General Comments:

In this work, stratospheric partial columns of NO₂ and NO obtained from FTIR measurements performed at Zugspitze during 25 years, and at Garmisch during 18 years, are presented. The authors explain how they split into tropospheric and stratospheric columns taking into account the maxima observed in the retrieved concentration vertical profiles, and from their corresponding partial columns averaging Kernels (above and below 16 km of altitude). The proximity of these 2 stations, which are however at very different altitudes (2964 and 745 m.a.s.l, for Zugspitze and Garmish respectively) are used to assess de influence of the tropospheric components on the stratosphere. Results show that stratospheric partial columns are weakly influenced by the tropospheric pollutants. However some stratospheric columns are affected by high pollution tropospheric events. To remove these outliers from the series of measurements, a contamination filter is applied to data. Then, the daily and seasonal evolution of NO₂ (Zugspitze and Garmish) and NO (Zugspitze) is obtained. NO₂ presents a linear increase along the day that is consistent for both stations. NO has a linear increase during day until noon. Due to the solar elevation and the increase in the O₃ concentration, NO₂-NO equilibrium changes after noon, and from that moment on, NO concentration has a non-linear behaviour. Thus, only NO data before noon are compared with NO₂ daily evolution. NO₂ and NO present similar seasonal evolution with maximum values in September and minimum values in December and November respectively.

The entire manuscript (text and figures) is clearly presented. The used approaches/methods are well described. I think that the results and methods used in this work will be useful for the validation of models and satellite data. Thus, I think that this paper should be publish in AMT. However, before publication, some questions should be better clarified.

Specific Comments:

1. Line 34: Please, include some references about the lighting and air traffic controlling the NOx concentration in the upper troposphere.

1. Line 64: MAX-DOAS measurements generally obtain information for lower SZA (compared to the high SZA of the twilights).

1. Lines 66-67: That is not entirely true, if the Free troposphere is considered to be representative of concentrations between the Boundary layer and the tropopause. In fact, by applying the method described in (Gomez et al., 2014), the Free Troposphere NO₂ concentration can be estimated from MAX-DOAS measurements performed at mountain stations. In (Gil et al., 2015), for instance, that method was applied to Izaña MAX-DOAS data carried out over 3 years to study the seasonal evolution of NO2.

1. Section 3.2: What temperature and pressure vertical profiles are used in the model?

1. Line 194: Please, explain how are the partial column averaging kernels obtained.

1. Section 4.2: How often are these pollution outliers observed out of the studied cases? It would be interesting to study also the high pollution episodes and how these tropospheric events affect the stratosphere.

1. Section 6.1: How do you explain the difference of the NO2 seasonal evolution observed at both stations between April and June? (Figure 6).

Technical Corrections:

1. Page 91: Do you mean “solid” instead “sound”?

2. Page 112: “excited” instead of “exited”.

3. Line 139: MCT meaning.

4. Line 192: “1a” instead of “1 a”.

5. Line 217: Something is missing in “of ca. 1”?

6. Line 248: you could mention the horizontal distance in km to show clearly how close the stations are.

References:

• Gil-Ojeda et al; NO2 seasonal evolution in the north subtropical free troposphere. Atmospheric Chemistry and Physics, vol. 15, pp.10567-10579, 2015.

• Gomez et al; Long-path averaged mixing ratios of O3 and NO2 in the free troposphere from mountain MAX-DOAS. Atmospheric Measurement Techniques, vol. 7, pp. 1-14, 2014.