The impact of El Ni&ntilde;o&ndash;Southern Oscillation on the total column ozone over the Tibetan Plateau

Li, Yang; Feng, Wuhu; Zhou, Xin; Li, Yajuan; Chipperfield, Martyn P.

doi:https://doi.org/10.5194/egusphere-2023-1452

Preprints

https://doi.org/10.5194/egusphere-2023-1452

Preprints

18 Jul 2023

| 18 Jul 2023

The impact of El Niño–Southern Oscillation on the total column ozone over the Tibetan Plateau

Yang Li, Wuhu Feng, Xin Zhou, Yajuan Li, and Martyn P. Chipperfield

Abstract. The Tibetan Plateau (TP, approximately 27.5–37.5° N, 75.5–105.5° E) is the highest and largest plateau on Earth with a mean elevation of over 4 km. This special geography causes strong surface solar ultraviolet radiation (UV), with potential risks to human and ecosystem health, and which is controlled by the local total column ozone (TCO). The El Niño–Southern Oscillation (ENSO), the dominant mode of interannual variability on Earth, is characterized by warming of central and eastern equatorial Pacific Ocean sea surface temperature anomalies (SSTA) for the El Niño phase and cooling for the La Niña phase. Although some studies have suggested that there exists a positive correlation between ENSO and the TP TCO, the mechanism underlying this effect of ENSO is not fully understood.

Here we use the Copernicus Climate Change Service (C3S) merged satellite dataset and the TOMCAT 3–dimensional (3D) offline chemical transport model forced by ERA5 meteorological reanalyses from the European Centre for Medium–Range Weather Forecasts (ECMWF) over the period 1979–2021 to investigate the influence of ENSO on the TCO over the TP. The correlation coefficient of the time series of monthly TCO anomalies over the TP between the TOMCAT output and C3S dataset is ~0.92 with statistical significance above 99 %. In particular, the correlation coefficients (December–May) are above 0.95, indicating that the TCO variability in TOMCAT is very consistent with the merged satellite observations. Based on the lead correlation between the ENSO index (Niño 3.4 index from the National Oceanic and Atmospheric Administration) and TP TCO time series, we find that ENSO has a significant impact on the TCO, especially from the December of its mature phase to the May of its decaying phase (December–May). That is, the El Niño (La Niña) events induce positive (negative) TCO anomalies over the TP. Through studying the ozone profile from the Stratospheric Water and OzOne Satellite Homogenized (SWOOSH) dataset and TOMCAT results over their overlapping period (1984–2021), we attribute the positive (negative) TCO anomalies mainly to the increased (decreased) ozone at the 200–70 hPa levels caused by the downward (upward) shift of ozone profile associated with the lower (higher) tropopause height during the El Niño (La Niña) events. Our results suggest that the El Niño events lead to the descending upper–level geopotential height and hence cause the decreasing air column thickness, which in turn induces the cooling tropospheric temperature over the TP. This cooling associated with El Niño events causes a decrease of the tropopause height, which contributes to the downward shift of the ozone profile and hence leads to the increase in TCO. The La Niña events affect TP TCO during December–May in a manner resembling the El Niño events, except with anomalies of opposite sign.

This work provides a systematic understanding of the influence of ENSO on ozone over the TP. Since climate models project an increase in the frequency of strong El Niño and La Niña events under greenhouse–gas–forced warming, we can expect more ozone variation associated with ENSO, with important implications for 21st–century ozone recovery, surface solar UV, and ecosystems over the TP.

Received: 30 Jun 2023 – Discussion started: 18 Jul 2023

Download & links

Yang Li, Wuhu Feng, Xin Zhou, Yajuan Li, and Martyn P. Chipperfield

Status: final response (author comments only)

RC1:
'Comment on egusphere-2023-1452', Anonymous Referee #1, 24 Jul 2023

Review of egusphere-2023-1452 “The impact of El Niño–Southern Oscillation on the total column ozone over the Tibetan Plateau” by Yiang Li et al.

General comments

This article aims to explore the impact of ENSO on the variability of the total ozone column over the Tibetan Plateau. The topic of the work is absolutely interesting, as the subject is not much investigated and the mechanisms driving the variability of ozone over the Tibetan Plateau are not fully elucidated. Also, the paper is well written, so below there are only a limited number of technical corrections to improve the readability. However, the paper suffers from many important drawbacks in the methodology applied, and therefore calls for major improvements before it can be accepted for publication on Atmospheric Chemistry and Physics. First of all, the analysis is mostly based on the use of correlation coefficients, which alone cannot be used neither to document the performance of model simulations against observations, neither to prove the existence of physical links between two (or more) atmospheric processes. Indeed, the presence of a correlation in two variables is not alone sufficient to claim the existence of a cause-effect relationship. Even more so as in this case the analysed variables present well different spatial patterns (which is never clearly discussed). Secondly, the authors focus the analysis on composite averages analysing simultaneously multiple El Niño (La Niña) events, and it is not clear if the results can be ascribed to all events or just to some major leading ones. Tropopause height, but also ozone, present relevant day to day changes, which have not been thorousgly analysed and documented here. Thirdly, the reanalysis, satellite measurements and model analysis utilized have very different (and coarse) resolution, which can lead to significant drawbacks in the analysis. On the basis of my review, I guess that the paper would benefit from a major revision and explanation of the methodology applied and from an improved discussion to motivate, interpret and prove the results obtained. Something also in the direction of the suggestions from the editor, highlighting that other modes of variaiblity can impact on the total ozone time series, can be also beneficial.

Specific comments

Lines 14-15: Well, more precisely, the solar ultraviolet radiation and the derived risks to human and ecosystems health are controlled by the stratospheric ozone. It is true that the total ozone column is essentially equal to the stratospheric content as the tropospheric ozone concentration is too low as compared to the stratospheric one, but the sentence should be more precise.

Lines 15-18 and 62-64: ENSO has both an oceanic (El Nino) and an atmospheric component (Southern Oscillation), but here you refer only to the oceanic one. Please describe better.

Lines 28-32: The explanation is quite confused and not straightforward, I would suggest to rephrase to make it clearer.

Lines 32-36: I do not understand the meaning of “descending upper-level geopotential height”. Explain better. Also I suggest rephrasing: “Our results suggest that the El Niño events lead to a descending upper–level geopotential height and hence cause a decrease in air column thickness, which in turn induces reduced tropospheric temperature over the TP.”

Lines 34-36: I do not understand how just a shift in the ozone profile can lead to an increase in TCO.

Lines 38-39: And what about the frequency of the events? If the events of the two kinds change intensity of the same magnitude, but with opposite effects on ozone, wouldn’t be the final change almost null? Please explain better.

Lines 12-41: The abstract is too long and reporting many results of the correlation tests, while it is better to focus on the interpretation of the results and the underlying mechanisms. Please revise.

Line 47: Please explain better how the elevation of the site is linked to the low air density and high atmospheric transparency.

Lines 48-49: How is this connected with the previous sentence?

Lines 54-58: Those mechanisms that you listed here are various and of different nature. I would suggest better description.

Lines 58-60: Before you were talking just about summer TOL, now you talk about winter and spring events, perhaps (not clear) connected to different processes. Revise.

Lines 61-62: As commented above, the link between QBO and TOL is not explained. Also, in this sentence, it is not clear if there is a link between QBO and ENSO. Revise.

Lines 64-66: Are you talking about climate or about meteorology or of the Earth system when you talk about the interannual variability of ENSO?

Lines 68-69: The sentence and the reason why most studies have focused on the polar regions and the tropical stratosphere is not clear.

Line 72: I would not label 1979-2002 measurements as “very limited”, if not explained better what the limitations are.

Line 74: You never talked of a limitation in spatial coverage, so the reader has no way to compare the claimed much wider range of your work against previous ones.

Line 75: It is not exactly clear, as stated now, what the chemical transport model adds to the analysis.

Lines 81-83: This sentence is more appropriate for the conclusion section rather than for the introduction.

Lines 102: The resolution seems to be rather coarse, which can pose limitations to the study.

Lines 86-119: I would suggest explaining better what kind of measurements/observations are derived from each source, as the sources are many and of different kinds: satellite, reanalysis, …

Line 128: Usually ECMWF means that you are not using always ECMWF data? From what does this depends? Also, which ECMWF reanalysis?

Lines 127-129: The sentence is not clear, as presented now, as it repeats that the model is forced with ECMWF winds and temperatures to specify atmospheric transport and temperatures. Revise.

Lines 130-131: Yet another different, and coarse, resolution...

Lines 135-136: This also depends on the purpose of the investigation...

Lines 157-168: The correlation coefficients alone cannot prove the goodness of the simulations against observations, as they indicate only that the model reproduces the temporal variability observed, but the presence of biases cannot be detected by correlations. In any case, the fact that in some seasons (months? Please see next comment) you have low correlations points out that there could be differences also in the simulated temporal patterns, at least in some seasons, and this needs to be better discussed. Revise.

Figure 1: If the correlations are 3-months, and we have 4 seasons, why do you have 12 columns in the plot?

Line174: The seasons are already utilized previously, so must be explained previously.

Line 177 and also 136-137: The use and the explanation of the lead-lag correlation coefficient is not clear.

Line 183-186: Not clear, revise.

Line 191: Is it standard deviation or variance?

Table 1 and 2: What is the meaning of the mean Niño 3.4 index?

Lines 211-221: The spatial pattern of the variability is different in the two phases, and should be discussed. I would also recommend discussing the relevance of the changes in percentages (are 8 DU anomalies relevant?) The discussion is in any case not clear and should be revised.

Figure 4 and 5: The plot documents that there are significant disagreements between the model and the observations, especially in the La Nina phase where the simulated vertical profile is remarkably different than the observed one. I am not sure if this depends from one particular event which is not simulated correctly or it is a general problem, since the authors have always used the composite seasonal means considering together DJF and MAM rather than analysing single events..

Figure 6: Are the spatial changes in TH similar to those in TOC? The spatial pattern seems different, so it is not clear how the two changes can be actually connected. Also, the implied changes in ozone documented in Figure 6c and d are remarakbly lower than those presented previously...

Lines 282-283: Please explain better how TH and the SSTA are linked to atmospheric circulation, and if they are linked. Up to now you were talking of just TH...

Lines 288-308: the discussion is confused and not straightforward.

Lines 317-319 and 336-349: The analysis of correlation coefficients alone cannot justify the physical mechanism that you are implying. Please better discuss.

Figure 8: Yet another spatial pattern, different than those presented previously...

Figure 9 and line 338: It is not clear the meaning of “temperature associated with air thickness..

Technical comments

Line 19: Perhaps remove “of ENSO”?

Line 25 and 363: “lead”?

Lines 69-70: I would suggest rephrasing: “The effects of ENSO on ozone changes at mid-latitude and in particular over the TP are less studied and discussed.”

Lines 70-71: Rephrase “... suggest the amplitude of ENSO signal in TCO over the TP to be of the order of 20 DU (their figure 3)”

Line 78: Add “following” after “three”

Line 86: what do you mean by “merged”?

Line 220: Change “understanding” to “understand”

Citation: https://doi.org/10.5194/egusphere-2023-1452-RC1
- AC1: 'Reply on RC1', Yang Li, 09 Oct 2023
  
  We thank the reviewer for taking the time to evaluate our manuscript and for their positive and helpful comments. You can find a detailed response to the reviewer's comments on the attached supplement.
  
  Citation: https://doi.org/10.5194/egusphere-2023-1452-AC1
RC2:
'Comment on egusphere-2023-1452', Anonymous Referee #2, 28 Jul 2023

General comments

The article ”The impact of El Niño–Southern Oscillation on the total column ozone over the

Tibetan Plateau” submitted by Yang Li et al. studies the connection of the ENSO to the total

column ozone (TCO) above the Tibetan Plateau (TP). By investigating long-term satellite data

from the C3S, the chemical transport model TOMCAT, and the water vapor and ozone data set

SWOOSH, the authors connect the positive (negative) anomalies in the Niño 3.4 index to anomalies

in the TCO and ozone profiles.

The study of this topic is very interesting and the article would be well suited for ACP and

an important contribution to the community. In addition, the article is well-written and mostly

understandable. There are, however, some aspects to the scientific presentation and content that

need major revision before recommendation for publishing. First of all, the analysis restricts to the

assessment of anomalies averaged over multiple Niño events, there is no mention of the spread

between events. A comprehensive study would benefit greatly from assessing, or even briefly

showing, the variability between different events and the dependence of the anomalies on the Niño

3.4 Index. Furthermore, the analysis of this article is restricted to correlations between different

anomalies. Drawing conclusions on the causation of the TCO stays, therefore, difficult. Especially,

since, as the authors mention, The TCO is influenced by multiple effects. There is, however, no

apparent attempt to decouple the considered effect of the Niño from the other processes. Lastly,

the explanation of the positive TCO anomaly by a downward shift of the ozone profile is lacking.

How would a mere downward shift alter the total ozone in a column? Or is partial column ozone

considered? The authors should explain the mechanism behind profile shifting leading to increased

ozone in a clearer way in order to make it comprehensible.

Specific comments

• Line 47: ”high atmospheric transparency” This region is below the Asian Tropopause Aerosol

Layer, which should affect the atmospheric transparency as well. Consider adding a comment

on its effect.

• Lines 54–58: Please expand a bit on these processes: Brief description of the mechanism

(at least for the dominant effects). Is there seasonal varying importance of the different

processes?

• Lines 62–64: You explain the EN part of ENSO here. Please add a sentence on the Southern

Oscillation, i.e., the atmospheric anomalies of the ENSO.

• Lines 64–69: Why are these regions ”showing the significant interannual variability”? Please

rephrase or expand.

• Lines 72–73: Expand on what limits the satellite measurements. Probably, there is only a

limited number of ENSO events in this time period. The sentence could also be understood that there are deficiencies in the measurements themselves. Please clarify.

1• Lines 95–96: ”The long-term stability of the TCO product is within the 1% per decade level”

It’s not clear to me what this means, please expand.

• Line 103: ”and has 12 levels per decade in pressure ranging from 316 to 1 hPa (31 pressure

levels)” Do you mean 12 time steps per decade, i.e., 10-monthly data? Please clarify what

the ”12 levels per decade” refer to.

• Lines 109–111: Why not use the SST from ERA5? Please briefly comment.

• Lines 111–113: Why use the Niño 3.4 index instead of other indices? Please briefly comment.

• Line 158: This is a running 3-month mean, right? Consider stating this in the text.

• Lines 179–181: Please state that this refers to the bars in Fig. 2.

• Line 185: ”as one could be expected from”, remove either ”one” or change ”be expected” to

”expect”.

• Lines 181–186: The variance is much higher in TOMCAT than in the observations (50–

100%). Is this accounted for in the following considerations? I would not call this ”reasonable

magnitude” but still the variability is well-matched. Maybe you could use the systematic

difference found here to put the later results into perspective.

• Lines 212–221: Here, all Niño events have been composited. It is unclear how the composition

was performed: e.g. average or weighted average according to the Niño 3.4 index? Please

specify. In addition, it is unclear what the behavior for individual Niño events is. Please give

at least a comment about the variability throughout the different events.

• Figure 4: Plotting the (standard) deviation of the profiles, i.e., the profile ± variability,

would be an easy way to show the variability between different events. Consider adding

these intervals (e.g. as shading) to the figure.

• Lines 253–254: It is unclear to me how a downward shift of a profile could singularly alter the

total content of ozone in the respective column. Either the partial column ozone is changed,

e.g., ozone up to 50hPa, or there has to be an increase in production/mixing from adjacent

regions. A stretching (or compression) of the profile, for example, would change the TCO.

Please clarify the mechanism that, in the end, leads to increased TCO.

• Line 262: Are the latitude-height sections averaged in longitude or taken as a cross-sectionat a fixed longitude? Either way, please specify.

• Figure 6: The change in TH alone (located mostly between 30◦ –35◦ N) does not explain the

widespread change in ozone stretching to at least 42◦ N. Locally, I agree that the TH might

contribute but do you have hypotheses on the cause of the northern part of the anomalies?

• Lines 299–304: Could there be a surface temperature anomaly above the TP due to Niño

events? If not or of the opposite sign to the Indian Basin SSTA, it could strengthen the

argument of the land-sea contrast.

• Line 338 & Fig. 9: Specify what ”temperature associated with air thickness” refers to more

clearly. I suppose this is the temperature as calculated from Eq. 3?

• Lines 344–346: There are some severe outliers in Fig. 9, e.g., La Nina with -2T thickness.

Are the outliers generally corresponding to a weaker Niño index? Consider coloring the scatter

plot with the Niño 3.4 index instead of blue/orange. But, of course, there could be various

other processes involved in singular events.

Technical comments

• Line 119: ”to the 1984–2021.” There seems to be a word missing here: average, period?

• e.g. line 219: ”from the December of the ENSO’s mature phase to the May of” the use of

”the” in front of a month is usually incorrect and reads cumbersomely. Consider removing

”the”. The same is true for time periods throughout the text, e.g., ”the YEAR–YEAR” →

”YEAR–YEAR” (unless using a trailing noun such as ”the YEAR–YEAR period”).

• Line 266: ”further results” → ”further contribute”

• Line 288: Consider dropping ”as” in ”is considered as an important”.

• Line 295: ”is a response of SSTA” → ”is a response to SSTA”

• Figure 7: Consider enlarging the text on the color bars.

Citation: https://doi.org/10.5194/egusphere-2023-1452-RC2
- AC2: 'Reply on RC2', Yang Li, 09 Oct 2023
  
  We thank the reviewer for taking the time to evaluate our manuscript and for their positive and helpful comments. You can find a detailed response to the reviewer's comments on the attached supplement.
  
  Citation: https://doi.org/10.5194/egusphere-2023-1452-AC2

Yang Li, Wuhu Feng, Xin Zhou, Yajuan Li, and Martyn P. Chipperfield

Viewed

Total article views: 500 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
341	129	30	500	19	22

HTML: 341
PDF: 129
XML: 30
Total: 500
BibTeX: 19
EndNote: 22

Views and downloads (calculated since 18 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	120	28	9	157
Aug 2023	39	12	1	52
Sep 2023	19	12	2	33
Oct 2023	47	21	7	75
Nov 2023	16	8	0	24
Dec 2023	28	18	3	49
Jan 2024	21	4	1	26
Feb 2024	12	13	1	26
Mar 2024	24	10	2	36
Apr 2024	15	3	4	22

Cumulative views and downloads (calculated since 18 Jul 2023)

Month	HTML	PDF	XML	Total
Jul 2023	120	28	9	157
Aug 2023	39	12	1	52
Sep 2023	19	12	2	33
Oct 2023	47	21	7	75
Nov 2023	16	8	0	24
Dec 2023	28	18	3	49
Jan 2024	21	4	1	26
Feb 2024	12	13	1	26
Mar 2024	24	10	2	36
Apr 2024	15	3	4	22

Viewed (geographical distribution)

Total article views: 476 (including HTML, PDF, and XML) Thereof 476 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Apr 2024

Short summary

Tibetan Plateau (TP), the highest and largest plateau, experiences strong surface solar UV radiation, whose excess can cause harmful influences on the local biota. Hence, it is critical to study TP ozone. We find ENSO, the strongest interannual phenomenon, can induce tropospheric temperature change and thus cause tropopause movement, which in turn leads to ozone change. Our understanding implies that future more strong ENSO events will play a key role in ozone recovery and ecosystems of the TP.


Total:	0
HTML:	0
PDF:	0
XML:	0