I think the manuscript is well written and it is a fine contribution to the literature.

However, I believe the manuscript might improve a bit more with a few adjustments, which are given below.

1) Line 150: Authors are neglecting the effect of the pitch and roll angles on the wind measurement of Trisonica mini which is not a bad estimation if the copter does not pitch and roll during flight (i.e. maybe only during hovering at a certain altitude). However, in line  211, the flight speed was given as between 2 - 5 m/s which I think will force the copter to roll or pitch at about 10 - 15 degrees. Considering the placement of the anemometer (~0.8 m over the propeller plane), the angular momentum might become non-negligible I think. Therefore, there might be a bias in the wind speed measurements during the flight. I think this should be clarified, if there is a small effect this needs to be shown by the authors. I think checking Donnel et al. (2018) paper might help with this (https://doi.org/10.2514/6.2018-2986).

2) Figure 3, from the figure it looks like the widths of each spike are not similar. If the flight characteristics are the same for each repeated crosswind flight why than the width of these spikes are different? Is this because of the environmental conditions? Additionally, why the background CH4 is increasing over time? CH4 concentrations are between 2.2 - 2.3 ppm before 800 s, and it increases at about 2.5 ppm at the end of the measurement. I would expect that the background signal will be more or less similar during the flight and when the drone sees the plume the spikes will occur. Maybe this figure needs more explanation. 

3) Line 259: Why the flight was conducted 130 m away from the source? Why not closer, was there any restrictions?

4) Line 265: This is a bit cryptic. How did the authors find the lower quantification threshold here? What are the standard flight conditions? Also, in Table 1, how did the authors come up with the Mira Pico uncertainties? When I check the manual only information given about the instrument is the sensitivity which is <1 ppb/s and the drift which is given as 30 ppb. Maybe adding a bit more explanation for Figure 7 might help here.