Review of the paper #tc-2023-148 entitled “Multiscale modeling of heat and mass transfer in dry snow: influence of the condensation coefficient and comparison with experiments” by Bouvet et al.

In the wake of a previous study of Calonne et al. (2014b), this paper aims presents a multiscale approach to follow heat and mass transfers in dry snow focusing on the peculiar role of the condensation coefficient α in order to mimic natural snow evolution during changes in the snow microstructure called temperature gradient metamorphism (TGM). Using a two-fold homogenization of a model coupling the heat conduction through ice and air, the water vapor diffusion in air and the sublimation of ice and deposition of vapor at the ice grain interface, this study’s interest mainly consists in the fact that the effect of the condensation coefficient α is poorly described in the literature. Thus, considering a large range for this parameter (from 10−10 to 1), different effective behaviors are obtained through the upscaling process according to a transition value α_T ≈ 3×10−4. Moreover, the homogenized modelling results were compared with three experimental tests of TGM of snow, providing a solid discussion.

In general, the manuscript is well-written and constructed, and the mathematical statement, the upscaling procedure and the experimental comparisons are well presented.

The quality of this paper is indubitable, and I only have minor concerns.

i/ The main articulation between this study and the previous results of the group on the same topic should be highlighted. Indeed, even if many new results are presented here, some interesting links can be made.

ii/ The consequences of the model’s assumptions (isotropic materials properties, no convective effect, no curvature effects, no natural convection at the pore scale, etc.) could be outlined. In particular, the use of the Hertz-Knudsen at the fluid-solid interphase is one of the key point of this model and requires to be justified in this context.

iii/ The evaluation of the dimensionless numbers defined by Eq. (19) is a key point of the upscaling procedure. Thus if the sensitivity to the coefficient α is well introduced, one may wonder if other similar dependencies of some dimensionless numbers (according to the temperature for instance) may not be discussed.

Notwithstanding these general remarks, this is a complete work coupling models, numerical simulation and experimental comparisons. It is rigorously presented and detailed in the appendix. That is why, if these minor suggestions are addressed, I suggest to accept the publication of this work.