the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Mar 2025
Interannual variability in air temperature and snow drive differences in ice formation and growth
Abstract. Recent warming of northern, high-latitude regions has raised alarms for the safe and efficient use of frozen lakes for winter transportation and recreation. This concern is significant in Canada’s Northwest Territories (NWT), where seasonally constructed roads over lakes, rivers, and land (winter roads) span thousands of kilometers and act as vital links to isolated communities and resource development projects. Current climate change and weather variability is altering the evolution of lake ice, challenging predictions of freeze-up, ice growth, and ice decay. The accurate simulation of ice evolution is imperative for safe and efficient planning, operation, and maintenance of winter roads under a changing climate and heightened weather variability. This is particularly significant in the early winter period when ice road planning and design is undertaken. Here, we investigate the effects of weather variability on ice formation, growth, and evolution in a small lake near Yellowknife, NWT, Canada. High-resolution measurements of air, snow, ice, and water temperatures were collected continuously from a floating research station between October and December in 2021, 2022, and 2023 and variability in ice evolution and weather examined. Combinations of above and below average snowfall and winter air temperatures resulted in variability of up to 17 days in freeze-up dates (FUD) and 8 days in freeze-up durations. End of December ice thicknesses (hi) varied up to 12 cm, while the duration between the FUD and hi=30 cm varied up to 10 days. hi were effectively simulated (RMSE=1.11–2.33 cm) using empirical relationships developed using cumulative freezing degree days (CFDD) and seasonally cumulative snowfall (ST), while snow-ice thicknesses simulated (RMSE=0.83–1.21 cm) using CFDD and daily snowfall. Developed relationships between air temperatures, snow, and ice thicknesses can be used for predicting minimum ice thicknesses required for commencing ice road construction, and to assist in the effective management of construction activities.
Competing interests: One of the authors is a member of the editorial board of The Cryosphere.
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.- Preprint
(1928 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 23 Apr 2025)
-
RC1: 'Comment on egusphere-2025-975', Anonymous Referee #1, 04 Apr 2025
reply
The manuscript presents an empirical model for lake ice formation and growth based on three-year field observations at Landing Lake, Canada. While the methodology demonstrates potential for winter road management and climate change monitoring, several critical issues require clarification to strengthen scientific rigor and practical applicability. Specific recommendations are organized as follows:
Specific comments:
- Line 61: Correct "Xinjing" to "Xinjiang"
- Lines 39-71: Condense discussions on ice phenology studies.
- Lines 72-77: Expand on the disadvantages of conventional techniques (manual observations and numerical modeling) compared to the FRS system to emphasize research significance. (1) Manual measurements: Labor-intensive with discontinuous temporal coverage. (2) Numerical models: Computationally demanding.
- The study focuses on freeze-up, ice-onset, and ice growth. However, the capability of the FRS system to monitor the complete ice thickness cycle (including break-up and melt processes) remains unclear.
- While the manuscript emphasizes the importance of ice simulation for winter road management under climate change, the empirical model is derived from a small lake (1.07 km2). Address whether such site-specific relationships can be generalized to larger water bodies or regions with distinct climatic/hydrological conditions.
- Lines 96-101: Add references to support statements about meteorological data requirements.
- Section 3.1.1: Clarify MODIS data usage: (1) Specify product version (MOD11A1/MYD11A1?). (2) Justify spatial representativeness: How were pixel quality issues addressed for a 1.07 km² lake under 1 km resolution? (3) Indicate whether day or night data were used.
- Line 167: State the distance between Yellowknife Airport station and Landing Lake.
- Lines 278-288: While agreeing with the 2021 ice-onset (IO) and freeze-up date (FUD) determinations, we recommend utilizing Sentinel-2 imagery for independent validation.
- Figure 5a: (1) Explain discontinuous Ts curves: Were data gaps caused by cloud masking or quality filtering? (2) Replace connected lines with discrete markers (e.g., circles) for non-continuous MODIS data.
- Definition inconsistency: Ice-onset (IO) and freeze-up dates (FUD) are defined as horizontal lake-wide phenomena (Line 18), yet 2022–2023 determinations rely on vertical SIMBA temperature profiles.
- Figure 6: Include time-series plots of SIMBA-recorded vertical thermal profiles to illustrate water column stratification dynamics.
- Table 4: Small lakes exhibit low thermal inertia, leading to rapid air temperature responses (11- and 3-day freeze-up durations in 2022–2023). However, the stable water temperature (Tw) in 2022 contradicts this pattern. Analyze potential causes: (1) Assess vertical stratification using mixed-layer depth calculations. (2) Evaluate whether the lake remained fully mixed.
- Figure 7a: Explain the abrupt snow depth reduction on 7 November 2023 (25 cm → < 10 cm). Was this due to melting, compaction, or sensor artifacts?
- Equations 5a–5b: Replace ambiguous coefficient symbols (e.g., use β, γ instead of α, a) to avoid confusion.
- Line 380: Strengthen analysis by presenting SIMBA thermal profile time series
- Figure 8: Provide model results across BT = 0 to -10°C (not 0 to -5°C) to justify selecting BT = -5°C as optimal. Include sensitivity analysis of BT variations.
- Lines 387 vs. 395: Conflicting descriptions of BT experimental ranges ("0 to -5°C" vs. "0 to -10°C").
- Lines 379–399: Reorganize logic: Step 1: Present BT sensitivity experiments. Step 2: Identify optimal BT (-5°C). Step 3: Report corresponding α, a, b, c Step 4: Show final model performance (Figure 8).
- Lines 411–417: Improve readability by integrating Table A1 into the main text.
- Lines 399–417: Relocate to the Discussion section to critically evaluate: (1) Model applicability across lake types. (2) Limitations in parameter transferability.
- Lines 461-463: The statement is debatable, as freeboard can be estimated using Archimedes’ principle.
- While the authors aim to develop a simplified empirical model for ice thickness estimation, the interannual variability of coefficients (α, a, b, c) necessitates field-based calibration, severely limiting practical utility. To strengthen conclusions, I recommend: (1) Comparative studies across lakes to establish parameter ranges. (2) Explicit guidance on minimum data requirements (e.g., duration and type of meteorological/hydrological inputs) for reliable model application in the future.
- Section 8 (Lines 488–524): Restructure content: (1) Relocate technical discussions (e.g., model assumptions) to the Discussion section. (2) Retain application scenarios and future research directions in Conclusions.
Citation: https://doi.org/10.5194/egusphere-2025-975-RC1 -
RC2: 'Comment on egusphere-2025-975', Anonymous Referee #2, 06 Apr 2025
reply
Review on “Interannual variability in air temperature and snow drive differences in ice formation and growth” by Arash Rafat and Homa Kheyrollah Pour
Climate change involves many complex processes. For the cryosphere, the timing of the events is a critical concept. The freezing and melting of lake/sea ice alter the energy balance between the atmosphere and the underlying water bodies (lakes or oceans), thereby influencing climate dynamics. One of the most critical practical concerns is that freeze-up timing directly affects the usability of ice roads. This is especially vital for North American Arctic communities, where ice roads serve as lifelines for remote regions in Alaska and Canada.
This manuscript investigates ice formation and growth in a small boreal lake in Canada’s Northwest Territories (NWT). The authors conducted in-situ observations over three consecutive winter seasons on a single lake. The dataset includes local meteorological parameters such as wind speed, air temperature, and turbulent and radiative heat fluxes. A platform was installed on the lake to collect high-resolution snow and ice temperature measurements using a novel, cost-effective automated device (SIMBA).
These observations, combined with long-term meteorological data from weather stations, were used in a statistical model to calculate ice thickness employing an exponential function of snowfall as input.
The manuscript investigates local variability in climate and weather, particularly ice formation and growth, with a focus on ice freeze-up dates (FUD) and the evolution of snow and ice cover. The authors argue that the derived relationships between air temperature, snow depth, and ice thickness can be used to predict the minimum ice thickness required for ice road construction, aiding in the effective management of construction activities.
The topic of this manuscript is highly relevant to the scope of TC. The observations were made without flaws, and the configuration of the SIMBA platform is solid and well-justified. The statistical model is conventional yet robust, and the data analysis is convincing. However, I have some concerns and comments regarding certain aspects of the content, which I hope the authors will address through a proper revision before the manuscript's final acceptance.
Major comments
1 The manuscript’s overall structure could be improved for better clarity. a) I don’t see a clear chapter on the “Results”. The presentation of data and results was somehow mixed. I suggest restructuring the entire manuscript. For example, a chapter entitled “results” that contains partial Chapter 4 and Chapters 5 and 6 may yield better clarity of the manuscript; b) I am not sure why Chapter 8 is needed, especially after the conclusions have been made. I suggest this chapter can be placed before the conclusion, e.g., Discussions.
2 Based on the study's objective, as stated in the abstract and final chapter of this manuscript, the discovery of robust relationships between air temperatures, snow cover, and ice thickness is intended to assess the feasibility of ice road construction and support effective construction management, which I agree. However, this work has been carried out in a tiny lake (1.1 km2). The questions I want to ask: a) How representative are the results of this work? b) Can those derived formulae be applied to obtain FUSs in other parts of the NWT? c) Would it be possible to assess the performance of the formula for the other small lakes in NWT? d) At least a discussion of the general applicability of the formula should be included in this study.
3 I have problems understanding the presentation of figures and tables. Many captions are currently insufficient for readers to easily grasp the key information. I recommend revising them accordingly (see my detailed comments below).
4 Authors investigated several snow parameters: date of the first snowfall (𝑆𝑂𝑁), the cumulative snowfall (𝑆𝑇), the peak hourly snowfall rate in a given day in each month (𝑆p ), and the number of snowfall days (𝑆𝑑 ). Please explain a bit more about Sp. Based on the definition, I understand the other parameters are one number for each winter season. However, Sp has multiple numbers for each winter, right? The snow measurement was made in the “Yellowknife Airport weather station (1942-2023)”. Please write more information about snow observations, e.g., instrumentation, data quality, and possible errors. If there are good snowfall data, I encourage authors to apply an analytical model to calculate the ice thickness and snow-ice, taking into account the effect of snow. See Lepparanta (1993) for a good example of such an analytical model.
Leppäranta M. 1993. A review of analytical models of sea-ice growth. Atmosphere-Ocean, 31(1): 123-138, doi:10.1080/07055900.1993. 9649465.
Detailed comments:
5 Please add a Canada map as a background for Figure 1. I think Photo A can be dropped since Figure 2 shows the details of the SIMBA floating station.
6 “Photographs 1, 2, and 3 were taken on October 23, 2023, 105 at 09:00 local time.” I don’t see any close text to explain those photos. I found at L290, a description “culminating in a FUD of October 23, 2022, 3 days”, I would assume this was the explanation of those photos. If so, maybe write: ,,,at 09:00 local time (see explanation in 5.1), and correct the typo 2022 to 2023. Otherwise, please add text somewhere to explain those photos.
7 Could you edit photo 3 in figure 1 to show a horizontal coastal line?
8 It seems to me that the PAR and pressure transducers' data on FRS, as well as turbulent and radiative heat fluxes measurement at the weather station on land nearby, are not used in this study. Please include a brief description of the purpose of these data.
9 Figure 2. Please add “lake water surface or ice surface close to the black inverted triangle symbol.
10 Section 3.3 Heat storage: please explain the physical meaning of negative heat storage in this section. Such values were calculated around L315.
11 Figure 3. Please explain the symbol “x”. It is hard to see the yellow color of the triangle.
12 L219: “The same year saw colder than normal conditions by the end of December with 𝐶𝐹𝐷𝐷 being 123% of normal”. Please add the number before 123%. Also for those >100% in the following text until L230 if possible.
13 Table 1. Please explain what those numbers with parentheses (4.1, 10.5). In the Table, I see (Tmin - Tmax),*mean (min, max) monthly cumulative snowfall between 1981-2010 and 1992-2021.
14 I would remote “-“ on the 3rd line. Please explain a bit more Sp, see my major comment 4.
15 Figure 5. Please explain how the surface temperature was measured by SIMBA. Could you add air temperature measurements from the nearby weather station on land for comparison?
16 Table 5. What is “X”? maybe N/A is better?
17 Figure 8. Please explain how those dots (blue, black and red) were obtained.
18 Figure 11. Please explain how the observed snow-ice thicknesses were made?
19 L406: CFDD should be CFDD.
20 L79: “inter- and intra- annual variability”, maybe good to write “inter-annual and seasonal variability”.
21 “Code and data availability”. I am not sure whether the statement the authors made is acceptable to the TC. The data link (https://climate.weather.gc.ca/) is the main page of ECCC. I think the authors should provide the data link that can direct access the air temperatures and snowfall measurements at the Yellowknife Airport weather station between 1942-2023. The lake measurement data sets (weather station, SIMBA) were missing and should be publicly accessible.
Citation: https://doi.org/10.5194/egusphere-2025-975-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 64 | 32 | 7 | 103 | 5 | 4 |
- HTML: 64
- PDF: 32
- XML: 7
- Total: 103
- BibTeX: 5
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 36 | 36 |
| Canada | 2 | 17 | 17 |
| Brazil | 3 | 14 | 14 |
| China | 4 | 6 | 6 |
| France | 5 | 5 | 5 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 36