Chen et al. reported a comprehensive analysis of the abrupt increase of surface ozone observed in the North China Plain on the night of July 31 2021. The authors employed various datasets and tools in their analysis, including (1) the air pollutant observational data from the national monitoring stations, (2) high-frequency ground-based observational data during a campaign in this region, (3) vertical profile observation of ozone, (4) radiosonde meteorological data, (5) reanalysis data, (6) regional meteorological model, and (7) back-trajectory model. They found out that the sudden increase of surface ozone was not due to horizontal transport; instead, the dying typhoon and a local mesoscale convective system brought the high-ozone/low-CO air to the lower troposphere and even the ground surface. The authors did a commendable job by applying various tools to analyze a unique atmospheric phenomenon that has implications for air quality management. I support acceptance of the paper once the following concerns are addressed. 

General comment:

1.    The authors emphasized that this paper is about the effect of a “dying” typhoon and “shallow” convection in the title, abstract, and many places in the main texts. While this is correct based on the authors’ analysis, I wonder would it be better to generalize the mechanism? If I understand correctly, a typhoon (dying or not) likely causes stratospheric ozone that brings the stratospheric ozone to the upper and middle troposphere, and a follow-up mesoscale convective system (shallow or not) would then transport the high-ozone/low-CO air further down to the lower troposphere. 

2.    Meng et al. (2022) reported a very similar process (an anomaly in surface ozone due to the passing typhoon) in the same region (NCP). I am aware that the authors of the present work started their analysis before this recent paper (Meng et al., 2022) was published, but it would be beneficial to the readers if the authors could add some relevant discussions. 

3.    The designs of the manuscript and figures require some improvements:

(1)    Section 2.3, please add a figure (at least in the supplement) showing the domain setting of the WRF simulation. I would like to know whether the inner domain covers the region with strong vertical transport. 

(2)    In The paragraph starting at line 174, a table listing all WRF model parameterizations and setups will be very clear. 

(3)    Line 191, why not describe the setup of FLEXPART-WRF here? How many simulations? What is the location (lat, long, and altitude) of particle release? How many days for each simulation?

(4)    Line 194-197, these two sentences seem out of scope. Consider removing. I don’t think it is necessary to mention the increase in ozone in the past decade in this region. 

(5)    Line 208-210, this sentence is a bit odd too. Consider removing. 

(6)    Figure 2, showing the “departure from the 10-day averaged ozone”, is not a good choice to demonstrate the sudden increase of surface ozone. Instead, I believe Figure S2 (with the average diurnal pattern of ozone) is a much better option for showing the anomaly of surface ozone in this region. Similarly, I would recommend drawing a similar figure of CO to replace the original Figure 3. 

(7)    If possible, Figure 4 should also be replaced with one similar to the original Figure S2. In fact, in line 249, the authors stated that “Compared with the normal nighttime ozone concentrations (an average of 36.6 ppbv), the magnitudes of surface ozone surge due to stratospheric intrusions were approximately 40-50 ppbv”. If the “normal nighttime ozone concentrations” were already shown in Figure 4, i.e., with the average diurnal pattern, readers would easily see the “departure” of ozone/CO from their “normal nighttime concentration”. Also, I suggest only including the “hour” in the X-axis should be informative enough. 

(8)    Figure 5, what do the positive/negative vertical velocities represent? Positive values (blue) are winds going down to the surface? Or the other way around? Please clarify in the figure caption. 

(9)    In Figure 7, similarly, I don’t understand why an average level of ozone between surface and 700hPa is used as a baseline. Shouldn’t the baseline be the 10-day average vertical profile? With the current figure, the readers must be puzzled why the surface ozone concentrations on July 31 and Aug 1 are lower than the average, while the other sections repeatedly show that the surface ozone concentrations in NCP are larger than the average. 

(10)    Line 360-364, this information should be moved before mentioning Bow-echoes. 

(11)    Figure 8, why not show the ozone data at all sites at all times and use a colour scale that covers 10 to 100 ppbv? With the current layout, there is no way to tell how much ozone is increased from 2100LST to 0100LST at stations like JN/BZ. It could just be 1 ppbv of increase (if ozone at JN/BZ were 79 ppbv at 2100LST and 80 ppbv at 0100LST) or >80 ppbv of increase (if ozone were <1 ppbv at 2100LST and >80ppbv at 0100LST). 

Specific comment:

4.    line 37, this line reads like both “water vapour” and “carbon monoxide” are primarily emitted from combustion processes, while only CO is. Consider revising it. 

5.    Line 75, what problems “require in-depth investigation”?

6.    Line 199, is this 36.6 ppbv calculated in this study or from a previous study?

7.    Line 228-230, this sentence sounds important. Any figures/data to support it?

8.    Line 237, “filed” should be “field”. I have spotted a few more typos. Please check through the manuscript. 

9.    Line 239, it is >30% increase from 45 to 60 ppbv. I would not call it “slightly higher”. 

10.    Line 250-251, an increase in ozone from ~36 ppbv to ~80 ppbv is a large enhancement, but this level of ozone (80 ppbv) should be very common in this region. I suggest toning down the phrase “great threats”. 

11.    Line 262, somehow the authors missed “anthropogenic emission”?

12.    Line 266-267, this is probably true, but it will be better if the evidence is presented. 

13.    Line 271, “indicate” should be “indicated”. 

14.    Line 417, “the analysed at detail” should be “be analysed in detail”. 

Reference:

Meng, K., Zhao, T., Xu, X., Hu, Y., Zhao, Y., Zhang, L., Pang, Y., Ma, X., Bai, Y., Zhao, Y. and Zhen, S., 2022. Anomalous surface O3 changes in North China Plain during the northwestward movement of a landing typhoon. Science of The Total Environment, p.153196.

General comment:

This is an interesting case study of significant stratospheric ozone transport down to the Earth’s surface by a dying typhoon, affecting local and regional air quality. Several observational data and modeling tools are applied to analyse/confirm the downward transport of ozone with stratospheric origin. Overall, this is a well-designed study which is relatively easy to follow. Such events of direct ozone transport have implications for air quality, contributing in ozone standards exceedances. The paper fits well within the scope of ACP and I recommend publication after the following comments are addressed.

Comments:

The only thing I found missing in the analysis are humidity measurements near the surface from ground-based meteorological stations. This would offer the temporal variability of humidity near the surface, likely supporting the case that the observed ozone increases are of stratospheric origin. Is this feasible?

L41-42: A more scientific definition of the tropopause is rather necessary here.

L59: Some additional references of direct SI impact on surface ozone concentrations are needed here like Akritidis et al. (2010), Dreessen (2017), and Knowland et al. (2017).

L60: Meul et al. (2018) and Akritidis et al. (2019) also suggested an increase of STT in a future climate.

L74-76: Maybe some references are needed here. Which are the fundamental problems requiring in-depth investigation?

L78-79: “the stratospheric ozone-rich airmass was transported downward to the surface”. This is Introduction and such statements are not yet supported. I suggest removing or rephrase.

L137-138: “along with ground-based automatic weather station observations”: Which exactly? Do you mean the radar data? If not, are these shown anywhere in the paper?

L298: Why is the PV = 2.5 pvu isosurface selected for tropopause representations? Usually, 2 and 1.5 pvu are used. A reference/rationale for that selection would be helpful.

Figure 4: Vertical lines delimiting the O3 increase and CO decrease (similar to Figure 3) would be helpful here.

Figure 7: Here the 10-day average "between the surface and 700 hPa" is used as baseline for the O3 profiles. What is the rationale behind this selection (between the surface and 700 hPa)? As O3 increases in general with height, I think it is likely that the positive (red) departures in the troposphere are partially normal, masking the STT effect.

L407-408: “when the stratospheric airmass had reached the surface”. Where does this arise from? If it’s based on a Figure, please include it in parentheses e.g. (see Fig. 2). 

Figure 11: What do the magenta contour line labels describe? Is this percentage (%) of total number release? Please include this information in the respective caption.

Technical comments

L16: “on 31 July 1 to 6:00” delete “1”

L25: Please move FLEXPART and WRF full names in the previous line where are referred.

L34: and the troposphere

L34: atmospheric composition

L56: STT usually stands for Stratosphere-to-Troposphere Transport which is not the case here. Please remove STT or change the phrase.

L76: origin

L80: Compare with -> Compared to

L87: Since here you are referring to a specific study I suggest to directly mention it. “Chen et al. (2021) evaluating the impacts of typhoons on tropospheric ozone showed..”.

L134: and they show -> showing

L151: the stratospheric dryness -> dry stratospheric air

L199: Please include nighttime definition (hour range).

L228: “confirms” is somehow strong here, I suggest “supports the case”

L357: lasting->lasted

L458: “occurring at nigh”. As this is the beginning of the conclusions, the date of occurrence should be also stated.

L459-460: “while the impacts of stratospheric intrusions on surface ozone are relatively less studied”. This is somehow not connected to the previous part of the sentence, thus, I suggest to split in two sentences.

L512-513: “which underscores the necessity of considering these processes into the global model of atmospheric chemistry.” This phrase is somehow strange. What do you mean by global model of atmospheric chemistry? Please rephrase.

References

Akritidis, D., Zanis, P., Pytharoulis, I. et al. A deep stratospheric intrusion event down to the earth’s surface of the megacity of Athens. Meteorol Atmos Phys 109, 9–18 (2010). https://doi.org/10.1007/s00703-010-0096-6

Akritidis, D., Pozzer, A., and Zanis, P.: On the impact of future climate change on tropopause folds and tropospheric ozone, Atmos. Chem. Phys., 19, 14387–14401, https://doi.org/10.5194/acp-19-14387-2019, 2019. 

Dreessen, J. (2019). A Sea Level Stratospheric Ozone Intrusion Event Induced within a Thunderstorm Gust Front, Bulletin of the American Meteorological Society, 100(7), 1259-1275.

Knowland, K. E., Ott, L. E., Duncan, B. N., & Wargan, K. (2017). Stratospheric intrusion-influenced ozone air quality exceedances investigated in the NASA MERRA-2 reanalysis. Geophysical Research Letters, 44, 10,691– 10,701. https://doi.org/10.1002/2017GL074532

Meul, S., Langematz, U., Kröger, P., Oberländer-Hayn, S., and Jöckel, P.: Future changes in the stratosphere-to-troposphere ozone mass flux and the contribution from climate change and ozone recovery, Atmos. Chem. Phys., 18, 7721–7738, https://doi.org/10.5194/acp-18-7721-2018, 2018. 

This paper discussed the downward transport of stratospheric ozone to the troposphere as well as down to the surface through a combined effect of a dying typhoon In-fa and shallow local mesoscale convective system (MCS). They analyzed the ozone and CO concentration, meteorological reanalysis data, radiosonde data, and FLEXPART-WRF simulation. The downward transport of stratospheric ozone-rich air to the surface will degrade surface air quality and affect human health. Overall, the paper is good. It studied an important topic, used various observations. However, it still has some major weak points.

General comments:

• Because the downward transport was caused by typhoon In-fa, it would be nice to have a brief introduction of typhoon In-fa in section 2. Please include a plot showing the development of the typhoon In-fa (e.g., radar reflectivity for different times), and a plot showing the path of the typhoon In-fa. This will help the reader to understand the discussion of the second part.
• Lightning-generated NOx could also increase downwind ozone level. The paper did not prove that the ozone increase is not caused by LNOx generated by previous storms.
• In the paper, they calculate the 10-day mean O3/CO as the baseline. However, the 10-day mean included the days affected by typhoon In-fa. Therefore, is hard to tell what’s the normal condition. It might be better to use the 10-day mean before the typhoon period as the baseline.
• In this paper, they run WRF with tracer instead of using WRF-Chem. However, LNOx and other ozone precursors could also affect the results. Please explain why you choose not to use WRF-Chem or other chemistry models. The ozone production is not significant in the first few hours, however, previous studies found that there would be a great ozone increase in the downwind side on the next day. If you insist to use WRF with tracers, you need to convince the reader that your results would not be affected by any ozone chemistry reactions.

Specific comments:

• Line 44, here are some references for deep convective transport of surface pollution and ozone precursors to upper troposphere:

Dickerson, et al. (1987). Thunderstorms: an important mechanism in the transport of air pollutants, Science.

Pickering, et al. (1991). Photochemical ozone production in tropical squall line convection during NASA Global Tropospheric Experiment/Amazon Boundary Layer Experiment 2A. J. Geophys. Res.

Pickering, et al. (1992). Ozone production potential following convective redistribution of biomass burning emissions. J Atmos Chem.

Li, et al. (2017). Evaluation of deep convective transport in storms from different convective regimes during the DC3 field campaign using WRF-Chem with lightning data assimilation. J. Geophys. Res. Atmos.

• Line 110, it would be nice to have a brief introduction of typhoon In-fa in section 2. Please include a plot showing the development of the typhoon In-fa (e.g., radar reflectivity for different times), and a plot showing the path of the typhoon In-fa. This will help the reader to understand the discussion of the second part.
• Line 166, please include the reference for WRF.
• Line 167, please add a figure showing the location of each domain in supporting information.
• Line 185, why do you choose WRF instead of WRF-Chem? See general comments 4.
• Figure 2, see general comments 3.
• Figure 2, please add a map showing the storm location during the ozone surge period.
• Line 226, CO is also an important tracer for deep convective transport. Please include references here. “CO is offen….(add references)”
• Line 250, please mention the ozone exceedance level, and compare the observed ozone level to the ozone exceedance level. Otherwise, you cannot conclude that “which can pose great threats to human health…”
• Line 268, please explain more about “no influence from ozone precursors from biomass burning or LNOx”. See general comments 2.
• Figure 5, please label time in each plot.
• Line 400, could you add a forward trajectory experiment of stratosphere tracers?

Review of Transport of large stratospheric ozone to the surface by a dying typhoon and shallow convection

By Zhixiong Chen, Jane Liu, Xiushu Qie, Xugeng Cheng, Yukun Shen, Mengmiao Yang,  and Xiangke Liu

This is an interesting paper, addressing an important topic, and appropriate for publication in ACP. At times the description of the dynamics is hard to follow, and so I suggest some attention be given to making the arguments simple and clear.

The paper mainly falls down, I think, in presenting the meteorology of the convective system that is alleged to be responsible for bringing the ozone down to the surface. This may simply be because the authors are meteorologists by training, and forget that most ACP readers are not. Please explain more! A prime example is Figure 11, which purports to demonstrate the downward transport: “We performed cross-section analyses of the bow-echo MCS (Fig. 11b-c), and the results clearly show a rearward pathway through which the stratospheric ozone-rich airmass was transported to the surface by the rear inflows descending from stratiform clouds to the leading convective line.” Maybe they clearly show that to the authors, but unfortunately not to this reader. Is the salient point that the “tracers” are now mostly below 3.6 km? Or that some of the wind vectors are pointing down? Is the reader supposed to be able to see “the rear inflows descending from stratiform clouds to the leading convective line”? Remember that many of your readers won't know where the "leading convective line" is found. 

Minor points:

The presentation and grammar need some good editing. It is sometimes difficult to understand what the authors are trying to say.

Abstract: This is a bit long, and some of it reads like an introduction to the paper, rather than a brief summary of new results. At the least, the last sentences of each paragraph (lines 20-21 and 30-32) should be moved or deleted.

Lines 66, 449, 450: “…wrapped around the anvil”. An “anvil” is a block of iron that a blacksmith hammers upon. “Rearward anvil” and “forward anvil” are meteorologist’s slang. Most readers will know that an anvil-shaped cloud is often associated with a thunderstorm, but no more. Please be clear about what you are describing, and why.

Lines 111-116: The instrumentation should be identified, and/or the uncertainty and detection limits cited.

Lines 189-190:  I think this is saying that FLEXPART-WRF used the 3-km resolution output of WRF-ARW, but it isn’t really clear.

Lines 208-201: “It is a common practice to use 25th percentile of ozone concentration distributions as a background value (e.g., Parrington et al., 2013), which yields an even severer ozone enhancement in the surface.”  I think the authors are trying to suggest that the ozone amount is more significant because it is all transported from elsewhere, and so could be measured against some “background” value (arbitrarily defined). This is a dubious comparison that will only serve to confuse the reader. Delete.

Line 226: After “…not reduced in Qingdao and Weihai” I suggest adding “…which were outside of the path of influence of the MCS, as noted in the preceding paragraph.”

Line 230:  I suggest referring to Figure 1 here.

Line 281: Bohai Bay is not indicated in Figure 1.

Lines 287-288: Is that the blue areas? In other words, does a positive vertical velocity in Pa imply downward motion? This is not clear.

Line 307, Figure 6: Why use dewpoint depression? This metric will be unfamiliar to those without meteorological training (most ACP readers!).

Line 344: The term “bow-echo MCS” is used here without definition. The description appears later, beginning on line 362. Please move it ahead of this.

Line 363:  “produce”? Perhaps “are associated with” would be better. The radar echoes don’t cause the winds.

Line 367: Perhaps they should be shown? I find the evidence of descent unconvincing at present, and this is an important part of the paper. Also, a few lines below, you claim that “…strong radar reflectivities were confined below 6 km altitude (480 hPa, -9 ℃) suggesting limited vertical extension of convective storms.” It would be helpful to see those data.

Lines 388-394: This plot and description give me no useful information with which to evaluate the model performance. What exactly is being simulated? What observations are being compared? Does a POD of 0.8 mean we have 80% perfect agreement, or 80% chance of seeing something similar within 20 km? What does the SR of 20-80% mean, and what is a frequency bias (FR)?

In contrast, I do get some information from comparing Figures 8 and 9. Perhaps instead of S9 you could simply describe the agreement between these figures. It looks to me like WRF is simulating a system of similar size and strength in pretty much the same place.

Lines 442-448: This description and Figure 11 are quite confusing to me, as noted above. I’m not at all sure what the lines labelled “tracers” represent. Are they contours of particle counts? At 3.6 km?

Figures 2, 3 & 4: I find the times on the X-axis hard to read.