the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Larger variability of winter snow depth promotes the soil thermal regime instability over boreal high latitudes
Abstract. Permafrost degradation in Siberia is assessed as one of the global climate tipping elements. Soil thermal regime instability (STRI) describes the amplitude and frequency of soil temperature extremes at the interannual scale, and such thermal instability can propagate downward through the soil column and affect permafrost thermal conditions, making STRI a useful indicator of permafrost vulnerability. Yet, how STRI changes and the underlying mechanism remain poorly understood. Here, regarding soil temperature variability as a proxy of STRI on interannual scale, and through combining a merged soil temperature dataset with in-situ observations, it was detected that the subsurface (40 and 160 cm) STRI over Siberian continuous permafrost region is higher compared to other regions over boreal high latitudes, most pronounced in winter. Moreover, winter STRI across the Siberia continuous permafrost region has increased suddenly since early 1990s. Increase in STRI is particularly pronounced in regions where soil thermal regime was once stable, with increase rate of STRI exceeding 0.5 °C/10a in some regions. Statistical analyses and numerical experiments reveal that this increased STRI is strongly associated with the larger interannual variability of winter snow depth through its insulation effect, with snow depth accounting for more than 50 % of the STRI increase in most regions, and influence of snow depth on STRI is detectable down to 3.6 m. These findings highlight that snow-soil interactions dominate winter soil thermal dynamics, particularly over the thick snow regions, with potential implications for permafrost stability, carbon dynamics, and ecosystem responses.
- Preprint
(1837 KB) - Metadata XML
-
Supplement
(1152 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
CC1: 'Comment on egusphere-2025-5676', Geoffrey Henebry, 20 Feb 2026
-
CC2: 'Reply on CC1', Chenghai Wang, 21 Feb 2026
Thank you for your comments and suggestions. Our main concern is focused on the high-latitude regions. However, you have provided excellent suggestions. We will take your suggestions into the revised version and cites the relevant references.
Citation: https://doi.org/10.5194/egusphere-2025-5676-CC2
-
CC2: 'Reply on CC1', Chenghai Wang, 21 Feb 2026
-
RC1: 'Comment on egusphere-2025-5676', Anonymous Referee #1, 30 Apr 2026
The manuscript of Lingyun Ai et al. is interesting as it tackles the impact of winter snow depth and snow depth variability on soil temperature and soil temperature variability in permafrost regions using observation, reanalyses, and model sensitivity studies. It is clearly written and illustrated, but it lacks some explanation and discussion like:
- Why using “Soil Thermal Regime Instability” (STRI) all the time and confusing it with its “proxy” soil temperature variability? The latter has a clear definition, but the former is not defined in the draft. STRI could be deleted from the draft and it would be much more straightforward.
- Can you really trust in 160 cm soil temperature provided by the reanalyses? For example, in ERA5 the soil is not deep enough to provide trustworthy values.
- In Figs. 2 & 4 the strong T_soil increase and T_soil and d_snow SD decrease before the 1980s are striking, but are not discussed in the text. But, these are important as their trends make the SD increase of T_soil later on difficult to understand. The simple explanation that with larger T_soil its variability increases or that the impact of the extremes accumulate are somehow contradicted by the pre-1980 trends.
- In Sec 4.1 it is discussed/concluded (I suggest to have a Conclusions section to not mix discussion and concl.) that upper-layer temperature perturbations can penetrate deep (3.58 m) and thus below the active layer (ca. 2.4 m). Might be, but this is based on limited model results and thus I suggest to clearly state that more observations and model studies are necessary.
- Line 441: “more frequent intense warm or cold anomalies”. As the draft considers the winter perturbations and the snow variability, is it snow insulation anomalies? So, more like more or less cooling anomalies in the upper soil which penetrates downward?
Line 140: I guess ambiguous is meant
Line 160: How a fixed heat flux ensure a robust/sensible response in the surface soil? This is a formulation issue. Of course, it needs a reasonable bottom boundary condition.
Line 173: months → seasons
Citation: https://doi.org/10.5194/egusphere-2025-5676-RC1 -
RC2: 'Comment on egusphere-2025-5676', Anonymous Referee #2, 24 May 2026
This manuscript tried to reveal the influence of the changes of winter snow changes or its variability on soil thermal regime instability over boreal high latitudes based on mostly the statistically analysis of in-situ ground temperature observations and merged soil temperature dataset, experimental simulation by LSM. The topic is important in reveal the changes of thermal state in this study region, especially to the changes of permafrost there. The sow effects to permafrost were well addressed in many permafrost related textbooks and published papers. And it is really important to qualify these relationships and mechanics in models. But this manuscript was not. The conclusion of this manuscript raised the issue again which is the snow playing significant influences on permafrost changes. For the manuscript, my questions listed as:
- It is a common knowledge that all reanalysis datasets have great bias in permafrost regions. So how about the method and the accuracy of merged monthly average Tsoil dataset from seven reanalysis datasets after interpolated to 0.5° ×625° grid using linear interpolation? And how to ensure that the accuracy of merged monthly average Tsoil is smaller sufficiently than the small temperature anomaly caused by abnormal snow cover and its changes?
- The authors used the Tsoil variability as a proxy of the STRI on interannual scale, and actually the standard deviation (SD) was used to represent the variability on interannual scale. Why? And why the monthly average values were used? It needs more evidences for all the questions raised.
- The study eliminated the impact of global warming by removing the linear trends for site observation, and detrended the data by a 9-year sliding standard deviation (SD; the detail given in Section 2.2 ). Did the influence of snow variability be bigger enough than the bias caused by averaging 9 years data?
- The authors removed the influence of autumn and winter Tair variability using multiple linear regression. Do these influences are linear?
- A fixed deep soil heat flux is imposed at the lower boundary to ensure stable thermal responses in the surface soil? How about the depth for the lower boundary?
Citation: https://doi.org/10.5194/egusphere-2025-5676-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 814 | 438 | 85 | 1,337 | 139 | 77 | 146 |
- HTML: 814
- PDF: 438
- XML: 85
- Total: 1,337
- Supplement: 139
- BibTeX: 77
- EndNote: 146
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Thank you for an interesting study. However, I think there is another source of uncertainty/limitation; namely, changes in snow seasonality, which have been observed in mountainous regions in northern Eurasia and globally, albeit at lower latitudes. Here is a selection of recent studies (in chronological order):
Li, Q., Yang, T., Zhou, H. and Li, L., 2019. Patterns in snow depth maximum and snow cover days during 1961–2015 period in the Tianshan Mountains, Central Asia. Atmospheric Research, 228, pp.14-22.
Notarnicola, C., 2022. Overall negative trends for snow cover extent and duration in global mountain regions over 1982–2020. Scientific Reports, 12(1), p.13731.
Tang, Z., Deng, G., Hu, G., Zhang, H., Pan, H. and Sang, G., 2022. Satellite observed spatiotemporal variability of snow cover and snow phenology over high mountain Asia from 2002 to 2021. Journal of Hydrology, 613, p.128438.
Henebry, G.M. and Tomaszewska, M.A., 2025. Snow-cover seasonality in Kyrgyzstan: variation and change over 20 years (2001–2021) as observed by the MODIS Terra snow product. Environmental Research Letters, 20(2), p.024018.