The paper "Nonlinear effects of the stratospheric Quasi‐Biennial Oscillation on ENSO modulating PM2.5 over the North China Plain in early winter" by An et al. investigates the combined effect of the QBO and ENSO on the PM2.5 concentration over the North China Plain. Main findings are that PM2.5 is enhanced for El Nino and westward QBO conditions, while PM2.5 is reduced for La Nina and eastward QBO. This effect is explained by variations in wind speed and direction, boundary layer height, and humidity.

Overall, the paper is a thorough study, it is well written, and of interest and relevance for the readership of ACP.

The paper is therefore recommended for publication in ACP after minor revisions.

My main comments are:

(1) the authors should include the model equations they assume for their multivariate linear regressions

(2) you should explain a bit why enhanced humidity would lead to enhanced PM2.5, but not to precipitation that would wash out air pollution

Please find more Specific and Technical comments below.

Specific comments:

(1) Fig.1d: what is the source of the SST data?

(2) l.116 onward:

You should state more clearly in the text that multivariate regression analyses are performed. The underlying linear models should be given as additional equations.

(3) l.120: In the text you should add the information that the regression coefficients in Figs.2b and 2d have opposite sign in most regions of China, while in the North China Plain (NCP), which is the main focus of your work, PM2.5 anomalies do not change much.

(4) l.153: The point with the humidity was not completely clear to me!

If humidity is high enough, precipitation would form and wash out air pollution.

Please clarify whether or not this is a relevant mechanism during conditions of enhanced humidity in November/December over the NCP.

This should also be clarified in the discussion around l.246.

(5) l.442: In addition to the white dots, there are also stippled areas in Fig.2. Do these areas refer to another different level of significance?

I would suppose that the white dots refer to 80%, and the black grid areas to 90% of significance, like in Fig.5.

Please explain and correct!

(6) Fig.3: to avoid confusion, please use same notation as in Fig.2

(ENSO- instead of ENSO^-1 and QBO- instead of QBO^-1)

(7) Fig.6, Fig.7: Green lines represent the Yellow River and the Yangtze River? Please add this information.

Technical corrections:

(1) l.40:

Given its the crucial in shaping global weather and climate

->

Given its crucial role in shaping global weather and climate

(2) reference Austin and Tu (2004) is missing in the References list

(3) l.105: Palmer -> Palm

(4) l.107: reference Chen et al. (2013) is not in the references list, did you mean Chen et al. (2003) ?

(5) l.117: It is clearly -> It is evident

(6) l.119:

whether is La Nina or El Nino, these is no

->

whether ENSO is in the La Nina or El Nino phase, there is no

(7) l.130: The now question is -> The question now is

(8) l.138: Fig.3 -> Fig.4

(9) l.142: (Fig. 4a and 4c) -> (Fig. 5a and 5c)

(10) l.148: 4e and 4g -> 5e and 5g

(11) l.155: remining -> remaining

(12) l.199: 9c 9e and 9g) -> 9c, 9e, and 9g)

(13) l.203: OBO -> QBO

(14) l.442: areas of significant -> areas of significance

(15) Caption of Fig.7: Is interval = 1, or 0.5? Please check!

The authors analyze the effects of the QBO on the ENSO influence on 

North China Plains pollution. The present work would benefit from substantial

improvements, namely in the methodology, quality of presentation and 

level of discussion. 

My impression is that the ENSO is the leading factor, and the QBO importance

is secondary. This needs to be quantified numerically, beyond the simple Fig. 3

analysis.

The description and selection of data needs improvement (see detailed comments).

The use of a fairly dated reanalysis product is not justified; while it may be useful

to retain it as a reference for other studies, the study should mainly rely

on up-to-date datasets, such as ERA5 (already used, partially). 

Please move the legacy reanalysis contents to the supplement, or remove; 

any difference should anyway be explained.

The period of analysis needs to be given, currently many dates are provided

for the various products, which is confusing.

The authors mention using two PM2.5 products, but the results seems to

agree only qualitatively. How do the the products differ?

The methodology should be improved, as it is unclear why the authors

selected the Nov-Dec season. In the introduction, Nov-Jan is also mentioned.

Given that nonlinear diagnostics are computed, the authors need to 

explain if high-resolution data (e.g., at least daily) were used,

as signals would otherwise be smoothed.

The authors ascribe PM2.5 changes to changes in atmospheric circulation. However, my understanding

is that haze conditions can be extremely local, e.g. due to the presence of urban settlements

in valleys. The discussion needs to be enhanced to explain relevant local features, including sources.

Relevant analysis should be included, to understand the relative contributions of

meteorology but also of primary and secondary sources.

The motivation of section 4 is lacking. Why do these diagnostics matter?

The authors need to clarify how their work is distinct from previous

papers (several papers from the first author are cited) and what is the new understanding, 

see e.g. the non-exhaustive list of works below (none cited, as far as I can see).

Based on a simple online search, a number of references seem missing:

An et al. 2018 https://acp.copernicus.org/articles/18/1863/2018/ on ENSO and winter pollution in China

Jiang et al. 2017 https://link.springer.com/article/10.1007/s13351-017-6412-z on drivers on haze in China

Li et al. 2023 https://doi.org/10.5194/acp-23-1533-2023 on summer ozone pollution in China and QBO

Lu et al. 2022  https://doi.org/10.1029/2022JD036938 on ENSO and EASM and summer ozone pollution in China

Ray et al. 2020 https://www.nature.com/articles/s41561-019-0507-3 on QBO and trace gases

Sun et al. 2018  https://doi.org/10.1029/2018JD028534 on ENSO and winter pollution in East China

Zhang et al. 2025 https://doi.org/10.1029/2024JD041825 for a model study of QBO/ENSO influence on ozone transport

Zhao et al. 2016 https://www.nature.com/articles/srep27424 on decadal variability of haze in China and tropical Pacific

The authors are advised to revised the text with care to correct various typos and unsupported statements,

a few reported in the list below. Also the text structure could be improved to avoid long paragraphs

and focus the discussion.

Comments by line

L32 This reference is not suitable. Are there peer-reviewed papers on this 'comprehensiveness'? 

E.g. comparing in-situ or satellite-based emissions/concentrations?

L40 unclear 'its the crucial'

L44 I don't get this, which would be a review?

L47 You cannot expect the reader to know in detail the regional geography. Please include a maps

with all the relevant names and features mentioned, perhaps in the supplement.

L64 typo 'underling'

L70 there is virtually no description of the data. Is it in-situ? Satellite? Model? The zenodo

repository does not show the information neither. The link should be given elsewhere, while

here readable information.

L80 The expression 'Fifth major global reanalysis' makes no sense. Is there a minor version as well?

The information on the spatial resolution and temporal coverage are both incorrect,

You probably refer to what you retrieved. Also the reference is to the website, not the paper.

L86 10 hPa is quite high, and here you are looking at instantaneous signals.

Why not using 50 or 70 hPa to characterize a tropospheric process?

Moreover, periods 2015/16 and 2019/20 showcase distinct QBO evolutions,

which should be acknowledged

L87 Based on the plots, there is one point per year. Is this an average of Nov-Dec? This

needs to be explained, and points added in Fig. 1 b/c

L89 The methodology of "Jan Null" is not explained. Please use some method justified in the 

literature

Table 1 caption: period should be lowercase

Fig.1 are the units of regression coefficient degC as reported? 

L104 The period used should be explained

L117 typo 'clearly'

Fig. 2 one among g or h should be for the QBO?

Fig. 3 I do not understand why you have "-1" subscripts. Are these referring to lagged values?

L142 You are presumably referring to Fig. 5

Fig. 5 Units are not indicated anywhere

L165 by whom?

Fig. 6 This figure definitely differs from Fig. 4. Verify

Fig. 7 'disturbed' in what sense?

L187 A similar statement was already done before

Fig. 8 is hard to read and needs to be improved graphically

L203 what's OBO now?

L211 Unclear sentence

L226 The 'Eurasia hinterland' sounds odd

L237 All this discussion does not seem relevant. 

L252 What do you mean? What numerical model? A transport model?

L260 the link for the QBO has moved, verify. Links have been visited one year ago or more!

L263 I can't locate experimental results in this work

Fig. S2 Since all terms have an 'I', it could be safely omitted. What is the 'forecast data' mentioned?

Is it again ERA5?