This manuscript presents a  new technique to calculate the age of air from measurements of trace gases.  Having better (more accurate) constraints of age of air is a very important topic, especially given the possibility the stratospheric circulation and age of air are changing due to  climate change and/or ozone depletion and recovery.  The paper is well written (although it did take me a while to understand the details of the method) and I think publishable with only some minor revisions (on the results section).

Specific Comments:

Line 470: The authors state that “the ‘wing plot’ of 𝛤 in the 18-20km altitude range” has been extensively used for model comparisons. So why is this plot not shown for the new calculations? 

Fig 6 This is not apples to apples comparison given the different reference location of the age calculations. Can you subtract a characteristic surface to tropical tropopause age (e.g. 3 months quoted on line 489) to make this a cleaner comparison.

Line 610  “A follow up study will examine the differences between the 1990s and 2020s ages in detail”.  I think there needs to be at least a brief discussion of this issue. You show figures for the two periods, so if you don’t comment on this you will leave it up to the reader to draw their own conclusions. Better that you say what your data shows.  This doesn’t prevent a more detailed analysis. 

Fig  10 Can you identify the season of DCOTSS and SABRE, so can see straight away which of the air core they should be compared with. I would also suggest removing the annual mean for air core, as figure is cluttered. .. 

Data Availability: The data is available but I am not idlsave files are the best way to archive the data. Why not save as ascii or netcdf files that can be easily read by any software.   I think there will be a lot of interest in these data (e.g. for model comparison) and I think it is in the authors interest to present the data in the simplest form for others to use. 

Ray et al. present a new technique to calculate both the first (mean age of air) and second (ratio of moments, width of age spectra) moments of age by using in-situ measurements of multiple long-lived tracers. This method uses transit times from the Earth’s surface, instead of the commonly used assumption of the tropical tropopause, to any location in the atmosphere. Having an accurate constraint on stratospheric age is especially important to better understand possible shifts in the Brewer-Dobson Circulation in response to a changing climate and its further implications on radiation, chemistry and dynamics.

The main conclusions of this study are: (1) This work presents age derivations using simultaneous in-situ observations of SF₆ and CO₂ from both 1990s and 2020s, which could help address ongoing questions in the field. (2) These derivations were done in the upper troposphere and throughout the stratosphere. In addition, (3) ratio of moments derived from in-situ observations agree well with recently published results from chemistry-climate model output, and lastly, (4) results from records spanning multiple decades will allow for future age comparisons.

Overall, this paper was well-written and concise. The research question(s) and descriptions addressing the science goals were clearly stated. However, the methods section could be written more clearly, which will be further explained in the comments section. Given that this manuscript introduces new methods to better constrain stratospheric age, I believe this manuscript is publishable and of interest to prospective readers in the field and ACP. Below are some suggestions for authors to consider for minor revisions before submission. I am very much looking forward to the follow-up study.  

Technical Comments:

Equation 2: Past literature used the same assumption that “age spectra are assumed to have an inverse Gaussian functional form.” Given that gamma, spectrum width, and ROM are provided, are there any other assumptions made when deriving the result? 

Line 195: “The fraction f has an age dependence with f(t’ <t’i)=0, f(t’ >t’f)=1 and an exponential form between t’i and t’f, which are the transition ages between the latitudinally varying and purely tropical source regions.”

I wanted to be clear about the notation described in this sentence:  t’i is the transit times of the latitudinally varying source region and t’f is the purely tropical source region? 

Specific Comments:

Line 156: Including a brief one or two sentence description of what the age spectrum is (after introducing Equation 2) would be useful. In reference to this study, it is a mass weighted function generated by different pathways (or colored lines) to the sample region. This would help bridge Equations 1 & 2 and Figure 1. 

Line 233-234: Adding a brief sentence explaining the reasoning of the latitude-season gradients in CO₂ (and not SF₆) would provide more context when presenting results in Figure 2.

Line 340: Include a reference to Figure S5 in the caption of Figure 3 so the reader can refer back to labels of the dashed lines in supplemental.

Line 470: There is a reference to a ‘wing plot’ in the manuscript and Age-N₂O relationships, where there are Age-N₂O results from this study. Given that there are zonal mean ages as a function of height and latitude derived from both flight and balloon data in this work and that CO₂-age and SF₆-age have been “been used extensively” with the wing plots, why was this comparison not done in this study? 

Figure 6: Which latitude ranges were used for both - balloon and aircraft - relationships from 1) this study, 2) Andrews et al. and 3) Volk et al., and are they the same? Tracer interrelationships vary with latitude as a result of the overturning circulation, and in the case of SF₆, mesospheric influence/ sinks. For example, one would expect the tropics Age-N₂O relationship to have younger air at almost all normalized N₂O bins compared to that of Age-N₂O at 60N where air has been in the stratosphere longer and has been subjected to mixing/ sinks. This is important to include in both the text and figure description and for comparing Age-N₂O relationships from different studies. 

Figure 7 and 8: the Age color bar ranges are slightly different for Figures 7a and 8a. The scaling ranges and pixel spacing should be included in the figure descriptions. (This can also be applied in description of Figure 11 as well)

Section 4.2: Everything was well-written and concise, although a bit short. I see that “a follow-up study will examine the differences between the 1990s and 2020s ages in detail,” but I think stating what the results show for the reader will not take away from the follow-up study. For example: there is a seasonal shift in AirCore relationships- why?

Figure 10: Similar to Figure 6, what latitude ranges were used for the Age-N₂O relationships and are they similar enough to be comparable. Including the latitude ranges in the figure descriptions would be useful.  

This is an interesting study in determining several statistical terms related to age-of-air, including some terms that have not previously been shown in other studies.  It also shows nicely how the use of two tracers can be used to better constrain the age-of-air.  I apologize for my lack of expertise in evaluating this study, but hopefully I can provide suggestions that help to clarify some concepts for a wider audience.

Please explain why, in equation (3), G is partitioned into a tropical surface source and a latitudinally varying surface source (which seems to include the tropics).  I certainly understand that it is useful to show the tropical term (e.g. the dashed line in Figure 2), but couldn’t one simply determined this by letting gV indicate the complete latitudinally varying surface source and then summing up gV from 30S-30N to find the tropical contribution?  Maybe this is all explained in a previous study, but some explanation here would be helpful.

The term on the right-hand side of (3) is “non-normalized”, while the terms on the left-hand side are “normalized”.  I am not sure what exactly this means, but I do not understand how adding two non-normalized terms can produce a normalized term.

The notation y_oTR in (7) is confusing.  I don’t think this is used elsewhere, and the presence of a surface source latitude parameter on one side of this equation and not on the other seems problematic.

Figure 4 – the polaris and solve colors are very similar and difficult to distinguish.  Also, please give some indication of the altitudes or pressure that are being shown here.  Of particular interest is how much of this data is in the stratosphere?

Line 405 – I think this is the first time the subscript ‘s’ is used in the text.  If so, please define it here.  I did eventually find the definition in the Figure 4 caption.

Paragraph starting at line 430 – I think the authors are saying that there was an SF6 measurement problem during SOLVE.  If this is correct please state so clearly.  If not, please clarify the paragraph to explain the problem.