I have read through the authors response to the reviewers and confirm that
I agree their replies to the comments are accurate and that I also approve
the suggested revisions that have been made to the manuscript, which in
response to the helpfully thorough reviews have made some significant
improvements to the explanations in the manuscript.

There are however several minor technical issues which I list below, and
once these revisions have been made, I can confirm my approval for the
manuscript to proceed to publication in the main Atmospheric Chemistry
and Physics journal.

The required technical revisions are listed below numerically:

1) Re: reviewer 1's request for detail on the vertical resolution, the suggested
addition is fine, but are the units for the values given there in hPa?
(I.e. is the mid-point of the lowest level at 992.56 hPa [not withstanding
pressure variations from the terrain-following element of the hybrid
sigma-pressure co-ordinate system]?). If so please add that unit to the
eta values given there.

2) Re: reviewer 1's comment on the "self-nudging", the first of the two
suggested added sentences here is fine, but in the 2nd sentence please
rephrase the last half of the sentence. In particular, the phrase "which in
turn leads to more confidence in, e.g. estimates of effective radiative forcing,
but is less useful for the purposes of comparing modeled aerosol properties
with observations." Although this explains the issue re: the distinction between
nudging to "best estimate fields" from meteorological re-analyses, and nudging
to some previously archived sequential instantaneous fields (e.g. 3-hourly)
from some reference simulation with the same model) the phrase "which in
turn lends more confidence in e.g. estimates of effective radiative forcing"
needs to be revised. I think the main point is not that the self-nudging gives
more confidence re: radiative forcings necessarily (although it does for
comparisons to observations), rather the approach suppresses (or over-writes)
fast feedbacks in the simulations, which then derive "radiative forcing" rather
than "effective radiative forcing" (the latter term including the fast feedback
effects but the former term explaining the effects have been deliberately
excluded as part of the experiment-design).


3) On the next reviewer 1 suggested addtion (page 33 line 7), just a
terminological point to please change "ERF ACI" to "ERFaci", the lower-case
aci rather than ACI being the recommended terminology from the radiative
forcing of the IPCC 5th assessment report (Myhre et al., 2013).
Please also replace other instances of "ERF ACI" and of "ERF ARI" with
"ERFaci" and "ERFari" respectively.

4) Re: the added text in reply to reviewer 1's comment page 21 line 3,
that text is fine but since that statement refers only to SO2 (some of the
aerosol metrics indicating an apparent worsening of skill with the addition
of the nudging) please add "SO2" after "climatological averaged".

5) Re: reviewer 1's comment re: the BC in the UT (page 34, line 13),
please add "BC compared to" between "overestimate" and "the aircraft
measurements" so that is also more specific, also replacing "aloft" with
"in the free tropopshere" for the same reason. Please also correct the
typo "participatig" with "participating".

6) That's great re: the reply to reviewer 1's comment re: the query on
the clear-sky flux info from the models, but when I checked that link
https://wiki.met.no/aerocom/optical_properties that seemed to be a
broken link or typo in the http-address (or maybe the site was down
for maintenance at the time I checked?).

7) Re: reviewer 1's question about page 62, please correct spelling of "froplet"
and add the word "cloud" before "droplet effective radius" so it's absolutely
clear you mean cloud droplet reff not aerosol particle reff.

8) In the reply to reviewer 1's question about page 71, I don't think this
reply is correct because the wording "direct radiative forcing" and
"cloud radiative forcing" are still shown in Figures 9 and 10, and in the
caption. Please correct these to "ERFari/RFari" for Figure 9 and
"ERFaci/RFaci" for Figure 10 in both the image and the caption.

9) In the reply to reviewer 2's suggestion re: p6 lines 13-16,
please revise the existing text to clarify what is meant by
"the background tracers and size-modifying tracers are treated
as if they were particles". Is this referring to the parts of OsloAero5.3
(or another module in the host model) that deals with the oxidation/conversion
of the gas phase tracers (i.e. the aerosol precursor gases)? And is that
what is meant by "aerosol chemistry" (i.e. shorthand for "aerosol precursor
chemistry")? If so, suggest to change "aerosol chemistry" to "oxidations",
add "of aerosol precursor gases" and replace "(the aerosol lifecycle scheme)"
with "(the gas phase chemistry scheme)" or similar. Isn't that the key thing
here re: the difference between gas-phase and aerosol-particle tracers here
in the context of the "aerosol chemistry, transport and dry deposition" stated?
Is that what you meant in terms of the tracers are treated as particles (i.e.
are they given non-zero sedimentation velocity according to their particle
size in the dry deposition routines [and surface area density and uptake coeff
for the "aerosol chemistry"? i.e. is it actually heterogeneous chemistry you
mean here when you say aerosol chemistry?]. Please can you re-word the
suggested text-addition to make this clearer.

10) Re: reviewer 2's comment re: the isoprene and monoterpene reactions,
suggest also to add mention that the chemistry follows offline oxidant approach,
so the products from the chemical conversion of the second-reactants
(the oxidants) do not need to be included in the products on the right-hand
side of the chemical-equations.

11) Re: reviewer 2's comments re: the AR5 emissions (Lamarque et al., 2010)
you have written "CMIP6" but you mean "CMIP5" here. Please correct.

12) Re: reviewer 2's request for reference for the nudging, I'm wondering
about the stated "relaxation time of 6 hours". I'm wondering if maybe you
meant "frequency of 6 hours" in that the model relaxes back to (interpolations
between?) different successive meteorological re-analysis field every 6 hours?
If that is what was meant please replace "relaxation time" with "frequency"
as I think that's clearer what's meant. I guess the relaxation time must be
on a shorter timescale than that?


13) Re: reviewer 2's request to explain the rationale for weighting the
averaging of the "instantaneous-calculated AODs" by the clear-sky fraction
rather than the recommended approach to calculate the hygroscopic-growth
applying the clear-sky-RH rather than the gridbox-mean RH, OK, yes that
response in explaining this alternative approach. I see that, when calculating
means, doing so with a weighting to one-minus-cloud-fraction will then make
that mean-AOD be more representative of clear-sky conditions (and therefore
minimise the effects from overestimated-aerosol-optical-depth due to
elevated humidity in cloudy gridboxes). And the added text now explains
that the CAM4-Oslo and CAMS5.3-Oslo diagnosed this differently than the
model, so that is good, and clear. However, again -- I agree it would help
in the reply to reviewers comments to be able to refer to the discussion on
the Wiki (although maybe contributors might I guess potentially object to
that then being referenceable and citeable in this way for some other reason).
In any case the link there did not seem to work when I tried it on 27th Aug.
As per the other comment, please check this and add the corrected link
(assuming the AeroCom team approve the link can be used in this way for
the review without the consent of the contributors).