Dear authors,

I have now received two reviews on your manuscript. The official report by Ref#2 is extensive and critical, but the bottom line is that the remarks should improve the manuscrip. Ref#1 is still sceptical, and there is quite some overlap in the criticism on the calculation of the sulfur content. As in the first round, I would like you to address all the issues raised by Ref#1 and Ref#2 in a point-by-point response, and explaining afterwards how you have updated the manuscript. Obviously I consider the concern on the calculation of the sulfur content and its undertainties the main issue still.

The report by Ref#1 got to me by email after urging the referee a few times. I included it below here:

*** REFEREE #1 ***
The authors did lots of efforts to correct the manuscript and most of the reviews' comments have been addressed. Hovewer, I have a serious concern about the calculation of the fuel sulfur content. They keep calculating the FSC from peak values instead of calculating the ratios of the areas under the S and CO2 time plot. This input a huge uncertainty in the method. I am, as an experimentalist, not fully convinced that I could believe in those ratios even if the calculated FSC is close to the laboratory given value.
For this reason, I can't recommend the paper for publication until the authors can prove that the peak value calculation is correct, for example, all the peak values in one measuremet give the same FSC with an acceptable statistics.
*** END REVIEW ***

Looking forward to a revised manuscript and response - the Associate Editor

Congratulation on your manuscript that is now accepted for publication in AMT, after you make the following clarifications and improvements:

(please use Author's response file to make changes, not revised manuscript)
Abstract, L12: *The* Waigaoqiao port ... was selected
Abstract, L17: "fuel samples were collected *at the same time* and sent"

Intro, L26: fuels --> fuel
P16, L9: "received" is a strange word, do you mean accepted?
P16, L17: The FSC limit was set to 0.1% (m/m) *in those areas* beginning in 2015.
P16, L28: cannot --> are not
P18, L21: a UAV to *simultaneously* measure ...
P23, L20: "can be obtained" --> is obtained
P23, L31: to calculate the *instantaneous* taio of SO2 and CO2
P24, L1: "We determined" --> We found
P24, L9: 'Regarding sensor uncertainty, the nonlinearity ,,,' would be a better way to start the sentence.
P25, L1: "a completely different uncertainty" --> please write that this is a systematic uncertainty.
P26, L10-11: "to the" can be removed
P26, L11: "peak values *were* not discussed
P26, L26: no more than 20 s apart?
P27, L9: "dividing line *between* plumes *with* high-sulfur and ..."
P30, L5: "not enough for reference" ... why? It looks like a reliable result to me, explain why it should be discareded.
P31, L14: "Poor quality data or rejected plumes" ... derive from what sort of situations? Please clarify this sentence.

P2, L24: "are not be"--> 'are not being' or 'are not'
In the acknowledgment, irt would be appropriate to thank the reviewer's comments which have improved the manuscript. I leave that up to you.

Otherwise, congratulation on the paper, which is accepted for publication in AMT.