|This paper is undoubtedly improved from the first version, but retains several problems, and in fact the revision has introduced has some new ones. |
Therefore I still do not think it acceptable in its present form. I would have no objections if the paper was revised a second time, as it can play a useful role in stimulating more detailed calculations, even if the results in this highly-idealized model setup are found not to be supported by those detailed calculations.
1. I was encouraged to read in their response to my initial comments that the authors had “deleted … the pure radiative equilibrium case”. I was disappointed to find that they had not done so. It features prominently in the figures and in Table 1, which is the main results table. I maintain my view that results from the pure radiative case are of little relevance to any real world cases, and the authors have not defended its retention, but continue to highlight the results from this case in, for example, the abstract.
2. My original comment *5:15 seems to have been ignored. The authors say they “agree on the facts” but the revised text does not make clear that the model includes ONLY a cirrus layer, and no other clouds. (and I apologise if I miss it, but the authors have not signposted where they have done so in their response)
3. In addition, I consider my comments on *7:10 and 8:6 to have been effectively ignored. In response to *7:10 the authors say they “have eliminated this discrepancy” but I do not know what this means, as the Figures in question remain in the manuscript. For 8:6 I had a specific comment on a value in the text which has neither been challenged nor corrected and my contention that the text is incorrect remains. The response to 12:18 is also somewhat unsatisfactory. The upper tropospheric amplification in the CO2 case is not the result of condensation at low temperatures, but due to the divergence of moist adiabats at lower altitudes. The same can also be true for the contrail case.
4. At line 298, Figure 8 has serious problems and has gone backward since the earlier version (when it was Figure 11). The right-hand side of the figure purports to show the instantaneous heating rate change due to CO2, but in fact the red lines are the contrail heating rates repeated from the left hand side. To add to the confusion the caption says that the left hand side is CO2 and the right hand side is contrails, when it is the other way round.
5. The discussion surrounding Table 1, the main results table, is confusing. As far as I can tell, the values in the table have changed (due to the change in insolation parameters) but the text is essentially unchanged and now inconsistent. At line 326, it says the contrail RFa is small, but it is no longer small, but 0.42 W m-2, and so a substantial fraction of the CO2 value (0.72). The same is reported at line 400 and again at line 408 where the SW and LW forcings are said to be “nearly cancelling” where they are not. The LW forcing is almost double the SW forcing now, and the cancellation is only partial. In the radiative-diffusion case, the climate sensitivity for the separate LW and SW cases is 0.24 and 0.27 respectively. So while the authors write at lines 333 and 334 that SW efficacy is larger than the LW efficacy, and this is of course correct, perhaps it is as notable that they are so similar, given the different signs of the forcing and the different surface atmosphere partitioning. At line 405, presumably RFi should be RFa (as efficacies for the instantaneous case are never shown) and the following text says there are “strong” departures from unity. But they are not actually that strong (the efficacy is about 0.75). Part of my point here is that the authors could easily be quantitative in the text – “strong departures” is a matter of opinion, and citing the actual values would provide a better perspective for the reader to judge.
6. The two sentences at line 402 are misleading. The “zero mixing case” (by which the authors mean the radiation-only case) is not a credible real-world case and hence the deduction from this case that the temperature sensitivity could be negative is beyond speculative, on the evidence presented. This issue is also present in the abstract at line 20, where it refers to a “low mixing case”, even though the text only ever shows that a cooling occurs for a “no mixing case”. I am not sure why the abstract doesn’t focus on the diffusive case, which is more realistic. One could easily interpret the results as showing that the difference in efficacies between the CO2 and contrail cases are really rather small, given the quite different nature of the forcings, and perhaps by being quantitative in the abstract it would allow the readers to make their own minds up, instead of being left with a potentially misleading impression.
7. In passing I note that the new text focuses on the mid-latitude summer continental case, while also maintaining that the timescales for mixing are determined by large-scale eddies. I have struggled to find anything quantitative to challenge this statement, but some care is needed. During summer in mid-latitude continental regions, my guess is that convective heat transport is at least competitive with the vertical heat transport from large scale eddies, and may even be larger.