Review provided in the attached pdf file

Overall, this is an interesting study making use of long-term MIPAS retrievals to investigate the potential sources controlling the spatio-temporal distribution of upper tropospheric pollutants. This study exploits information from enhancement ratios (ER) in comparison to emission ratios from known source types (e.g. forest fires) to determine the likely source of the pollutants. The manuscript is generally well written, figure presentation is good and would be an interesting addition to ACP. Therefore, subject to some minor comments, it is suitable for publication in ACP.

1. The GFED data used in Figure 3 is from an older version (3.1). Therefore, it makes sense to exploit a newer version e.g. (vn4.1s or vn5).
2. The authors use the ERs to infer source information about the pollutants retrieved by MIPAS. However, in places these statements are overall “conclusive” and need to be weakened (e.g. Page 17 Line 33, Page 20 Line 7 and Page 9 Line 20, Page 12 Lines 21-22) without the support of e.g. a model.
3. The methods used in Section 4.3.2 need some more detail. From reading the text (Page 13 Lines 15-22), it is not overly clear what ESDs are and how they are calculated and why λ is needed in the regression model.
4. In Section 4.3.2, why focus on individual months instead of the seasons as used earlier in the manuscript and why choice those specific months (i.e. Feb, Apr, Jul, Oct)?

Page 9 Line 17: Reword “even better visible” to something like “more evident”.

Page 9 Line 18: Can you provide a motivation for looking at 11 km (e.g. why not 10 km)?

Page 16 Lines 14-19: Suggest rewording as not overly easy to follow.

Page 16 Line 28:  Add “a” between “as” and “possible”.

Page 17 Line 28: This line needs rewording as not clearly written. Also, what is the “two times larger relative retrieval error” based on? Is this from the literature, the data product, do you calculate this?

Page 18 Line 33: Add “any” between “hardly” and “information”.

Page 21 Line 8-9: This sentence needs rewording as difficult to follow.

Page 21 Line 22: Add “an” between “as” and “important”.

Page 21 Line 26: Replace “better” with e.g. “stronger”.

Page 22 Line 24: Replace “become better” with “becomes more”.

Table 2: For “a” why give the value as ppbv. For consistency and presentation, keep as pptv.

Table 3: Same as Table 2, why use “a”? Best to use consistent units and not a %.

Figure 6: Panel title and x/y-axes are missing for PAN.

Figure 8: I do not understand the delta HCN / delta CO units of 0.01 pptv/pptv while the colour bars show ER %.

Figure 3: For the colour bar units, what is 10 to the power of? Just says “10^”.