Dear authors,

thank you very much for submitting the revised manuscript. It took some time to get it reviewed by the Referees, who either acted during the open discussion phase or not. As you can see, one of the Referees is satisfied with your responses and applied changes to the manuscript. However, the second Referee has some important objections regarding the acceptance of the manuscript in its present form. The most important objection refers to the need of applying the relationship obtained in Figure 13 to an independent set of measurements in order to assess its potential value for other users, cities, or pollution events. Since this part of study is essential and the most important one, it is indeed very useful to assess the usefulness of the derived relationship in order to generalize and expand its use and applicability.

Besides, I have made (in the annotated pdf manuscript) several notes and suggestions for modifications/improvements throughout the manuscript, which should be taken into account.

I would be happy to receive your further revised manuscript along with point-by-point responses and description of the applied modifications.

Sincerely yours

Dear Authors,

please revise the manuscript addressing the comments of Reviewer#5 concerning the limitations of applicability of the derived relationship between AEC and PM2.5.

The results of Figure 14 show that the poor correlation of the 3rd event (please report it on the Figure), leads to the reduction of R-squared from 0.75 (for 2 extreme pollution cases) to 0.61 (for 3 cases). It is clear that the relationship derived from the 2 extreme cases is not applicable to the 3rd case, which is not very strong. This is actually acknowledged in the manuscript. Therefore, this should be further emphasized, i.e. the prerequisites and limitations of the applicability of the derived relationship from the 2 extreme pollution cases. Actually, this is an interesting finding and deserves to be highlighted, in order to warn the community about it.

Another issue that should be addressed and discussed in the manuscript with relevance to Figure 14 is the slope of the applied linear fittings. Please apply linear fitting and show the slope for the 3rd pollution case, independently. What changes obviously between the scatterplot of the 2 extreme events and the scatterplot of the weaker 3rd event is not only the correlation, but also the slope. In investigations of such relationships and interdependencies between two parameters, like AEC and PM2.5 in the case of this manuscript, it is important to assess the slope, which appears to change drastically in the 3rd case. This is in great part due to a population of pairs with strong PM2.5 values, between 40 and 50 micrograms/m3, which correspond to rather low AEC values. The result is a probable reverse of the slope value from positive to negative. The causes of this difference and change should be discussed.

In general, it is very important to further strengthen the discussion of the relationship between AEC and PM2.5, with emphasis to its applicability and limitations, and to highlight this, which is the most interesting part of the analysis and makes this manuscript approppriate for a published ACP research article.

In summary,

(i) please address the concerns of Reviewer#5
(ii) emphasize the applicability and limitations of the derived relationship from the 2 extreme pollution cases with respect to the weaker 3rd event. You should do this in the Abstract, where now only the two last sentences refer to generalized findings and not to descriptions and reports of measured quantities, as well as in Conclusions where two sentences in the last paragraph (lines 7-10) do so.   
(iii) discuss the slopes of scatterplots in Figure 14, also including the slope of the 3rd dataset in this Figure

Best regards

Dear Authors,

thank you for your detailed answer and the revised manuscript. I am fine with the provided reply to the comments, and especially with the improvement of the results of Figure 14 concerning the new data relevant with the third examined dataset/pollution event.

There are only the following two very minor points that I would like you to clarify/address:

1. In the new Figure 14, it seems that the group of points in previous Figure 14 with PM values between 40 and 50 μg m-3 and AECs below 0.4 km-1, have disappeared in the new Figure 14, where the new/extended dataset seems to now fit very well, and also to provide a very similar slope to that of the two major pollution events. Please, explain how this was achieved. In fact, the third dataset has been extended, from 05-08/10/2016 to 03-10/12/2016, thus the group of points would be expected to be enriched (more days/points) but also to include the previous (05-08/12) group of points.
Also, relevant with this change, please correct the caption of figure 14 referring to the third dataset, i.e. blue crosses, and modify the third sentence referring to the time period.

2. Please make the requested modifications shown in section 5.3 of the annotated document.

Best regards

Dear authors,

thank you very much for your reply and the last amendments made to the manuscript.

I am delighted to let you know that it is now accepted for publication in ACP.

Best regards