The reviewers agree that the paper is a useful contribution that presents a novel explanation for the land-ocean contrast in surface temperature change seen in models and observations. The author's have presented a sufficient response to the reviewer comments.

Regarding the first several comments of referee 1, I encourage the authors to look for observational literature references that can be added to the textbook references. (Just because something is in a text book doesn't necessarily mean it is actually correct!) While I don't seriously doubt that these assertions are, indeed, correct, since these points are central to the arguments in this paper, having more direct observational references would strengthen the paper. (And also allow readers to access the magnitude of deviations from the assumptions made.)

Please make sure the references to the textbooks that are used are given in proper format in the revised manuscript.

For the reply " Observations of the land surface energy balance generally show a net longwave cooling at night with values typically well below 100 W m−2. " it would be useful to provide an observational reference for this. (or, if this is not readily available, show this information in supplementary materials.)

In the response to Reviewer 1 comment 11:, please add this analysis, associated discussion, and related figure wither in the revised manuscript or, perhaps, as either supplementary material or an appendix in order to not breakup the flow of the manuscript.

In line with some of the reviewer comments, I urge the authors in their revision to discuss a bit more how the current analysis could be extended. Again, the referee's have agreed that the author's arguments are sound - so the purpose of the paper should not solely be to justify the methodology and assumptions used in the current analysis. It would improve the contribution of this paper to the literature to discuss a bit more the potential implications of taking into consideration the next order of analysis. (For example, could additions improve the comparison between this analysis and the CMIP models in Figure 3? In particular I note that this figure seems to show that the result from this analysis appears to be at the low end of the model range. Given the author's methodology and interpretation, what inter-model differences might explain this range?)

In the discussions of the current simplifying assumptions it would be useful if the authors could also discuss (perhaps at the end of the paper) the potential implications if these analysis were taken a step further. For example, in the revised introduction the authors note that "the ground heat flux does typically not exceed more than 100 W m−2, which is comparatively small to the absorption of 800 W m−2 ". What might be the implications of taking this into account? Would this be a O(10%) correction to the result?

How about the opposite case, moist land-surface regions (including rivers, wetlands, lakes, etc.)? Presumably there would be greater diurnal heat buffering there as compared to a desert area? Would incorporating such differences significantly alter the estimate?