| 20 May 2026
High-Resolution Modelling of Landfast Ice Formation and Midseason Breakout on the Siberian Shelf
Abstract. Substantial amounts of landfast ice form over the Siberia shelf seas during winter. Under certain conditions, midseason break out may occur, affecting local ecology and human activities. This study employed a 2 km resolution coupled ocean-ice model incorporating a basal stress parameterization to simulate landfast ice in the Siberian Seas. The high- resolution model yields realistic simulations of landfast ice distribution in the Siberian Seas. The contribution of the grounding mechanism to landfast ice stability varies regionally and depends strongly on local geography. Grounded fast ice accounts for 56% of the total simulated landfast ice on the Siberia shelf, with the East Siberia Sea having the largest proportion (70%), followed by the Laptev Sea (54%) and the Kara Sea (41%). The simulated midseason breakout events primarily occurred along the outer edge of the landfast ice in the East Siberia Sea, where the basal stress was ineffective. Further analysis reveals that all the midseason breakouts were associated with the passage of cyclones, whose offshore winds fractured the landfast ice into floating ice.
Review report on "High-Resolution Modelling of Landfast Ice Formation and Midseason Breakout on the Siberian Shelf" by Bai et al.,
This study investigates landfast ice formation and midseason breakout. The topic is highly relevant to the TC Journal. However, I have some doubts and questions for the authors. I cannot provide an adequate assessment of this manuscript without more information. My recommendation is to reconsider and resubmit after a major revision. I hope the authors will find my comments useful to improve this study
1 To my knowledge, the landfast ice is largely immovable; the major evolution of landfast ice is largely driven by thermodynamics. Yet I see no discussion of sea ice thermodynamics in this manuscript at all. This makes me wonder what the definition of landfast ice is in this study? May the modelling be more relevant to claim as "High-Resolution Modelling of coastal sea ice"?
2 Ice formation was mentioned in the title, and this was not adequately investigated in this manuscript. The authors need to offer strong arguments on why sea ice thermodynamics was not investigated in this study. I would rather encourage authors to add modelling of sea ice thermodynamics in the manuscript. To be more specific: a) How does the landfast sea ice grow and melt?
3 L69-70 “a chart identifies areas of landfast ice and their boundaries (Yu et al. 2014)” : What is this? Do authors use some sort of input data to define landfast ice area? Or the landfast ice extent was modelled?
4 L70-L72: Is there a particular reason to use Japanese 55-year Reanalysis? Do authors applied 6h time interval to drive the coupled ice-ocean model? JRA-55 data have a spatial resolution of 0.25° x 0.25°, while the coupled ice-ocean model has a resolution of about 2km. How to deal with the mismatch of spatial resolution between forcing and model configuration? A 6h time interval is also questionable as it will filter out the diurnal variations, in particular solar radiation, which played a crucial role in sea ice melting.
5 L102: “the initial sea ice conditions, including sea ice concentration and sea ice thickness, are from National Snow and Ice Data Center (NSIDC)” This is the initial condition, as far as I understood. While on L112: “The model was initialised from the nearly ice-free conditions of September 1, 2012.” Which one is actually used? The authors claimed, “We conducted two three-year simulations from September 1, 2012 to September 1, 2015”, which means each model run lasts 3 years. Is the landfast ice seasonal or interannual? If there is no landfast ice (e.g. Fig.4), how was the model run made during ice-free conditions?
6 Can authors provide an evolution of ice thickness time series?
7 How about snow conditions? Does snow play a role in landfast sea ice modelling?