Convective Heat Transfer of the Spring Meltwater Accelerates Active Layer Phase Change in Tibetan Permafrost Areas

By Zhao et al.,

In this study, the authors applied the SHAW model to investigate temperature dynamics at the Tanggula site at the Eastern Qinghai-Tibetan Plateau. The modeling experiment includes the cases: 1. control (fully coupled diffusion convection), 2. NoSurf (no convective heat flux between surface and subsurface, but preserves the convective flux in the subsurface as a results of water movement in soil pores), 3. NoConv (no convective heat flux from surface to subsurface as well as no flux due to water movement in the subsurface). If I understand correctly, in case 2, the second left term in eqn. 2 remains for the subsurface formulation and is completely removed from the eqn. 2 for case 3.

Overall, the study is interesting and well presented. I suggest including a discussion on a similar mathematical formulation described by Painter et al., (2016). How is the current mathematical formulation different from Painter et al. What are pros and cons of each formulation? Is q_l a Darcy flow or water head potential in the current formulation? My understanding was that the current model considers both saturated and unsaturated soil conditions. It would be nice to include the discussion on the compute time difference between saturated and unsaturated cases.

What is the equation for the vertical water flux? It is not clear how one would define subsurface water transport without addressing the soil permeability? I suggest including the permeability paragraph in the Discussion section. During snowmelt, the surface is typically frozen. That said, most if not all of the snowmelt water should runoff the surface. Not clear why it would percolate into the subsurface? Westermann et al., has a publication on the effect of water percolation in the snow layer and its impact on surface temperatures, worse to mention as well.

I found Figure 3 not necessarily helpful for understanding the effect of convective heat flow. Does it mean that convective flux adds to the 5C difference in soil temperatures over a short period of time?

Can the changes that are shown in Figs 3 and 4 be verified against observations or in lab measurements? L353 concludes that convective flux could contribute to up to 10C temperature change during snowmelt. This is massive warming or cooling, which I am having a hard to believe. Short-term 10C warming/cooling should lead to substantial changes in subsurface thermal state over time. Are there any pieces of evidence of the accelerated ground warming in the region? I suggest quantitively estimate the impact of these short-term warming/cooling pulses on mean annual ground temperatures. Are these pulses significant, and on what temporal scale? For example, on the yearly scale, those pulses could be less important than on a monthly scale and so on....

Taking into consideration the above comments and questions, the paper needs more clarification on the mathematical formulation, application, and discussion with similar studies.

Minor

L45-48 …influenced. Influenced by what? Not clear. I suggest to re-write.  

L121-124 need a reference.

In Table 1. Are any of the parameters been tuned during this study?

On Figure 2. It is hard to tell the time of snowmelt.

Additional language improvements are needed as well.

This manuscript applied the SHAW model to investigate the impacts of CHT on active layer thermal dynamics on the Tanggula station, a typical continuous permafrost site located at the eastern Qinghai-Tibetan Plateau with abundant meteorological and soil temperature/moisture observation data. The 2008-2009 observed hourly data were used to calibrate the model parameters and those of 2010 for validation. The control experiment was carried out to quantify the changes in active layer thermal regime affected by vertical advection of liquid water, consisting of three setups: using (1) the original SHAW model with full consideration of CHT; (2) a modified SHAW model ignoring the CHT due to infiltration from the surface, and (3) a modified SHAW model ignoring complete CHT processes in the system. The results show that the CHT events mainly happened during thawing periods when the active layer melted at shallow (0-0.2m) and middle (0.4-1.3 m) soil depths, and its impact on soil thermal regime at shallow depths was significantly greater in spring melting periods than in summer. The impact was minimal in freezing periods and in deep soil layers. The overall annual effect of CHT by liquid flux is to increase soil temperature in the active layer and accelerate the thawing of permafrost at the study site. The topic is interesting and suitable for the journal The Cryosphere. I do have some major concerns such as: The introduction should be consistent with the topic. The discussion should be stated as different format with the results. The concepts such as the difference between unfrozen water and liquid water content need to be defined in the context. I listed my major concerns as follows:

1. The abstract is too long and needs to be shorten.
2. My main concern is that the data in 2008-2010 was used in this study, which was too old. I recommended to update the time series of the data.
3. “Tibetan” or “Tibet”? It should be maintained consistent throughout the text.
4. The logic of the introduction is a little bit unclear, and some of its details are not adequately rigorous. As I know, there were many works that has been done on this work based on different models, such as Yu et al., 2018, Liquid-Vapor-Air Flow in the Frozen Soil, JGR, and He et al., 2018, A coupled model for liquid water-vapor-heat migration in freezing soils, Cold Regions Science and Technology. I suggest that the authors to provide a thorough review on this aspect and state the previous pros and cons on the previous work, then clearly state that why this need be done to make a real progress and what is the different between them?
5. How representative is the Tanggula site for such a large area of the Qinghai-Tibet Plateau? It is suggested to add more points in different regional to illustrate the problem. Why did the authors select this site?
6. Why to use SHAW model, the introduction is not clear.
7. In model settings: Table 1 should remove to this section, and what I want to know is how this parameter is obtained, which has important effects on the simulation results.
8. The simulation error is large, especially for soil moisture (Fig. 1), which will bring large uncertainty to the simulation results of sensitivity experiments in this paper, and the results are doubtful.
9. Why is the analysis divided into shallow, middle and deep layers, and what is the basis for the layers?
10. What is the spring meltwater referred to in this article? If it is snow melt, how much snow is there on the Tibetan? How much melt water is there and how much of it is infiltrated into the soil, all these questions are not answered in the article.