Dear author, co-authors,

It has taken quite some extra time to come to a decision also due to a high workload but also checking carefully again your long paper that contains a large amount of information. I have checked critically your response to the comments raised by the reviewers and this has further increases the information presented in this paper that, according to me, could now easily cover two papers; one more on the GEM-MACH modelling system (so more a paper suitable for GMD) and one on the actual analysis of the ship emission impacts. However, this is not what I recommend at this stage now that you have addressed all the comments and provided this revised document. Overall it seems that you have properly addressed the comments and have made the edits that also reflect these suggestions for arriving at an improved version of the ms. Consequently, I accept the paper for publication in ACP after you have addressed my last two small comments that came up in reading-over again the paper.

In the description of the model experiments: “The eight-month simulation was conducted by a series of staggered 30-hour runs with a 6-hour (meteorology-only) “jump-back”, starting at 00 Z daily, to allow meteorological spin-up; the meteorology is initialized at 00 Z using the Canadian Meteorological Centre’s regional objective analyses while chemistry (delayed for 6-10 hours from run start time) continues from the preceding run”.

This is explaining the set-up of the model runs using quite detailed/technical jargon and was wondering if you could state this in a manner that is also easily to apprehend by readers not too familiar with such technical details. What is 00 Z?

Pp 12-13; You show in figure 4-7 a comparison of the long-term time series of the simulated and observed tracer concentrations. First of all, although it is pretty clear it would be good to state here that these are surface observations. Then it is also good to know the typical reference height of these observations given that the reference height of the model simulated concentration is generally ~10m (given surface layer depths of 20m in low level terrain). And then in your discussion of what might explain the differences between the simulated and observed O3, I wonder also to what extent this might be due to issues on BL mixing and dry deposition as well as the imposed boundary conditions (also in terms of high altitude ozone). Since you use daily mean value, these might be strongly effected by nocturnal biases associated with the representation of the nocturnal inversion layer and then the large sensitivity to small differences in dry deposition. You mentioned in the model introduction that you have adjusted the snow-ice deposition rates of O3 but also having this biases in summertime there might be issues on the representation of deposition to the bare soil/sparse vegetation/water surfaces

Dear author, co-authors, thanks for addressing these last list of minor comments on the revised version of your ms submitted for publication in ACP. I read your response and this appears to handle well these comments accept the added statement on the nocturnal O3 bias and the role of the SBL.
Your proposed revision on this does not come out well according to me:
"The nighttime model bias can be a result of the model’s difficulty in simulating (or resolving) the stable nocturnal boundary layer: the lowest model level may reside in residual layer rather than in the surface layer (where surface O3 monitors are located).”,
My point is that the model's lowest layer represents the surface layer and its processes and that the layer(s) above might represent the lowest part of the residual layer actually present in the atmosphere if indeed the SBL would be only about 20m deep (which I doubt, over snow/ice cover it could be indeed only some meters deep but generally more ~50-150m). I would suggest to phrase this sentence to something like:

The nighttime model bias can be a result of the model’s difficulty in simulating (or resolving) the stable nocturnal boundary layer and where small differences in the actual O3 sources and/or sinks, like O3 dry deposition can have a large impact on O3 concentration gradients that might also be reflected in significant differences in observed and simulated nocturnal O3 for different reference heights”

With this last minor modification, I accept the paper for publication in ACP,

Laurens Ganzeveld

Dear authors, co-authors, with this last minor modifcation, I accept the paper for publication in ACP and let's now see how this is further picked up by the interested readers,

Laurens Ganzeveld