Dear Dr. Joly and co-authors,
Thank you for your efforts in addressing the comments made by the reviewers to your submitted manuscript. Your responses and the additions you have included in your revised manuscript improve the depth and presentation of the material you have presented.
There remain some issues that should be addressed to improve the readability and clarity of the manuscript, and some issues with usage and formatting that will likely be addressed in copyediting.
A few specific comments:
P2,L23: "thermodynamic structure and atmospheric composition"
P6L5-7: "DFB" = distributed feedback; "ICL" = interband cascade laser
P6,L13: "Judson" is "Teledyne Judson Technologies"; PTU should be explained
P6,L14: "PTU" = pressure, temperature and humidity
P6,L23: "sensitivity" represents a signal/analyte quantity--do you mean "detection limit" (over what time?)? Table 1 shows values for both CH4 and CO2 in 2015 while P5, L21-22 states that in 2015 AMULSE was a single gas CH4 methane instrument and the dual-gas version was introduced in 2016. Table captions go at the top of the table, figure captions below.
Figure 1: comment in text on CO2 drift from early to late part of time series. Average of the difference does not capture--average absolute difference would be a better metric in this case. Stating the std dev as you have is valuable. Greater color difference between traces would be helpful.
P11,L7: "CO2 near the surface", "at night"; one reviewer suggested a reference for this. The artifact in H2O due to outgassing on ascent is mentioned, which explains the non-physical values shown in Fig. 5c, but why aren't descent data shown--the text mentions valid data then. A temperature profile plot would be a useful addition to Fig 5 and help with identification of the tropopause.
P15,L9: interpolate or average? Doesn't the in situ data have higher vertical resolution so that you would need to average the data over the depth of a model layer? Fig 7--would probably be better to put the reference trace on top, so that it is clearer that the profile in the shaded (non-measured) region is not the from the sonde.
P16,L11: "red"? Red might actually be better
P16,L13: "This" here is confusing since it naturally refers to the "good agreement" when you are addressing the observed differences in the window regions. "The difference seen between the simulated IASI and CrIS window channels is due to the temporal and spatial differences between the selected pixels."
P17,L7+: as a reviewer asked/suggested, explain 'EXP'. You did so in your response, but not in the manuscript.
P18: adjust colors in Fig. 9 to be more distinct. Clearer structure in the figure caption. Why is SONDE capitalized?
P18,L8: What "specific case"? "For the case of 2018-04-07, it can be seen in Fig 9 that..."
P20,L24: reference for the higher difficulty in CH4 simulation in CTMs (just source term or chemistry as well?) and the major scientific questions regarding CH4 trends.

Dear Dr. Joly,
Thank you for your efforts in addressing my final comments on your manuscript, I am pleased to accept it for publication in AMT, subject to minor technical corrections.
Troy Thornberry
AMT Associate Editor