scientific methods and assumptions valid and outlined clearly?
Assessment: Generally understandable, but only partly validated and not always sufficiently justified.
The overall workflow is described clearly enough to understand its main structure: DEM → forest mask → snowfall forcing → PRA generation → release-depth calculation → RAMMS simulation → avalanche extent/intensity mapping. However, several assumptions require more rigorous justification.
4.1 Snowfall downscaling
The original WRF information is at approximately 4 km resolution. It is remapped to 10 m, Gaussian-smoothed, and corrected using elevation. This produces a raster with 10 m spacing, but not meteorological information physically resolved at 10 m. The paper must state this distinction clearly throughout.
4.2 Lapse-rate assumption
The final implementation effectively uses a constant snowfall lapse rate of 0.0923 mm m⁻¹ throughout Southeast Alaska. This is a strong regional simplification that requires stronger empirical support and sensitivity testing.
4.3 Release depth
Release depth is calculated as: without the more complete corrections used in the SLF standard procedure. Given the direct influence of release depth on avalanche volume and runout, this assumption should be subject to sensitivity analysis.
4.4 Vegetation assumptions
The treatment of Deciduous, Mixed and Evergreen Forest and the use of a 30% vegetation threshold for filtering low-elevation avalanches are insufficiently justified quantitatively.
4.5 Recurrence intervals
The relationship between maximum 30-year snowfall and the assigned 30-/300-year avalanche scenarios is not sufficiently rigorous and represents one of the manuscript’s most important methodological weaknesses.
This criterion requires major revision.
Thus, the study framework is internationally relevant, but the current implementation and interpretation require strengthening before the results can fully satisfy international standards of quantitative hazard assessment.
5. Are the results sufficient to support the interpretations and the conclusions?
Assessment: Only partly.
The results support several qualitative conclusions reasonably well:
- avalanche hazard is spatially extensive in Southeast Alaska;
- vegetation classification strongly affects low-elevation PRA and runout modelling;
- projected warming may reduce hazard at lower elevations;
- colder high-elevation northern areas may respond differently;
- the automated framework is useful for regional screening.
However, some quantitative interpretations are too strong.
For example, Figure 20 interprets a 10% change in affected area within a 100 × 100 m grid cell as approximately a 10 m retreat of maximum avalanche runout.
Given the combined uncertainty in:
- 4 km atmospheric forcing;
- snowfall downscaling;
- land-cover data;
- PRA geometry;
- release depth;
- friction parameters;
- wind redistribution;
- model formulation;
This interpretation gives a misleading impression of metre-scale predictive precision.
Likewise, statements about future hazard should be framed as outcomes from the specific CCSM4–RCP8.5 simulation considered, rather than as deterministic predictions of future conditions.
6. Does the author reach substantial conclusions?
Assessment: Yes, but some conclusions should be reformulated more cautiously.
The manuscript reaches important conclusions regarding:
- spatial variability of avalanche hazard;
- implications of climatic warming;
- regional differences in future snowfall extremes;
- the usefulness of automated hazard-indication mapping in data-poor regions.
These are substantive scientific conclusions.
However, the statements that the outputs provide an “immediately useful basis for hazard-informed infrastructure planning and land-use decision-making” and a “defensible regional baseline” are somewhat stronger than warranted by the present validation.
It is recommended to describe the outputs primarily as: “a regional screening and prioritisation tool identifying areas requiring detailed site-specific hazard assessment”.
The maps should not be presented as substitutes for detailed engineering-scale avalanche hazard zoning.
7. Is the description of the data, methods, experiments, calculations and results sufficiently complete and accurate to allow reproduction?
Assessment: Partly, but not yet sufficiently.
The broad workflow is reproducible conceptually, but full computational reproducibility is limited.
The manuscript states that the complete workflow requires multiple operating systems and proprietary software, including ArcGIS, eCognition and IDL. The Code Availability section indicates that only portions of the code are publicly available.
For adequate traceability and reproducibility, the authors should provide:
exact version of RAMMS::LSHIM;
- exact friction parameterisation for each PRA class;
- complete PRA decision-tree settings;
- all filtering thresholds;
- exact software versions;
- WRF variable names and preprocessing steps;
- kernel implementation and boundary treatment;
- all scripts required to reproduce the snowfall downscaling;
- forest-mask generation rules;
- release-depth input tables;
- post-processing thresholds;
- configuration files;
- a workflow diagram;
- ideally, representative complete input/output examples.
The distinction between results availability and workflow reproducibility should also be made clearer.
This criterion requires major revision.
8. Does the title clearly and unambiguously reflect the contents of the paper?
Assessment: Mostly yes.
The title: Automated Avalanche Hazard Indication for Southeast Alaska is concise and generally accurate. However, the manuscript also contains a significant climate-change component. A more informative title might be: “Automated Avalanche Hazard Indication Mapping for Southeast Alaska under Historical and Future Climate Conditions” or: “Automated Regional Avalanche Hazard Indication Mapping for Southeast Alaska under Present and Future Climate Conditions”.
That said, the current title is acceptable and does not require mandatory change.
9. Does the abstract provide a concise, complete and unambiguous summary?
Assessment: Generally yes, but some wording should be moderated.
The abstract effectively summarises:
- the motivation;
- the data-scarcity problem;
- the modelling framework;
- the historical/future comparison;
- the number of simulations;
- the main regional climate response.
However, several aspects need improvement.
Recommended revisions: The abstract should explicitly state that:
- the future projection is based on a single CCSM4–RCP8.5 realisation;
- the products are hazard-indication maps, not probabilistic hazard maps or risk maps;
- the results should be interpreted as regional screening information.
The phrase “first region-wide, climate-informed assessment of avalanche susceptibility” is generally appropriate, but the abstract should avoid implying site-specific predictive accuracy.
10. Are the title and abstract pertinent and understandable to a wide and diversified audience?
Assessment: Yes, with minor improvements.
The title is accessible. The abstract is understandable to readers familiar with natural-hazard science, but terms such as:
- PRA;
- RAMMS::LSHIM;
- dynamically downscaled climate projections;
may not be immediately understandable to the broader NHESS readership.
The abstract would benefit from reducing specialised modelling terminology slightly and briefly explaining RAMMS’s function rather than merely naming the model.
11. Are mathematical formulae, symbols, abbreviations and units correctly defined and used?
Assessment: Mostly, but several problems require correction.
The manuscript does not contain an excessive number of equations, and therefore a dedicated notation appendix is not necessary. However, several technical issues remain.
11.1 Snow-depth units
The manuscript reports typical snow depths around Juneau as approximately “0.7 m² and 1.2 m²”. These should presumably be m, not m².
11.2 Velocity threshold
The avalanche extent criterion appears in the text as velocity “at least 1 m”, whereas the intended unit is presumably: . This should be checked systematically.
11.3 Lapse-rate units
The manuscript alternates between:
- m km⁻¹;
- mm m⁻¹.
These are dimensionally equivalent in this context, but the chosen unit should be standardised.
11.4 “Snowfall” versus “snow accumulation/depth”
These terms are occasionally used interchangeably. Snowfall amount, snow-water equivalent, settled snow depth and accumulated snow depth are not equivalent quantities.
11.5 Formula interpretation
Equation (1), deriving the lapse rate from spatial gradients, should include a clearer explanation of:
- sign convention;
- physical interpretation;
- treatment of flat terrain;
- numerical handling where approaches zero.
Moderate revision required.
12. Is the size, quality and readability of each figure adequate?
Assessment: Generally good, but several figures and captions require correction and improved interpretability.
The maps are mostly useful, and the Juneau/Valdez/Cordova comparisons are particularly informative. However, several problems exist.
Figure 8 - The caption contains inconsistent historical and future periods. This must be corrected. Figure 11 - The caption refers to “maximum height, maximum pressure and maximum depth”, whereas the plotted variables appear to be height, velocity and pressure. Figure 14 - The caption apparently describes both low- and high-pressure areas as blue. Figures 18–20 - These figures give a visual impression of quantitative precision that exceeds the confidence supported by the input data.
The captions should explicitly state: The mapped values represent modelled hazard-indication extent rather than probabilistic avalanche risk. Figure 20 should not imply an effective positional precision of approximately 10 m.
13. Does the author give proper credit to previous work and clearly identify the original contribution?
Assessment: Generally yes, with some important exceptions.
The key methodological antecedents are properly acknowledged, especially:
- Bühler et al. (2013);
- Bühler (2018);
- Bühler et al. (2022);
- Christen et al. (2010);
- Sykes et al. (2022).
The manuscript makes clear that it is adapting an existing automated hazard-indication methodology, rather than presenting RAMMS or automated PRA delineation as entirely new. However, the authors should more explicitly distinguish between:
Existing methodology
- RAMMS dynamics modelling;
- PRA delineation;
- Swiss-style hazard-indication workflows.
Original contribution
- adaptation to Southeast Alaska;
- alternative snowfall forcing;
- treatment of Alaska land-cover constraints;
- climate-change application;
- regional computation and resulting dataset.
This distinction would strengthen the novelty statement.
14. Are the number and quality of references appropriate?
Assessment: Mostly yes, but the bibliography should be strengthened in several areas.
The principal avalanche-modelling literature is appropriately represented. However, additional literature is needed on:
- uncertainty and sensitivity analysis in avalanche modelling;
- extreme-value estimation of snowfall;
- return-period estimation;
- climate-model uncertainty propagation;
- forest ecology of Southeast Alaska;
- possibly more recent work on avalanche response to climate change.
The bibliography is therefore adequate in quantity but somewhat uneven in thematic coverage.
15. Are the references accessible by fellow scientists?
Assessment: Mostly yes.
Most journal references have DOIs and are readily accessible. Several important supporting sources are technical reports, including:
- TetraTech (2022);
- WSL Institute (2011);
- Fesler and Fredson (2000);
- Mears et al. (1992).
These may be less readily accessible internationally. Where such grey literature is essential to validation, the authors should provide persistent public links or archived versions whenever possible. The data reference to Wolken et al. (2026) is particularly important and should be verified to ensure long-term public accessibility.
16. Is the overall presentation well structured, clear and understandable?
Assessment: The broad structure is sound, but it needs substantial editorial improvement.
The overall progression is logical: Introduction → data/methods → validation → regional results → limitations → conclusions.
However, the manuscript would benefit from clearer separation between:
- Results;
- Validation;
- Discussion;
At present, interpretation and results are sometimes intermixed.
The manuscript also contains a non-trivial number of typographical and grammatical errors.
Examples include:
- “Thane Red” instead of Thane Road;
- “estimaation”;
- “within a call” instead of within a cell;
- “interprested”;
- “comptuer”;
- inconsistent use of NLCD/NCLD;
- “deliminate” instead of delineate.
A comprehensive English-language and technical copy-edit is necessary.
17. Is the length of the paper adequate?
Assessment: Generally adequate.
At approximately 34 pages including references and numerous figures, the manuscript is not excessively long given:
- the large study area;
- methodological complexity;
- extensive validation examples;
- climate-change component.
It is not recommended to shorten it.
However, some descriptive passages concerning local examples could potentially be condensed if additional space is required for the more important methodological additions requested in this review, particularly:
- sensitivity analysis;
- uncertainty quantification;
- quantitative validation;
- recurrence-interval clarification.
Thus, the overall length is appropriate, but the content balance could be improved.
18. Are there parts that need clarification, reduction, addition, combination or elimination?
Assessment: Yes — several substantial modifications are recommended.
Clarify
The following require clarification:
- derivation of the 30- and 300-year scenarios;
- relationship between PRA size and recurrence interval;
- meaning of “maximum possible three-day snowfall”;
- statistical basis of the snowfall lapse rate;
- treatment of Mixed Forest;
- rationale for the 300 m elevation threshold;
- rationale for the 30% land-cover threshold;
- interpretation of the future climate scenario.
Add
The manuscript would benefit substantially from:
A sensitivity-analysis section
At minimum, testing: d0 x 0.8 d0 x 1.2
and alternative forest thresholds.
Quantitative validation
For example:
- intersection-over-union;
- false-positive area;
- false-negative area;
- runout-distance error;
- spatial overlap statistics.
Uncertainty discussion
Preferably with at least partial quantification.
Workflow schematic
A single figure showing the entire workflow would greatly improve clarity.
Reduce or modify
Reduce use of the term risk. Remove or substantially revise interpretations implying metre-scale predictive precision. The expression “maximum possible snowfall” should be removed.
19. Is the technical language precise and understandable by fellow scientists?
Assessment: Mostly understandable, but not consistently precise.
The technical language is generally accessible to avalanche and natural-hazard scientists. However, several terms require more rigorous usage. “Hazard”, “susceptibility”, and especially “risk” are used too interchangeably. This is one of the most important conceptual issues, although it should also be relatively straightforward to correct.
The modelling framework essentially estimates: release susceptibility + modelled avalanche propagation/intensity under prescribed scenarios. It does not, in a strict sense, estimate risk because it does not systematically incorporate exposure, vulnerability, and expected loss. Even hazard in the probabilistic sense is only partially represented, because it does not estimate event frequency using a robust probabilistic framework. Nevertheless, Section 3.2 refers to areas as being “currently susceptible to avalanche hazards” and then repeatedly shifts to the term “avalanche risk”. The same issue arises in Figure 19, where changes in the proportion of land affected by simulated avalanche extents are described explicitly as changes in “avalanche risk”. This is more than a terminological issue. A change in the area reached by simulated avalanches does not in itself constitute a change in risk.
Most importantly:
Risk - should not be used as a synonym for avalanche-affected area. Hazard - should be distinguished from hazard indication and probabilistic hazard. Return period - should only be used where the corresponding probability is adequately defined. Resolution - The distinction between:
- raster grid spacing;
- effective source-data resolution;
- predictive positional accuracy;
must be made explicit.
Maximum snowfall - “Maximum possible” should be replaced by “maximum simulated during the considered 30-year period”.
Technical precision would improve considerably if these terms were revised systematically. The authors should use terminology such as: avalanche hazard indication, modelled avalanche-affected area, potential hazard extent, or avalanche susceptibility. Moreover, reserve risk for analyses that explicitly incorporate exposure and vulnerability.
The term “buildable zone” is also overly normative. Elevation below 40 m does not necessarily imply that land is suitable for development. It would be better to use “low-elevation zone or development-relevant lowland zone instead.
20. Is the English language of good quality, fluent, simple and easy to understand?
Assessment: Understandable, but not yet publication-ready.
The manuscript is generally understandable, but the English requires systematic editing.
There are:
- typographical errors;
- grammatical inconsistencies;
- awkward expressions;
- duplicated words;
- inconsistent terminology;
- caption errors;
- occasional informal wording.
Examples noted include:
- “our simulations finds”;
- “estimaation”;
- “interprested”;
- “comptuer”;
- “light greet”;
- inconsistent singular/plural agreement;
- several malformed date ranges.
A professional or careful native-level language edit is recommended before acceptance.
This issue alone would not warrant Major Revision, but it should be addressed alongside the scientific revisions.
21. Is the amount and quality of supplementary material appropriate?
Assessment: Currently insufficient for a study of this computational complexity.
The manuscript does not appear to provide supplementary material commensurate with the workflow’s complexity. For a study involving more than 3.5 million simulations and a multi-stage preprocessing pipeline, supplementary material should ideally include:
- complete RAMMS parameter tables;
- PRA thresholds;
- land-cover decision rules;
- release-depth settings;
- numerical workflow;
- model versions;
- representative configuration files;
- sensitivity analyses;
- validation metrics;
- additional maps;
- possibly tabulated regional statistics;
- exact climate-processing methodology.
The current Code and Data Availability statements are useful but do not fully replace detailed supplementary methodological documentation.
It is therefore recommended to provide a substantial Supplementary Information document.
22. Principal scientific issues requiring Major Revision
Based on the 21 criteria above, the following issues are essential before acceptance.
Recurrence intervals and extreme-event methodology
The paper labels scenarios as corresponding to 30- and 300-year recurrence intervals, yet the forcing relies substantially on maxima derived from 30-year climate periods. The authors themselves recognise that such values may underestimate 100- and 300-year extremes and that Extreme Value Theory would provide a more rigorous framework.
The relationship between
Moreover, must therefore be explained rigorously.At minimum, the authors should distinguish from because the recurrence interval associated with release conditions is not necessarily equivalent to the recurrence interval of avalanche impact at a given point.
If probabilistic recurrence intervals cannot be robustly estimated, the manuscript should use more qualitative terminology, such as frequent-type scenario and rare-type scenario, rather than implying precise 30- and 300-year probabilities.
Quantification of uncertainty
Uncertainty enters at every stage: . However, almost all simulations are deterministic.
The paper should include at least a representative sensitivity analysis. It is strongly recommended to vary:
Release depth: Forest threshold: For example Snowfall lapse rate: At least a plausible lower and upper alternative. RAMMS friction parameters: If computationally feasible, test representative parameter ranges on selected avalanche paths.
The purpose need not be a complete formal uncertainty propagation across all 3.5 million simulations. A representative sensitivity experiment would already substantially strengthen the manuscript.
Validation should be made more quantitative.
The current validation is useful but predominantly qualitative. The authors compare their outputs with previous assessments in Juneau and Valdez. Where digital polygons are available, it is recommended to calculate:
Iou = (Amodel ∩ Areference) / (Amodel ∪ Areference)
as well as:
Hit rate
False positive rate
False Negative Rate
and, where possible:
ΔLrunout
between modelled and reference runout.
Even a limited quantitative assessment for Juneau would represent an important methodological improvement. It is also recommended to rename the section: “Evaluation against existing local hazard assessments” unless more rigorous quantitative validation is added.
Correct distinction between hazard, hazard indication and risk
This terminology should be revised throughout the manuscript. A useful conceptual hierarchy would be:
- where avalanches may initiate or propagate;
- probability and intensity of avalanche occurrence;
- people/assets within affected areas;
- expected degree of damage;
and therefore:
RIsk = ƒ(Hazard, Exposure, Vulnerability)
The current study does not systematically quantify the latter two components.
Consequently, most uses of risk should be replaced by:
- hazard-indication area;
- avalanche-affected area;
- avalanche susceptibility;
- potential hazard extent.
Interpretation of 4 km → 10 m downscaling.
This issue should be explained particularly carefully. The snowfall field is represented at 10 m after interpolation and terrain correction, but its fundamental meteorological information originates from kilometre-scale atmospheric modelling. Therefore:
10 m grid spacing ≠ 10 m meteorological resolution
and certainly:
10 m grid spacing ≠ 10 m predictive accuracy
The interpretation in Figure 20 of a 10% grid-cell change as approximately 10 m of runout retreat should therefore be removed or strongly qualified.
Climate-scenario uncertainty
The projected changes represent one climate-model/scenario combination. The text should therefore avoid statements such as: “avalanche hazard will decrease” and instead use: “the CCSM4–RCP8.5 simulation used in this study indicates a decrease in modelled avalanche hazard extent…”
This modification is particularly important in the Abstract and Conclusions.
Vegetation treatment
The land-cover model is one of the largest potential sources of spatial uncertainty. The authors acknowledge misclassification problems in NLCD. Moreover, the 30 m NLCD data are resampled to 10 m, increasing raster sampling density but not the underlying land-cover information.
The sensitivity of hazard-indication results to:
- Mixed Forest treatment;
- Evergreen classification;
- the 30% threshold;
- the 300 m threshold;
should therefore be tested.
Release-depth treatment
The simplified release-depth formulation is understandable given data scarcity, but its impact on avalanche propagation must be demonstrated. Because avalanche volume scales approximately with release depth,
V ∝ APRA d0
uncertainty in d0 propagates directly into dynamic-model forcing. Release-depth sensitivity should therefore be treated as a core methodological issue rather than merely a limitation mentioned in the discussion.
Citation and bibliography assessment
The bibliography is generally scientifically sound, but it requires several corrections.
Védrine et al. (2022) - The reference itself is appropriate for avalanche–forest interaction. However, it is used to support the statement that Western Hemlock and Sitka Spruce dominate Southeast Alaska evergreen forests. That citation does not appear appropriate for regional forest composition. Replace it with a regional forest-ecology reference.
Sykes et al. (2022) - This is highly relevant to:
- PRA mapping;
- remote terrain;
- forested environments;
- data scarcity;
It should play a considerably larger role in the Introduction and Methods.
White et al. (2024) - The reference is legitimate, but because it is primarily ecological, it should not be relied upon as the principal methodological justification for three-day snowfall forcing. A more directly methodological snow-avalanche reference would be preferable.
Additional work on uncertainty and sensitivity analysis in avalanche-dynamics modelling would be highly relevant, particularly given the very large number of deterministic simulations and the the modelling chain which propagates uncertainty through: climate model → downscaling → snowfall → PRA → release depth → RAMMS → hazard extent, without formally quantifying that propagation.
The manuscript would benefit from more specialised literature on extreme snowfall statistics and return levels, beyond the general reference to Naveau et al. (2005) which is insufficient on its own for a manuscript whose interpretation relies strongly on recurrence intervals. This is particularly important because the study explicitly uses 30- and 300-year scenarios.
Required editorial corrections
Several specific corrections should be made before publication:
- Correct the date ranges in Figure 8. The caption appears to state “Historical (1981–1990) and Future (2051–2060)”, whereas the panels and the rest of the manuscript use 1981–2010 and 2031–2060. This is almost certainly an error and should be corrected.
- Correct the snow depth units of m² to m where appropriate. Es. on p. 13, snow depths appear to be reported as “0.7 m² and 1.2 m²”. The unit should presumably be m, not m².
- Figure 11 caption should correctly identify the plotted variables. The caption of Figure 11 states “Maximum height, maximum pressure and maximum depth”, whereas the axes show height, velocity, and pressure.
- Correct the colour description in Figure 14. In the Figure 14 caption, “Blue areas indicate impact pressure <30 kPa and blue areas indicate pressure >30 kPa”. One of these colours is evidently incorrect.
- “light greet” → “light green”;
- “Thane Red” → “Thane Road”;
- “our simulations finds” should be our simulations find.
- “estimaation” → “estimation”;
- “within a call” → “within a cell”;
- “interprested” → “interpreted”;
- “comptuer” → “computer”;
- “ArCGIS” → “ArcGIS”;
- NCLD/NLCD should be standardised;
- “deliminate” → “delineate”;
- The expression “maximum 3-day snowfall, including settling” requires clarification, because snowfall and settled snow accumulation/depth are not equivalent terms.
- The manuscript should use “historical”, “historic”, “recent historic”, and “contemporary” more consistently.
- Date ranges should be consistently formatted as 1981–2010 and 2031–2060 avoiding corrupted forms such as “2031-±2060”;
- Van Vuuren et al. DOI formatting should be corrected;
- The spelling and diacritics in “Bühler and Kuhne (2018)” should be checked and standardised to Bühler and Kühne, where appropriate.
23. Summary evaluation table
1. Relevance to NHESS
Yes / Strong
2. Novelty
Yes / Mainly applied novelty
3. International standard
Partly — revision required
4. Methods and assumptions
Partly — major clarification required
5. Results support conclusions
Partly
6. Substantial conclusions
Yes, but some overstatement
7. Reproducibility
Incomplete
8. Title
Adequate
9. Abstract
Good, but requires qualification
10. Accessibility of title/abstract
Good
11. Formulae, symbols, units
Mostly adequate; several corrections required
12. Figures
Generally good; some important corrections
13. Credit to previous work
Generally appropriate
14. References
Adequate but should be expanded selectively
15. Reference accessibility
Mostly adequate
16. Structure and clarity
Generally sound; editing required
17. Length
Appropriate
18. Need for additions/clarification
Substantial
19. Technical language
Generally good but conceptually inconsistent in places
20. English
Understandable but requires editing
21. Supplementary material
Insufficient for full reproducibility
24. Final recommendation: MAJOR REVISION NEEDED
The manuscript addresses a highly relevant NHESS topic. It provides an original, large-scale application of automated avalanche hazard-indication modelling in a region where systematic avalanche hazard information is scarce. The principal scientific contribution is significant enough to justify further consideration for publication. However, the manuscript is not ready for acceptance because the current level of quantitative interpretation exceeds the robustness demonstrated by the methods.
The essential revisions are:
- clarify or reformulate the 30- and 300-year recurrence scenarios;
- introduce a representative uncertainty/sensitivity analysis;
- strengthen validation quantitatively where possible;
- distinguish consistently between susceptibility, hazard indication, hazard and risk;
- moderate claims of fine-scale spatial precision;
- qualify future projections as results of the specific CCSM4–RCP8.5 realisation;
- strengthen the treatment and uncertainty analysis of vegetation;
- test sensitivity to release-depth assumptions;
- improve reproducibility documentation and supplementary material;
- perform comprehensive technical and English-language editing.
Rejection of the manuscript is not recommended, because the study has clear scientific value, considerable computational investment, and strong potential as a regional-scale hazard-screening framework. At the same time, Minor Revision would not be sufficient, since several of the issues identified above require additional analysis rather than simple textual amendments. If the authors address these points satisfactorily, the revised manuscript could make a valuable contribution to the literature on large-scale avalanche hazard-indication mapping, climate-sensitive natural hazards, and hazard assessment in data-sparse mountain environments.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 621 | 279 | 52 | 952 | 50 | 69 |
- HTML: 621
- PDF: 279
- XML: 52
- Total: 952
- BibTeX: 50
- EndNote: 69
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General Comments
In this manuscript, the authors present hazard indication maps for southeast Alaska, United States. They use historical (CFSR) and projected snow depth data (NCAR CCSM v4) combined with downscaled 4km WRF projected climate data for estimates of maximum 3-day snowfall for historical and the mid-century period. They then used these data to conduct millions of simulations using RAMMS with specific forest and land cover inputs over a large domain. They provide results in several populated areas and compare with previous runout assessments and local knowledge.
The manuscript is well organized, written, and has a logical order. I am, however, a bit confused by the Results section and associated reference to Results. The authors reference Wolken et al., (2026), but the doi does not work. I understand not making data available to the public prior to publication. However, the Data Availability section references this citation, and reviewers should have access to these results.
I think the Discussion (included in Results) could be bolstered by comparisons with other avalanche-climate studies throughout North American and the rest of the world. This would provide some context to the projected results.
Input and validation data are scarce or non-existent and, as the authors acknowledge, leads to a substantial amount of uncertainty. I think the authors treat this accordingly for the most part and do not overstate their results. However, I think a more quantitative approach to comparing this study’s output to the previous studies (i.e. runout distance or extent differences) would be beneficial. This would provide some context to the influence of uncertainty in this study especially given these previous studies are the main source of validation data. Additionally, the authors mention local avalanche knowledge (Lines 49 and 254 and Figure 12 caption) as methods of validation. What does this entail? Are actual observations used? More information regarding this type of validation and the period of record knowledge of these observations would be useful for the reader.
Overall, this is a useful approach for a data sparse area, but with substantial limitations, uncertainty, and assumptions. The manuscript is suitable for the journal, but I would strongly suggest that the authors provide a statement, at least in the Conclusions, about using these hazard indication maps with caution for operational or planning purposes given the uncertainty and lack of in-situ validation data.
Specific and Technical Comments
Line 59/Section 2: There are a lot of data and processing steps in this study. I think this manuscript would benefit greatly from a figure explaining the workflow that includes what datasets are used at each step and how each informs the other and perhaps the most important point of uncertainty with each dataset. This would also help with understanding uncertainty propagation through the workflow.
Line 82: How do you account for the scenario when only part of the PRA releases in a medium or large PRA? I assume that the major change in simulation is the release depth, but how do you account for partial propagation across the PRA? Or are you simply treating full PRA release in every simulation? This probably deserves some explanation.
Line 85 and onward: First abbreviated as NLCD, but then NCLD is often used.
Lines 167-168: This is a bit confusing as written. I assume the 30-year and 300-year return periods for frequent and rare are based on 3-day maximum snowfall occurrence and that the runout is simply a function of that rather than a more sophisticated 30-year and 300-year avalanche probability, correct? Consider clarifying.
Line 187: snow depths are given in m2. Should be m.
Line 210: Why did you choose 1000 m as a buffer?
Line 228 (Sec. 2.10): Given the coastal avalanche climate, wet snow avalanches (both winter and spring) are common in southeast Alaska. How did you account for a difference between wet and dry snow avalanches or mixed flow regimes (i.e. those that start dry but entrain wet avalanche debris as they run downslope) and how this would impact runout distance?
Line 361: Suggest including a sentence about what Sykes et al. (2022) did that would improve forest cover related runout and PRA identification in your study.
Line 385: “…inherent limitations of climate modeling…” and also forest cover inaccuracies?
Lines 545 and 559: This doi does not exist. See General Comment above. Wolken, G., Fischer, E., Hendricks, M., and Wikstrom, J.: Alaska Snow Avalanche Hazard Database, Tech. rep., Alaska Division of Geological & Geophysical Surveys, https://doi.org/10.14509/32076, 2026.
Figure 10: This figure is not referenced in the text. Additionally, please include a legend that identifies what the dark green and light green represent.
Figure 11: Please add location dots for the town centers and N arrows or grid tick marks on the exterior as in other figures.
Figure 13 right: It is difficult to see the blue outlines in the dark green areas.
Figure 15: ‘light greet’ to green.
Figure 19: caption “…show that expected precipitation [phase] changes are the...”
Figure 20: The resolution is quite poor in this figure.