Dry and warm conditions in Australia exacerbated by aerosol reduction in China

Gao, Jiyuan; Yang, Yang; Wang, Hailong; Wang, Pinya; Liao, Hong

doi:https://doi.org/10.5194/egusphere-2024-3399

Preprints

https://doi.org/10.5194/egusphere-2024-3399

Preprints

18 Nov 2024

| 18 Nov 2024

Dry and warm conditions in Australia exacerbated by aerosol reduction in China

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Abstract. A substantial decline in anthropogenic aerosols in China has been observed since the initiation of clean air actions in 2013. Concurrently, Australia experienced anomalously dry and warm conditions in 2010s. This study reveals a linkage between aerosol reductions in China and the drying and warming trends in Australia during 2013–2019 based on aerosol-climate model simulations and multi-source observations. Aerosol decline in China triggered alterations in temperature and pressure gradients between the two hemispheres, leading to intensified outflow from Asia towards the South Indian Ocean, strengthening the Southern Indian Subtropical High and its related Southern Trade Winds. Consequently, this atmospheric pattern resulted in a moisture divergence over Australia. The reduction in surface moisture further resulted in more surface energy being converted into sensible heat instead of evaporating as latent heat, warming the near-surface air. Aerosol reductions in China are found to contribute to 19 % of the observed decreases in precipitation and relative humidity and 8 % of the increase in surface air temperature in Australia during 2013–2019. The intensified dry and warm climate conditions during 2013–2019 further explain 12 %–19 % of the increase in wildfire risks during fire seasons in Australia. Our study illuminates the impact of distant aerosols on precipitation and temperature variations in Australia, offering valuable insights for drought and wildfire risk mitigation in Australia.

Received: 31 Oct 2024 – Discussion started: 18 Nov 2024

Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Preprint (PDF, 1092 KB)

Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Preprint (1092 KB)

Supplement (3468 KB)

Download & links

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Journal article(s) based on this preprint

23 Sep 2025

Dry and warm conditions in Australia exacerbated by aerosol reduction in China

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Atmos. Chem. Phys., 25, 10949–10964, https://doi.org/10.5194/acp-25-10949-2025,https://doi.org/10.5194/acp-25-10949-2025, 2025

Short summary

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-3399', Anonymous Referee #1, 13 Dec 2024

Please see my comments in the attached PDF.

Citation: https://doi.org/10.5194/egusphere-2024-3399-RC1
- AC1: 'Reply on RC1', Jiyuan Gao, 01 Feb 2025
  
  Please see our responses in the attached PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3399-AC1
RC2:
'Comment on egusphere-2024-3399', Anonymous Referee #2, 18 Dec 2024

The manuscript investigates the role of the recent decline in East Asian anthropogenic aerosols in driving precipitation and temperature changes across Australia. The authors highlight the potential for aerosols to contribute to enhancing the fire risk via widespread drying and warming. The topic is timely and very important, as not many studies have examined the influence of recent changes in aerosols on recent climate variations. However, I find several major weaknesses in the study, including in the simulation design, that prevent the manuscript from being acceptable for ACPD. Unfortunately, I need to recommend a rejection. Happy though to reconsider the manuscript if reviewed accordingly.
Major comments:
I believe there is a serious overinflation of the recent observed rainfall changes over Australia. Firstly, the authors use observations and plot linear trends from 2010, while model results are shown as 3-year differences from 2013. While the latter accounts for the decline in aerosols, observations show that the years 2010 and 2011 were anomalously wet over Australia, and just a couple of years before rainfall was much less (I have checked by plotting observations since 2000). This can also be inferred by examining Fig 2, where clearly the trend in panel d is affected by the two early years. By eye, using 3-year composites from 2013, the rainfall changes are very modest. Looking at more recent years, it turns out that 2019 was also anomalously dry, and more recent years (excluding those affected by COVID-related reduced emissions) show a recovery. Therefore the results based on observations, which ultimately is the motivation of the study, are strongly affected by the extremely short record and cannot be trusted.
Related to the point above, it is not surprising that observed circulation trends (e.g., Fig. S18) are extremely weak. Yet, this is the mechanism driving Australian rainfall changes, thus is central to the proposed aerosol influence.
The model analysis is also weak in the sense that 3-year composites are extremely short to identify forced trends. In addition, observed trends should be first compared to those from a control simulation and the contribution of aerosols should be related to the model world, not observations.
Another major weakness is the experimental design. The authors examine long-term changes coming from equilibrium runs, which are far from representing the actual transient response over 10 years or so. This is likely far from being in equilibrium. I would have seen large-ensemble transient simulations with time-varying aerosols (2013-2024 for example) as more appropriate and suitable.
Trends or changes should always be plotted with the related statistical significance, such as in Figs. 1 and 2.
Fig 3: I think it would be more appropriate to display the divergent wind, rather than the total wind.

Citation: https://doi.org/10.5194/egusphere-2024-3399-RC2
- AC2: 'Reply on RC2', Jiyuan Gao, 01 Feb 2025
  
  Please see our responses in the attached PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3399-AC2

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2024-3399', Anonymous Referee #1, 13 Dec 2024

Please see my comments in the attached PDF.

Citation: https://doi.org/10.5194/egusphere-2024-3399-RC1
- AC1: 'Reply on RC1', Jiyuan Gao, 01 Feb 2025
  
  Please see our responses in the attached PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3399-AC1
RC2:
'Comment on egusphere-2024-3399', Anonymous Referee #2, 18 Dec 2024

The manuscript investigates the role of the recent decline in East Asian anthropogenic aerosols in driving precipitation and temperature changes across Australia. The authors highlight the potential for aerosols to contribute to enhancing the fire risk via widespread drying and warming. The topic is timely and very important, as not many studies have examined the influence of recent changes in aerosols on recent climate variations. However, I find several major weaknesses in the study, including in the simulation design, that prevent the manuscript from being acceptable for ACPD. Unfortunately, I need to recommend a rejection. Happy though to reconsider the manuscript if reviewed accordingly.
Major comments:
I believe there is a serious overinflation of the recent observed rainfall changes over Australia. Firstly, the authors use observations and plot linear trends from 2010, while model results are shown as 3-year differences from 2013. While the latter accounts for the decline in aerosols, observations show that the years 2010 and 2011 were anomalously wet over Australia, and just a couple of years before rainfall was much less (I have checked by plotting observations since 2000). This can also be inferred by examining Fig 2, where clearly the trend in panel d is affected by the two early years. By eye, using 3-year composites from 2013, the rainfall changes are very modest. Looking at more recent years, it turns out that 2019 was also anomalously dry, and more recent years (excluding those affected by COVID-related reduced emissions) show a recovery. Therefore the results based on observations, which ultimately is the motivation of the study, are strongly affected by the extremely short record and cannot be trusted.
Related to the point above, it is not surprising that observed circulation trends (e.g., Fig. S18) are extremely weak. Yet, this is the mechanism driving Australian rainfall changes, thus is central to the proposed aerosol influence.
The model analysis is also weak in the sense that 3-year composites are extremely short to identify forced trends. In addition, observed trends should be first compared to those from a control simulation and the contribution of aerosols should be related to the model world, not observations.
Another major weakness is the experimental design. The authors examine long-term changes coming from equilibrium runs, which are far from representing the actual transient response over 10 years or so. This is likely far from being in equilibrium. I would have seen large-ensemble transient simulations with time-varying aerosols (2013-2024 for example) as more appropriate and suitable.
Trends or changes should always be plotted with the related statistical significance, such as in Figs. 1 and 2.
Fig 3: I think it would be more appropriate to display the divergent wind, rather than the total wind.

Citation: https://doi.org/10.5194/egusphere-2024-3399-RC2
- AC2: 'Reply on RC2', Jiyuan Gao, 01 Feb 2025
  
  Please see our responses in the attached PDF.
  
  Citation: https://doi.org/10.5194/egusphere-2024-3399-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Jiyuan Gao on behalf of the Authors (01 Feb 2025) Author's response Author's tracked changes Manuscript

ED: Reconsider after major revisions (24 Mar 2025) by Laura Wilcox

Both reviewers raised major concerns with this manuscript, which I don’t think have been sufficiently addressed in either the response or the revisions. I share the reviewers’ concerns about the experiment design used in this study, and would like to see these addressed in more detail before returning the manuscript to the reviewers.

The observed precipitation trend is evaluated based on only 10 years of data, and there is no discussion of how this trend compares to either longer term trends, or internal variability. The trend is also based on only one observational dataset.
• I think it is important that we see a comparison between the GPM data and other precipitation datasets such as GPCP to give a sense of how robust this observed trend is.
• It is also important that the authors show a longer precipitation timeseries, and evaluate the interannual variability, to give some context to the magnitude and significance of the trend. Reviewer 2 is concerned that 2010 and 2011 were anomalously wet, and thus bias the trend. They have plotted data from 2000 – I ask the authors to at least do the same, and would strongly encourage them to plot an even longer timeseries to provide more context. It is stated in the abstract that Australia experienced anomalously dry and warm conditions since the 2010s, but, beyond a few citations, there isn’t any evidence in the manuscript to support this.
• In the response document, the authors state that the observed trend is calculated from 2010-2019 instead of 2013-2019 to minimize the internal variability. This is still a very short time period, and a quantitative assessment of internal variability is required if readers are to be confident that it is a forced trend.

The reviewers also raise important concerns about the experiment design, which warrant more discussion than they have been given in the current response and revision. I am not convinced by the response to reviewer 2, and think the authors undermine their own interpretation of their results in this response. Equilibrium runs are an excellent tool for understanding the physical response to forcing, and for identifying the mechanisms for particular responses, as the authors have done in this study. However, care needs to be taken when comparing them to real-world changes. As the authors state in their response, ‘the relatively short period from 2013 to the present may not provide enough time for the climate system to fully respond to aerosol changes’. But, this full response is exactly what their equilibrium simulations show. So, how can these experiments be used to quantify an aerosol contribution to an observed trend, as the authors do in their manuscript? Is it possible to use such simulations to make anything more than qualitative statements about observed trends? One could perhaps quantify the Chinese aerosol contribution to a modelled response, but this would require other forcings to be considered. Reviewer 2 is recommending rejection of the manuscript based on the experiment design. I would like to see a more detailed discussion of the design of the study, and the implications of comparing several hundred years of simulated equilibrium responses to just 10 years of observed data, in both the response to the reviewers and the revised manuscript before making a decision.

I am confused by the interpretation of the low bias in the PM2.5 and AOD trends in CESM1 found in the response and the revised manuscript. CESM1 is known to underestimate PM2.5. The authors have calculated a change in PM2.5 and found that it is small compared to the observed change. If this is based on an absolute change, and the background PM2.5 in the model is low, then one would also expect the change to be small compared to observations. However, even though aerosol concentrations and AOD in the model are low, CESM1 has a relatively large aerosol forcing. Thus, a small absolute change in PM2.5 or AOD may result in a large forcing in the model. I don’t think a small absolute change in the model necessarily implies an underestimate of the forced response. How do the percentage changes in PM2.5 and AOD in the model compare to the percentage changes in observations, for example?

Thinking back to the discussion of precipitation, it would also be good to know more about the significance of the observed trend. The emission reductions have been very large, but how significant is the change in AOD? Is there potentially a contribution from internal variability here that accounts for some of the model underestimate of the change relative to observations?

In addition to addressing the major comments above:
• it is necessary to improve the indication of significance on the map figures in the manuscript. The shading is difficult to see in many of the figures, both on my screen and on a printout.
• I would like to see some discussion of how these results compare to previous studies assessing the Australian precipitation response to Asian aerosols. Rotstayn et al. (2007) and Fahrenbach et al. (2023) are cited in the introduction, but are absent from the results and discussion sections.

Hide

AR by Jiyuan Gao on behalf of the Authors (08 Apr 2025) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (05 May 2025) by Laura Wilcox

RR by Anonymous Referee #1 (07 May 2025)

RR by Anonymous Referee #2 (28 May 2025)

ED: Reconsider after major revisions (23 Jun 2025) by Laura Wilcox

There is a nice experiment, and some nice analysis of it in this paper, but there are still major concerns about the framing of the work and the interpretation of the results, and we cannot proceed to publication until these are resolved. It might be that they cannot be resolved. In which case, the authors should motivate their study based on the large emission changes, rather than the observed precipitation, and frame it as a speculative study of the potential responses. However, the issues with interpretation raised by all reviewers must be addressed regardless.

Framing
The premise of the paper is that Australia has experienced anomalously dry and warm conditions since the 2010. The authors then hypothesise that aerosol reductions in China since 2013 have contributed to this.

The observed precipitation trend is based on only 10 years of data, and there is still no discussion of how this trend compares to internal variability. The additional Figure S12, provided in response to my comments, raises more concerns here than it solves. The abstract claims that the period since 2010 was warm and dry in Australia, but the mean of the two periods highlighted in Figure S12 are almost identical. There is no evidence in that figure that Australia was anomalously dry since 2010.

Figure S14, provided in response to my comments, is reassuring, as it shows that multiple datasets agree that there has been a drying since 2010. However, I disagree with the authors’ conclusion that Figure S12 ‘clearly shows an increasing precipitation trend during 2001-2010, followed by a decreasing trend during 2010-2019.’ What I see here is variability. To be convinced that the trend since 2010 is anomalous, it needs to be compared to all other 10-year trends in ERA5 (using the available data from 1940 to the present to evaluate this). Is the negative trend from 2010-2019 unusually large in the context of 1940-2019?

Figure S12 also highlights the dependence of the 2010-2019 trend on what appears to be anomalously high precipitation in 2010 and 2011. Without these years, the trend would be markedly reduced. This is important, and more so in this case, since the driver that is being explored in the study occurs over 2013-2019. The authors should address the concerns from Reviewer 1 about the sensitivity of the trend to these years, particularly in light of Figure S12.

Analysis
The authors perform an equilibrium experiment to examine the Australian response to emission changes in China, North America and Europe, and everywhere except China. This shows that the response to Chinese emission changes between 2013 and 2019 is larger than that to emission changes in other regions. It also shows that the pattern of the precipitation response to Chinese emission changes is different to the response to emission changes elsewhere. The authors also show a strong seasonal component to the response. They also discuss the mechanisms for the response.

Neither the reviewers nor I have substantial comments on the analysis at this stage, as the points raised in review have focussed on the framing and interpretation of the study.

Interpretation
The authors claim that the warming and drying seen in response to Chinese emission reductions in their equilibrium experiments accounts for some of the drying and warming seen in the real world.

Reviewer 1 is concerned about the use of an equilibrium experiment to interpret real-world changes that have happened over a 7-year period. They point to literature that shows that the fast response of Australian precipitation to Asian aerosols can have the opposite sign response to the slow response, and therefore caution that an equilibrium experiment cannot be used to interpret the drivers of a fast change in the real world. It is important that the authors address these concerns.

There are also some inconsistencies and errors in the response to my comments on the interpretation of the experiment. The authors correctly note that ‘the relatively short period from 2013 to the present may not provide enough time for the climate system to fully respond to aerosol changes’. There then needs to be some discussion of the differences between the fast and the slow responses to aerosol changes. As the climate system may not have had time to fully respond to the aerosol changes, how do we expect an equilibrium simulation (where this has happened) to compare to the real world (where it may not have happened)? Can the authors demonstrate that the equilibrium response can be reached within the observational period, so that we can use this experiment to interpret real-world changes? If not, can they comment in detail on where they might expect their experiment to differ from the real-world response? The perturbation is applied as a step change in the equilibrium experiment, so it should be possible to find the time taken for the system to reach equilibrium in this case, and to assess whether the short-term response differs to the equilibrium response. Reviewer 1 also suggests an additional experiment to examine the fast response explicitly. This is a relatively cheap simulation compared to the equilibrium runs, and would only need to be performed for the China experiment.

The authors have removed most direct attribution statements from the text, but Reviewer 2 highlights one that remains, and should also be removed.

Hide

AR by Jiyuan Gao on behalf of the Authors (15 Jul 2025) Author's response Author's tracked changes Manuscript

ED: Publish subject to technical corrections (31 Jul 2025) by Laura Wilcox

AR by Jiyuan Gao on behalf of the Authors (01 Aug 2025) Author's response Manuscript

Journal article(s) based on this preprint

23 Sep 2025

Dry and warm conditions in Australia exacerbated by aerosol reduction in China

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Atmos. Chem. Phys., 25, 10949–10964, https://doi.org/10.5194/acp-25-10949-2025,https://doi.org/10.5194/acp-25-10949-2025, 2025

Short summary

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Supplement

https://doi.org/10.5194/egusphere-2024-3399-supplement

Jiyuan Gao, Yang Yang, Hailong Wang, Pinya Wang, and Hong Liao

Viewed

Total article views: 851 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
728	101	22	851	51	20	40

HTML: 728
PDF: 101
XML: 22
Total: 851
Supplement: 51
BibTeX: 20
EndNote: 40

Views and downloads (calculated since 18 Nov 2024)

Month	HTML	PDF	XML	Total
Nov 2024	74	10	5	89
Dec 2024	104	50	2	156
Jan 2025	21	3	3	27
Feb 2025	31	6	4	41
Mar 2025	9	2	0	11
Apr 2025	34	4	1	39
May 2025	20	8	1	29
Jun 2025	30	1	5	36
Jul 2025	19	7	0	26
Aug 2025	107	5	0	112
Sep 2025	279	5	1	285

Cumulative views and downloads (calculated since 18 Nov 2024)

Month	HTML	PDF	XML	Total
Nov 2024	74	10	5	89
Dec 2024	104	50	2	156
Jan 2025	21	3	3	27
Feb 2025	31	6	4	41
Mar 2025	9	2	0	11
Apr 2025	34	4	1	39
May 2025	20	8	1	29
Jun 2025	30	1	5	36
Jul 2025	19	7	0	26
Aug 2025	107	5	0	112
Sep 2025	279	5	1	285

Viewed (geographical distribution)

Total article views: 825 (including HTML, PDF, and XML) Thereof 825 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 23 Sep 2025

Download

The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.

Preprint (1092 KB)
Metadata XML

Short summary

The decline in aerosols in China altered temperature and pressure gradients between the two hemispheres, which intensified the Southern Trade Winds and caused a moisture divergence in Australia, resulting in less rainfall. Reduced surface moisture led to more energy being converted into sensible heat instead of evaporating as latent heat, which raised near-surface temperatures. Our findings offer insights for managing drought and wildfire risks in Australia.

Dry and warm conditions in Australia exacerbated by aerosol reduction in China

Journal article(s) based on this preprint

Interactive discussion

Interactive discussion

Peer review completion

Suggestions for revision or reasons for rejection

Suggestions for revision or reasons for rejection

Journal article(s) based on this preprint

Supplement

Viewed

Viewed (geographical distribution)


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