Dear Authors,

Thank you for the substantial improvements. A reviewer has asked for some additional but minor changes prior to final acceptance, which I enclose below. I agree with the reviewer about the ordering of Figures 1 and 2 which should be swapped. Please also deal with points 2-5 and provide a revised manuscript. The manuscript will be copy-edited for language by Copernicus after acceptance.

Regards,

James

Reviewer comments:

1) I would still prefer authors to change the order of current Figure 2 and Figure 1. It’s better to show your research domain first. The SHEBA ice camp trajectory should be added in the Figure (* I cannot see SHEBA trajectory in MS paper copy, however, I can see in in online MS, please check, It is better to mark the direction of ice camp drift by adding the initial and ending points).
2) The Table 1 and Table 2 can still be improved. e.g. what do you mean “includes different treatments for snow or bare ice surface. What “different treatments” are you talking about here? “Initialized from Eq. (1)”, but at what time?
3) Explain the symbols in eq. (1).
4) P10, L35 “Both WRF-Noah and WRF-HIGHTSI overestimate the sea ice albedo due to the simulation of a too late snow melting over sea ice and the lack of melt-pond effect in summer. WRF-HIGHTSI has a larger overestimation of albedo than WRF-Noah.”
This is still difficult to understand. When snow is presented, the surface albedo is equal to snow albedo, say 0.8 since snow is pretty white. Then if snow-melting is delayed, the surface albedo will remain high, e.g. 0.8. I cannot see how a “overestimate the sea ice albedo (A)” can be linked with “due to the simulation of a too late snow melting (B)” When snow melted away entirely you will subject to deal with sea ice albedo. “A” is albedo and “B” is about time. Please explain.
5) Same paragraph: “Both WRF-Noah and WRF-HIGHTSI underestimate the surface upward shortwave radiation due to the underestimation of downward shortwave radiation. Such a bias in upward shortwave radiation simulated by WRF-HIGHTSI is partly compensated by its overestimation of the sea ice albedo” The upward shortwave radiative flux should be largely associated with surface albedo, e.g. if albedo = 0 there is no upward shortwave radiative flux, then why downward shortwave radiative flux was underestimated?
6) The current manuscript is quite long. The language can still be improved, e.g. sentences can be expressed in a more compact way. The language polishing maybe needed by a native English speaker.

I'm happy to accept the manuscript for publishing in GMD. Thank you for publishing here.