This study uses continuously measured methane measurements at three tower locations in and around Hangzhou, China, to investigate temporal variations of emissions, especially from the treatment of waste.  The authors use the WRF-STILT (Weather Research and Forecasting – Stochastic Time-Inverted Lagrangian Transport) model combined with a Bayesian inversion framework to compare the data driven results with the prior emissions inventory.  They conclude that emissions have been overestimated for the city of Hangzhou and that there is a seasonality to the emissions that can only be explained by the waste treatment sector. 

This topic is very timely and important for understanding the influence of climate change on emissions of this high global warming potential pollutant, but several issues in this paper need clarification before publication. 

• Of the three sites, it appears that only one is in the city (Hangzhou), and one is on a relatively remote mountain (Damingshan). Is the third site, Linan, in a suburb or also background region, as stated on line 177?   If this is true, then there is only one site that is truly relevant to determining emissions from the city, since the other two are described as background sites.  However, background values are taken from much more remote sites.  There can be significant sources between the very remote sites and the urban region being studied, including large cities such as nearby Shanghai between Hangzhou and TAP and RYO, the latter being used almost always as background.
  
  The footprint for the Damingshan site is only slightly influenced by emissions in the urban core.
• What emissions did you use for the prior? It seems like you used the EDGAR v6.0 inventory for anthropogenic sources (except rice patties) and WetCHARTs for wetland emissions, including from rice patties. Please state explicitly how you calculated the prior – “a priori” is not mentioned in the WRF-STILT model setup section.
• A major assumption of the paper is that waste treatment is the dominant source of emissions and the other anthropogenic sources do not contribute to the seasonality of the observed CH4 measurements. What you show in Figure 4d is that waste treatment contributes most to the CH4 signal, but the other sources are also important.  Perhaps you can show a map of the locations of the anthropogenic sources – power plants (especially natural gas powered), landfills, wastewater treatment plants, distribution lines for natural gas, refineries, dairies, rice paddies – especially close to the urban center.  Enlarge the urban center to show locations. I am not convinced that you have enough information to discount the influence of other CH4 emissions sources or to characterize the sources in the urban center with only one site, especially when the reader does not know the sources in the region or the general seasonal wind patterns.  A measurement that you might consider for the future is ethane, since fossil-fuel-derived CH4 contains measurable C2H6, whereas biological sources (including waste treatment and wetlands) do not.  Seasonality due to fossil CH4 is observed in cities, even as far south as Los Angeles.  Is rice cultivation seasonal – should you expect some seasonality from this sector?
• This paper uses all of the diurnal cycle of the measurements. I definitely agree that emissions at night are not captured if only afternoon measurements are used, as is commonly done.  However, one reason most investigations don’t use the entire 24-hour record is that WRF does not do a good job with the transport parameters at night, specifically the planetary boundary layer height (PBLH).  It is very important to get this right for modeling to produce meaningful results.  You don’t show how your model performed for this critical parameter.  Can you show how the modelled PBLH compares with measurements, even if only a limited number of measurements are available? 

More detailed comments follow:

Abstract: mention the types of waste included in this study

Line 72: Out of curiosity, what are the top five anthropogenic sources of CH₄ in China?

Line 91: USEPA

Line 106: replace “absence” with “omission”

Lines 143-145: City-scale studies have not focused on waste treatment sources because there are many sources, as in Hangzhou.  Yadav et al. (2019; JGR Atmospheres) were able to see the effects of the closure of a landfill in the Los Angeles, CA area that was included in the prior inventory and not seen in the modelled results.

Pages 6-7:  In the description of the sites, please summarize the regional, seasonal wind patterns and any differences between the sites.

Lines 188-190: How frequently were standards run?  What uncertainty, including both precision and accuracy, did you assign for the measurements?

Line 238:  What does “fuel exploitation from coal, oil, and natural gas” include?  Extraction, transportation, refining, distribution, and combustion, or some subset of these?

Line 239: How and where is the energy for buildings generated? E.g., natural gas power plants in the suburbs, coal burned in the buildings, …? 

Lines 241-245: Is 0.5° high enough spatial resolution for your study region?

Line 287: reference for CCGCRV? Thoning et al., 1989, JGR 94, 8549-8565; https://gml.noaa.gov/ccgg/mbl/crvfit/crvfit.html

Line 295:  Is it meaningful to give an annual average when 1-2 months are missing data?

Line 295: replace “variations” with “trends”

Line 296: What are the “similar atmospheric transport processes?” Summarize seasonal wind direction and speed patterns.

Line 309: replace “YON” with “TAP”

Line 310: replace “temporal” with “spatial”

Lines 320-321: Figure 3 does not show significant differences in the size of the footprints at the different sites.  You might want to expand the scale to show this.

Lines 323-326:  Cities shown significant diurnal variation in PBLH.

Line 345:  Not sure what you mean by “amplitudes” here – amplitude of the seasonal variations?  I don’t see obvious differences.  The absolute average abundances are different.

Line 348: The simulated data for Linan actually approximate the observations very well!

Line 364: It is very much to be expected that the Hangzhou site is more influenced by local emissions since it is in the urban core.  What are the major emitters within 5-10 km of the site?

Lines 366-368: How did you show that the Linan and Damingshan sites are influenced by emissions from a much larger region?  The footprints don’t indicate this.

Lines 375-378:  Can you give a reference for the statement that waste treatment emissions are larger during the daytime than at night? 

Line 420:  Emissions from waste treatment dominated the total CH4 and the seasonal pattern, but you do show significant seasonal variations for the other anthropogenic sources in Figure 7a.  Can you split those up at all?  Can you say anything about the natural gas distribution infrastructure – more leaks in winter than summer, …?

Line 479:  Where are these values of SFs shown?  They are not from Table S2.

Figures:  In general, please improve the resolution of the figures.  It is very difficult to impossible to read the small text, even when expanding the figures on the screen.

Figure 1:  What are the divisions within Hangzhou City?

Figure 2:  Use the same color schemes on all figures and parts within figures for the same sites.

Figure 3:  Replace “lg” with “log.”  Are the waste treatment CH4 emissions in panel (e) also from EDGAR v6.0?

Figure 4:  Need higher resolution graphics, especially for panel (d) and (e).  The note at the end of the caption may be incorrect.  In panel (e), is it true that the blue color for the bar charts include all of Zhejiang, including Hangzhou?  Do the blue regions in the pie charts represent Zhejiang minus Hangzhou?

Figure 8:  What region is this figure describing?

Summary:

Review of “Global warming will largely increase CH4 emissions from waste treatment: insight from the first city scale CH4 concentration observation network in Hangzhou city, China” by Hu et al. 2022 for Atmospheric Chemistry and Physics.

Hu et al. use atmospheric observations and modelling tools (lagrangian) to estimate methane emissions from an important Megacity. Relying on EDGAR V6.0 they analyse the sectorial contribution to atmospheric CH4 enhancements and then optimize fluxes using a Bayesian framework. The results indicate an overestimate of local emissions by EDGAR V6.0. The seasonal bias between a priori and a posteriori fluxes is attribute to waste sector emissions and a temperature sensitivity is calculated. Using IPCC scenarios the authors than quantify the temperature-specific component of the waste sector emission factor changes for the coming decades.

Overall, the paper is clear and can be followed easily. However, the study lacks some critical assessments around the choice of EDGAR V6.0 and the implications of that choice. Furthermore, the study should be clearer on the fact that the suggested effect could be fully compensated by other parameters affecting the waste sector emission factor. It also would be useful to specify that a single city study should not be scaled globally, but that is surely has an important message for CH4 emissions in Chinese Megacities. Given the importance of this region for future climate change this study is surely of interest to the wider scientific community and especially ACP readers. After addressing the general and specific comments this manuscript would appear suitable for publication.

General comments:

• The title implies a global impact; however, it only provides results for one urban region. Also, country-specific waste management strategies (e.g. highly localized waste separation stations) call into question how much the results from this region can be extrapolated beyond Chinese Megacities.
• This study only assess the influence of temperature on the emission factor for waste although previous work has shown the importance of other meteorological parameters such as atmospheric pressure changes, water content and management strategies. It is unclear that local climate change would not also affect these parameters as well. This could reduce or strengthen the suggest increase in emissions. The authors also do not discuss how relevant temperature is as a parameter when compared to the others mentioned above.
• This study uses EDGAR CH4 without critically assessing its limitations. EDGAR is coarse resolution 0.1x0.1 degree for urban studies and was shown to have biases in some high-density urban areas. e.g. Vogel et al. 2012 (https://doi.org/10.1080/1943815X.2012.691884). Why do you rely solely on EDGAR and why do you believe its spatial disaggregation to be correct?

Specific comments:

Line 36 and line 75:

Please provide a source for the claim that waste emissions contribute over 50% of CH4 emissions at city-scale. For which cities and regions does this apply?

Also please provide evidence that most household waste is located in cities and not in landfills outside the cities. In some regions landfills are located outside the city limits.

Line 75: Please add a critical discussion of the importance of active and closed landfills, waste water systems and household waste in residential areas. Recent work has shown that waste water can be a significant source at urban scale. E.g. Williams et al. 2022 (https://doi.org/10.1021/acs.est.2c06254).

Line 79-83: this review fails to mention the critical impact of atmospheric pressure changes on emissions. As shown by e.g. Kissas et al. 2022 (https://www.sciencedirect.com/science/article/pii/S0956053X21006310) and references therein. Emissions can be increased by orders of magnitude due to this effect.

Line 137: Given the strong influence from barometric pressure on landfill CH4 emissions it is critical to discuss the clear-sky bias of satellites here. Satellite observations are too sparse to be up-scaled to estimate annual totals.

Line 165: The described study can only assess the temperature component of the EF changes but neglects pressure changes as well as all the other factors outlined in line 79-83, e.g. water content oxidation efficiency, landfill gas collection.

Line 273-282: How where these prior uncertainties calculated/determined? They seem to strongly differ from Solazzo et al. 2021 (https://doi.org/10.5194/acp-21-5655-2021)

Line 287: Please provide a reference for the CCGCRV fitting method.

Line 336: Please provide a reference for the emissions from waste separation stations.

Line 344: Please quantify the consistency of the temporal patterns by providing Pearson’s r values for all time series shown in Figure 4.

Line 357: The finding that waste dominates emissions here strongly relies on the spatial patterns of EDGER being correct also previous work has shown limitations of EDGAR to capture CH4 emission patterns in urban areas, see e.g. Pak et al. 2021 (https://doi.org/10.1016/j.atmosenv.2021.118319)

Line 424: How much do daytime and all-day average concentrations differ at the Hangzhou site?

Line 425: Here you are assuming strong changes in waste-related methane emissions, without any references, while EDGAR V6.0, which you used as a prior assumes constant emissions.

Line 482: Your study nicely shows the temporal bias of EDGAR V6.0, what about a potential spatial or sectorial bias?

Line 493-496: Have you investigated the correlation of monthly CH4 emission changes with soil water content, precipitation or other parameters you listed in line 79-83?

Line 508: Please clarify that this is only the temperature component of the EF and does assume no changes in technology or other meteorological variables.

Line 529: Agreed that this is beyond the scope, but it seems prudent to mention that changes in management and technology can have a strong influence emissions in the future.

Line 570: What was the predicted emission change due to changes in activity data and management in the cited studies? How does you reported temperature sensitivity compare?

Line 606: Large parts of the conclusion sections are actually a summary.