Dear authors,
One of the previous reviewers has now has given a positive judgement on the paper. Therefore, I have decided to accepted your manuscript, subject to minor revision, suggested by the reviewer.

In addition to the reviewers’ comments, I suggest some further improvements.


The abstract could be a bit clearer to serve as a stand-alone text.

For example:
l. 7 I miss the assertion that the choice for SSP3 without and without air pollution policies, leads to a ‘maximum’ effect, and as a consequence using other SSPs would likely give lower impacts.

Abstract lines 9-15, the quantification of ozone/pm2.5 changes; as well as the temperature, precipitation: please clarify whether these numbers apply to the changes in the SSP3 baseline versus SSP-3 NTCF scenario by 2050, or averaged over the simulation period. Would it be possible to add uncertainty intervals?
Clarify also how the choice of these two scenarios provides information on “policies that optimally address both climate change and air quality”, as this is clearly not the case for the SSP3 vs SSP3-NTCF scenario. SSP3 rather expresses the absence of climate policy.


l. 21 Suggest to introduce also the term SLCF, along with NTCF, as many will know this term better than NTCF. IPCC considers the two synonymous.

l. 50 Clarify whether the use of the word likely (here and following sentences) implies a quantified uncertainty, as in IPCC uncertainty language? This is in particular of importance, as IPCC will probably use this paper in its forthcoming assessment.

l. 61 The risk factor is for premature death and mainly due to non-communicable diseases.

l. 116 Some more elaboration on the issue of 3 realisation vs 1 realisation for these transient simulations would be useful. What is the range within single models of the results, and what does it imply for uncertainties from single realisations. Is this information used, or could this information be used for improving error estimates of the multi model average?

l. 190 explain the rationale of not including methane the experiments.

l. 191 I didn’t understand why it is difficult to reduce air pollutant- while not changing CO2. Isn’t that the case for e.g. any filter, scrubber, catalysts on top of existing technology. Or is the idea that technology changes are needed to achieve further air pollutant emission reductions, and those technologies are probably also delivering on GHG gas emission. The terminology ‘keep CO2 fixed’ is confusing, you keep the scenario assumptions constant, but the emissions change according to SSP3.

l. 402 Conclusions. It seems that the omission of aerosol nitrate from 3 of 4 models, may lead to a substantial low bias in the estimations of aerosol and resulting effects. It would be appropriate to mention in this in the conclusions.