|The manuscript clearly presents the models which have been used in the POLARCAT model intercomparison (POLARMIP), characterizes some of the general behaviour of these models in the Arctic with useful analysis, and presents an extensive comparison with atmospheric composition observations made in the Arctic.|
The paper has tried to disentangle the impact of uncertainties in emissions from characteristics in the models (chemical and physical parameterisations) in describing the composition of the Arctic atmosphere. It is suggested that by using the same emission data set, differences seen between model results can be easier interpreted and attributed to model differences or deficiencies. Also, when for certain species close to their principal emission regions a similar behaviour/deviation is seen among the models with respect to the observations, it allows (with some confidence) to pinpoint possible uncertainties/inaccuracies in the emission data set. On the other hand, when specific models compare remarkably well or bad with observations, the authors have tried to find out which processes might be responsible for this behaviour.
It is appreciated that the authors have done this effort of attributing differences with observations to specific model parameterizations or characteristics, although the statements are sometimes a bit vague and could have benefitted from some more underpinning. However, I think that attributing firmly the observed deficiencies to specific paramaterisations demands for a large amount of extra sensitivity simulations, which is probably outside the scope of this model intercomparison exercise. That is however possibly the way forward to characterize deficiencies in CTMs and improve their capability to describe well the Arctic atmospheric composition. It is assumed that the two additional POLARMIP papers of Arnold et al.  and Monks et al.  focus more on some specific model behaviour.
I appreciate the response of the authors in taking into account the comments on the former version of the manuscript.
Therefore, there are only some small technical comments listed below.
page 2, one but last line : add ":" after "***now".
page 6 : "VOC" used but defined only later (on page 9).
page 11, line 14-15 : "suggested by (Mao et al., 2013)" -> "suggested by Mao et al. (2013)".
page 17, line 10 : "carbon monoxide (CO)" : CO is already defined on page 6.
page 18, line 20 : "June-July" with a short "-", whereas most other appearances have a long "-".
page 20, line 9 : "June 18-July 15" with a short "-", whereas most other appearances have a long "-".
page 20, line 22, 23, 28 : "Figure" -> "Fig."
page 22, line 16 : "Figure" -> "Fig"
page 23, line 14 : "Figures" -> "Figs."
page 23, line 20 : "Figure" -> "Fig."
page 23, line 23 : "J" is once italic, and once straight.
page 24, line 20 : "ppb" has a larger font here than on all other places in the text.
Figs. 3 and 4 : Please add some extra values on the vertical axis.
page 57, caption : maybe change "France" -> "POLARCAT France"