Primary comments:

This manuscript describes important emission information on key Montreal Protocol controlled and climate related gases.  It should be fully published.

The methods used to determine the emissions are widely accepted and appropriate for this study. 

My primary concern is one that can be remedied with some careful attention to the discussion to make the reading of the flow a bit better.  The discussion is rather disjointed, with a bunch of factual statements of decisions made for the analysis.  There is little discussion of why the decisions were made for many of the assumptions.  Such discussion would help to make the logic flow a bit better.   One such example is at line 220.  An assumption is mentioned with no statement of why.  The same is true for the discussion of the results.  All the discussion related to the figures (which I mostly like) does not naturally flow.  For example, a better discussion of the comparison of HFC-23 vs. HCFC-22 emissions and what that REALLY tells us would be helpful.  A useful discussion point that this paper COULD address is the distribution of the inferred emissions relative to what is seen for CFC-11.

Other areas that could be described more clearly are how the priors are really constructed relative to the TEAP information regarding the location of the probable production facilities.  It was not clear to me how the initial emissions were distributed among the various facilities for the priors, and how they changed.

A few minor points are below…

Abstract:  How are the known production locations used in the analysis?

Line 39: the term “basket” is a bit informal.  I’d change it to “group”….though did Kyoto really regulate?

The term “top-down” is loosely defined in the abstract, but not in the body of the paper.  I suggest doing so, and with a bit more detail than what is done the abstract. 

Figure 3…so that these maps can be shown without the caption, I’d put the years of the analysis in the upper left corner of each map.

Review of Park et al., on HFC-23 emissions

This manuscript describes new measurements and interpretation that are very important and novel for the international Montreal Protocol community in particular.  These results have been eagerly anticipated for some time, given the paper of Stanley et al on global HFC-23 emission changes a few years ago.  The current manuscript provides convincing evidence suggesting that the cause of the continued elevated global emissions of HFC-23 arise in large part from eastern China.  It will be read and studied extensively by scientists and the much broader audience that includes international policymakers.  As such, it needs to be as clear as possible.  At this point, I don’t have substantive issues with the scientific approach or methodologies, but I do find that the organization and emphasis in certain sections make the main messages opaque and not clearly conveyed.  With close attention to these issues in a revision, I believe this paper will be an extremely important contribution to the field.

For example, the title does not accurately reflect the findings of the paper: Emissions from East Asia as a whole are not presented in a table or figure, making it difficult to assess if the sum of emissions from this region have “continuously increased” throughout 2008-2019 as is asserted in the Title (and abstract, see line 26). Second, the emission estimates for smaller regions (E China, S. Korea, N. Korea, and W. Japan) provided in Figure 2 and Table 1 do not suggest a “continuous increase” in their sum over 2008 to 2019. Instead, I see emissions that vary between 4 and 10 Gg/yr from time to time depending on the effectiveness of the controls in place. Suggestion: revise title to better reflect the results and consider if showing the total from these 4 regions in Figure 2 would be useful. (note also that East Asia is a well-defined region that doesn’t correspond well to the region for which measurements at Gosan have sensitivity to; consider a different label.)

The lack of any map indicating the spatial sensitivity extent is unfortunate, particularly since Figures 3 and 4 suggests a strong co-location of HCFC22 production plants with HFC-23 emissions in some instances but not all. Is this because the knowns plants in southern China are not well sensed by measurements at Gosan, or because emissions from these plants were indeed small?  A map of surface sensitivity would help address this issue, even if it appeared only in the Supplement.

Further on figures 3 and 4, some indication of the magnitude of HCFC-22 production by plant (different sized red dots?) would strengthen your arguments suggesting HCFC-22 plants are co-located with HFC-23 emission. Aren’t these magnitudes available by production plant at least during the CDM period? It appears as if the distribution of production plants is the same in all three panels of Figure 3 despite the time changes in HCFC-22 production plants in China through this period.  Is it possible to remedy this issue and show only plants reportedly producing during the different periods?

Discussion related to HCFC-22 is confusing and hard to follow.  Some up-front text is needed in sections 3.3. and 3.4 that makes clear why production of HCFC-22 from eastern China is needed with respect to the main points of the paper, HFC-23 emissions, and how you will go about deriving those quantities. Maybe some explanatory text that is in section 3.5 can be moved forward. Consider moving section 3.4 to the Supplement, as it describes a fairly straightforward method for estimating HCFC-22 production in years where that info is lacking.

Conclusion needs a significant overhaul.  Issues not discussed in the paper are mentioned at length (unreported HCFC-22 production?) as are potentialities that are in the end discounted.  These possibilities can be mentioned, discussed in the paper or Supplement, but they leave this reader very confused when presented only in the conclusion.  The Conclusion focus should be on clear statements of the main messages and findings, otherwise they will be lost to most readers, which would be very unfortunate for this most important result.

Unclear wording in the abstract:

Line 22: on use of the word mandated.  The text (lines 55-64) indicates that the Chinese NDRC was a plan for reducing emissions, and the HPPMP is a plan with reporting of emission estimates.  None of these comes across as a “mandate”, so I’d argue the use of the word is inaccurate unless related laws or mandates were put in place in China that are not currently mentioned in the manuscript. 

Line 23- Likely sources of HFC-23 emission are well known (leakage or release of by-production); consider changing “sources of” to “regions responsible for”.

Lines 28-29, are these quantities cumulative for the 5-yr period or annual rates?

Lines 30-33.  This is a confusing sentence attempting to convey too many points in one sentence.  I recommend a rewrite.

Other details:

The TEAP data displayed in Figure 7 are puzzling given the expectation that production should have been capped in 2013 and declined thereafter.  Is this total production for all uses, dispersive and feedstock-related? Some explanation is needed but lacking.

Lines 210-215, without a full analysis of eastern Chinese HCFC-22 results, banks, emissions from banks, etc., it is inappropriate to suggest or “imply that HCFC-22 production and consumption in the region have been increasing” despite MP controls mandating a freeze and decline in these quantities destined for dispersive use in and after 2013.   These messages need clarifying.

Citing WMO reports and other assessment reports is appropriately done by using chapter lead authors for the appropriate chapter.

Lines 49-52, be careful with tense, as Kigali Amendment controls may be “in place” now, but controls on production wont apply for many years for many countries.

Lines 53-55. Citations appear mid-sentence so support the statement on HCFC-22 production, but the assertion that China is the largest HFC-23 emitter needs support (a citation) given that HFC-23 emission doesn’t necessarily track HCFC-22 production given the possibility of destruction.

Lines 69-72—this argument doesn’t hold unless we understand the contributions of these countries, which isn’t mentioned wrt this point. Logic needs improving.

Figure 1 and Figure 5, consider somehow indicating periods where the instrument was not operational to prevent the conclusion that there were no measured enhancements during late 2016 to mid-2017 and other times. Also, indicate in the caption if measurements made at Mace Head and Cape Grim represent all results or only those during non-polluted periods (the latter would be most appropriate to show).

Figure 2, indicate that the values plotted are annual averages.  How were the data gaps handled when deriving annual means and their uncertainties? This needs a clear description in the methods.  The uncertainties for 2017 don’t look much different than other years, despite there only being approximately only half as many measurements available for interpretation during 2017.

Figure 3, axis label units aren’t mentioned. Consider including a table next to the panels or within them providing mean emission totals by region.

Figure 7 has typos and duplicate text that needs fixing.

Figure 8, hatched regions need defining in the caption, particularly the yellow region, even if they are described in the main text.

Figure 9, assumptions (e.g., emission/production fraction) were made to derive the HFC23 emissions from Chinese HCFC-22 production information shown in this figure, and these assumptions need mentioning in the caption along with a description of the values that were chosen and how they were derived. (Don't make the reader have to search for important details in the main text.)