the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Feb 2025
Impact of volcanic sulfate aerosols on the stratospheric heating: implications on the Quasi-Biennial Oscillation
Abstract. Large and moderate volcanic eruptions significantly impact Earth's atmosphere by releasing sulphur emissions, thereby affecting atmospheric dynamics and QBO. Using the ECHAM6-HAMMOZ model, we show the impact of eruptive volcanoes on the tropical stratosphere and Quasi-biennial oscillation (QBO) from 2001 to 2013. Our simulations with volcanoes, when compared without volcanoes, show that volcanic sulfate aerosols enhance the stratospheric aerosol optical depth (SAOD) two months after the eruption of Rabaul (0.0034); Sarychev (0.0040) and Nabro (0.0097). The enhanced SOAD in the tropics (0.0014) led to a radiative forcing at the top of the atmosphere (TOA) by -0.92±0.34 W m-2 and at the surface by -0.88±0.18 W m-2 in the tropical region. The volcanic aerosol precursors enter the tropical stratosphere, propagating upward and enhancing sulfate aerosol concentrations by 46.95 ng m⁻³ and heating rates by 0.13±0.05×10⁻² K d⁻¹. The QBO estimated from model simulations using the wavelet analysis shows that stratospheric heating caused by the volcanoes reduces the amplitude of the QBO and disrupts its phases, resulting in the prolongation of the easterly phase by ~12 to 20 months and the westerly phase by ~16 to 24 months. The secondary meridional circulation induced by the QBO produces the double-peak structure in the amplitude near the equator, with peaks at 10 hPa and at 50 hPa. Our study points out that moderate and large volcanoes modulate the QBO. Since QBO also modulates tropical convection and weather, we suggest including volcanic eruptions and the QBO in weather prediction models for a better forecast.
Competing interests: More than one author is Editor of the journal 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.- Preprint
(1641 KB) - Metadata XML
-
Supplement
(245 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-3825', Flossie Brown, 16 Feb 2025
-
AC1: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-I
The concept of this study is very interesting, but as it stands, I am not convinced of the effect of recent volcanic activity on the QBO. The difference between the QBO in simulations with and without volcanic eruptions seems mostly to happen at the onset of the simulation, rather than in response to large volcanic activity happening after 2005. In my opinion, more analysis is needed to understand what is happening in the simulations and why. Here, I don’t find the conclusions sufficiently supported by the results. Furthermore, the final section of the study (Fig. 8) is a topic I am very interested in, but it is not clear what the findings of this work are or if any new conclusions have been reached. I think to meaningfully add to existing understanding, more than a simple lat-height anomaly is required. Beyond this, some statements would benefit from additional citations, and the supplementary could easily be extended to include the analysis suggested above. With these additions, I would find the study fits in the scope of the journal, and is well structured with a sufficient description of methods.
Reply: We thank the reviewer for careful reading and valuable suggestions. We have included additional analysis suggested by the reviewer in the revised manuscript. The changes are indicated at the line numbers mentioned in each reply and indicated in blue colour.
Major comments:
(1) I remain a bit confused with the statement ‘QBO is not included’ in the ECHAM model and whether that is an issue. Why did you choose to use this model over a model that has an internally generated QBO? What are the possible issues with not having the QBO and deriving it from wavelet analysis?
Reply (1): Thank you for the comment. The previously used ECHAM version with resolution T63L47 did not have internally simulated QBO. Since the reviewer pointed out we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results (L179-L186). This allows us to analyse QBO characteristics directly from the model output without depending on wavelet analysis.
(2) Could you compare observations to confirm your model results? The present day QBO reanalysis and observations are available.
Reply (2): Thank you for the suggestion. We compared the model-simulated QBO with ERA5 reanalysis for the period 2001–2013. A detailed description of this comparison is provided in Section 3.1 (L267-L275).
(3) I am unconvinced of the impact of the eruptions on the QBO and vertical winds. Why do larger eruptions not seem to have a larger effect on vertical velocity or the QBO? In addition to the vertical velocity anomaly, I would be interested to see the temperature anomaly and perhaps even the latitudinal temperature gradient.
Reply (3): This text is removed since it was creating confusion. In the previous version, we wanted to mention that heating caused by volcanoes will increase vertical velocity and therefore will increase the transport of sulfate aerosols in the UTLS region.
As suggested by the reviewer, we have plotted the temperature anomaly and the latitudinal temperature gradient (attached herewith). These figures show the influence of volcanoes on QBO in temperature and a double-peak structure at the equator. However, influence on temperature will divert the focus of the manuscript, hence not included in the manuscript. We plan to report it in the follow-up paper.
(4) The wavelet analysis and how it gets a QBO from the zonal wind is difficult for me to follow. Some of the mathematical terms are not defined and the purpose of each equation is not clearly explained.
Reply (4): In the revised manuscript, we performed the model simulation using a six-member ensemble mean approach and higher vertical resolution (L95), which is able to internally generate the QBO through resolved dynamics. Therefore, we have removed the wavelet analysis section from the manuscript and revised the text accordingly to improve clarity and focus on the model-simulated QBO.
(5) Fig. 6: There is almost no difference between sulfate burden in VAL and VAL0 between 2001 and 2005, however the vertical velocity increases right from the start of the simulation and doesn’t seem to increase any further after 2005 or after larger eruptions. The relationship between sulfate and vertical velocity is not that convincing to me.
Reply (5): As mentioned in reply 3. We have removed a figure description of vertical velocity since it was creating confusion.
(6) Fig. 7 is not intuitive to understand. It may benefit from showing VAL, VAL0 and the anomaly, since the anomaly is hard to interpret on its own. Panels c and d are also hard to understand. In Panel d, the anomaly and VAL0 look identical with a small shift, suggesting VAL0 is slightly out of sync with VAL from the start but not necessarily suggesting an increased QBO period over the course of the simulation. To indicate an impact of volcanic eruptions, I would expect the anomaly to have a different period to VAL0 and maybe even to change over time in response to different eruptions. I am happy to be proven wrong, but I think more work needs to be done to confirm the impact of the eruptions.
Reply (6): Thank you for your insightful comment. The high vertical resolution model simulations (T65L95), six-member ensemble simulations, with internally generated QBO, have changed the results. In the revised analysis, we have included only VAL0 and the anomalies (VAL–VAL0) to better highlight perturbations due to volcanic activity. These simulations show anomalies are out of phase with VAL0. This indicates a strong influence of volcanic eruptions. We have added this in the manuscript at L383-L393.
(7) The section describing Fig. 8 needs some work. What are the consequences of your finding that the zonal mean cross section changes? Is it important? Is it significant? Why explain the thermal-wind balance principles but not discuss or show figures of temperature or wind anomalies?
Reply (7): We thank the reviewer for highlighting this important point. In the revised manuscript, we have substantially improved the section discussing Fig. 8. As suggested, we have included a discussion of the associated temperature anomalies in the revised manuscript, ensuring that the explanation of the thermal-wind balance principles is well supported (L417-L469).
(8) The thermal wind relationship and the height-latitude structure of the QBO is related to the meridional temperature gradient. To explain the changes in Fig. 8, I would guess the sulfate aerosol has caused a change in the temperature gradient (Brown et al., 2023 has a detailed explanation). In general, I find the lat-height changes to the QBO of great interest but I am not sure of what message you are trying to convey here.
Reply (8): As explained in reply 7 this section is modified.
Minor comments
(9) L221: Why do you do the wavelet analysis on the anomaly? Surely this is not certain to have an amplitude or period.
Reply (9): In the revised manuscript, we have removed the wavelet analysis section. Since new model simulations with higher vertical resolution (T63L95) are able to generate a QBO internally.
(10) The model evaluation section 3.1 needs a slightly more nuanced analysis in my opinion. The baseline SAOD is underestimated by the model, but it simulates increases in response to volcanic eruptions that are much larger than observed. This could lead to overestimation of the volcanic effects. Adding a 1:1 line to Fig. b would demonstrate this.
Reply (10): Thank you for your helpful comment. In the revised manuscript, we have added a 1:1 line to Fig. 2b to better illustrate the relationship between observed and simulated SAOD. We acknowledge that the model slightly underestimates the baseline SAOD while showing stronger responses to volcanic eruptions. However, we would like to emphasize that our analysis is based on a six-member ensemble and uses a higher vertical resolution configuration (L95), which improves the representation of stratospheric processes. Therefore, we believe the simulated volcanic response is more realistic and not necessarily underestimated and overestimated. We have revised the discussion in Section 3.1 accordingly to reflect this nuance. (L256-L260).
(11) L286: Why would output on different levels create a bias?
Reply (11): Our Six-member ensemble simulations show good agreement with observations. Hence the above text is changed.
(12) L288: How do you know transport and resolution processes can produce the difference in your model vs GloSSAC?
Reply (12): Differences between the model and GloSSAC can arise because GloSSAC is based on assimilated satellite observations with relatively higher vertical resolution, whereas the model has a coarser vertical grid and simulates aerosol transport based on its own dynamics. Coarser resolution can smooth out fine vertical structures and gradients, while model transport processes can differ from the real atmosphere, especially after a volcanic injection.
(13) L298-306: Why do you just compare the year 2008? To me this is not the best way to decide if the QBO can be well represented. I would think the higher latitude zonal winds are not so important and may even compare a zonal mean zonal wind at all pressure levels, similar to Fig. 8.
Reply (13): We thank the reviewer for this valuable suggestion. In the revised manuscript, we have improved the QBO evaluation by comparing the model-simulated QBO from the VAL simulation against the ERA5 reanalysis over an extended period, 2001-2013, rather than focusing on a single year (2008). The discussion has been updated accordingly to highlight the similarities and differences between the model and reanalysis, focusing particularly on the tropical region where the QBO is strongest (L267-L275).
(14) L432: A disruption of 12 months between which time period? Or every cycle of the QBO is extended?
Reply (14): In the revised manuscript, we have updated the text to make this point clearer and avoid any ambiguity. (L386-L388).
(15) L434: Is this an observed phenomenon or just in your simulation?
Reply (15): We have removed the original sentence and revised the text for clarity
(16) L492: sentence doesn’t read well
Reply (16): We have corrected the sentence for clarity and improved the wording in the revised manuscript (L439-L441).
(17) L493: higher atmospheric pressure but lower atmospheric level
Reply (17): We have revised the text accordingly.
(18) L497-489: How does Fig. 8 add to this background knowledge? Is this paragraph relevant or is it better suited to the introduction?
Reply (18): We have kept this paragraph in the results section, as it directly supports the interpretation of Fig. 8. This background information is necessary here to help explain the observed double-peak structure in the zonal mean cross-section and the changes in the QBO-related circulation. Including this explanation in the results section ensures that readers can immediately understand the physical basis for the patterns seen in Fig. 8, rather than having to recall this information from the introduction (L424-L438).
Conclusion section:
(19) I think point 1 is fine as it is, but to be a great point perhaps you could add what you have done that is different to existing results on this topic.
Reply (19): In the revised manuscript, we have strengthened point 1 by emphasizing what is new in our study compared to previous work (L483-L485).
(20) Point 3, similar to an earlier comment: is 12 – 24 months an accumulated delay over the course of the simulation? There isn’t any evidence of responses from specific eruptions in this work.
Reply (20): In the revised manuscript, we have revised the conclusion and the related discussion (L488-L489). While specific eruption-related signals are difficult to isolate clearly in the current analysis.
(21) Point 4 is not a new finding of your study as far as I am aware. It is just a description of the zonal mean zonal wind at the equator when you average over a long period of time.
Reply (21): We have revised point-4 in the conclusion (L490-L491).
(22) For point 5, is ‘large’ impacts on weather perhaps an overstatement.
Reply (22): We have removed it.
Replies to Reviewer-II
(1) The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Reply (1): Thank you for your valuable comment. We have now performed six-member ensemble simulations at higher vertical resolution T63L95 so that QBO is simulated by the model. We have repeated the complete analysis using the new simulations. Details of the six-member simulations, their configuration, and the updated analysis have been included in the revised manuscript. The use of an ensemble helps to better account for natural interannual variability and improves the robustness of the results. (L179-L186)
Yes, we have used a double call radiative scheme to obtain Radiative forcing estimates. After using a six member-ensemble, the magnitude of radiative forcing is comparable with previous studies (e.g., Brühl et al., 2015; Schmidt et al., 2018).
(2) Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Reply (2): Previously, we used the ECHAM version with resolution T63L47, which did not have internally simulated QBO. Since the reviewer pointed out, we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results. We have repeated the complete analysis.
-
AC2: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-I
The concept of this study is very interesting, but as it stands, I am not convinced of the effect of recent volcanic activity on the QBO. The difference between the QBO in simulations with and without volcanic eruptions seems mostly to happen at the onset of the simulation, rather than in response to large volcanic activity happening after 2005. In my opinion, more analysis is needed to understand what is happening in the simulations and why. Here, I don’t find the conclusions sufficiently supported by the results. Furthermore, the final section of the study (Fig. 8) is a topic I am very interested in, but it is not clear what the findings of this work are or if any new conclusions have been reached. I think to meaningfully add to existing understanding, more than a simple lat-height anomaly is required. Beyond this, some statements would benefit from additional citations, and the supplementary could easily be extended to include the analysis suggested above. With these additions, I would find the study fits in the scope of the journal, and is well structured with a sufficient description of methods.
Reply: We thank the reviewer for careful reading and valuable suggestions. We have included additional analysis suggested by the reviewer in the revised manuscript. The changes are indicated at the line numbers mentioned in each reply and indicated in blue colour.
Major comments:
(1) I remain a bit confused with the statement ‘QBO is not included’ in the ECHAM model and whether that is an issue. Why did you choose to use this model over a model that has an internally generated QBO? What are the possible issues with not having the QBO and deriving it from wavelet analysis?
Reply (1): Thank you for the comment. The previously used ECHAM version with resolution T63L47 did not have internally simulated QBO. Since the reviewer pointed out we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results (L179-L186). This allows us to analyse QBO characteristics directly from the model output without depending on wavelet analysis.
(2) Could you compare observations to confirm your model results? The present day QBO reanalysis and observations are available.
Reply (2): Thank you for the suggestion. We compared the model-simulated QBO with ERA5 reanalysis for the period 2001–2013. A detailed description of this comparison is provided in Section 3.1 (L267-L275).
(3) I am unconvinced of the impact of the eruptions on the QBO and vertical winds. Why do larger eruptions not seem to have a larger effect on vertical velocity or the QBO? In addition to the vertical velocity anomaly, I would be interested to see the temperature anomaly and perhaps even the latitudinal temperature gradient.
Reply (3): This text is removed since it was creating confusion. In the previous version, we wanted to mention that heating caused by volcanoes will increase vertical velocity and therefore will increase the transport of sulfate aerosols in the UTLS region.
As suggested by the reviewer, we have plotted the temperature anomaly and the latitudinal temperature gradient (attached herewith). These figures show the influence of volcanoes on QBO in temperature and a double-peak structure at the equator. However, influence on temperature will divert the focus of the manuscript, hence not included in the manuscript. We plan to report it in the follow-up paper.
(4) The wavelet analysis and how it gets a QBO from the zonal wind is difficult for me to follow. Some of the mathematical terms are not defined and the purpose of each equation is not clearly explained.
Reply (4): In the revised manuscript, we performed the model simulation using a six-member ensemble mean approach and higher vertical resolution (L95), which is able to internally generate the QBO through resolved dynamics. Therefore, we have removed the wavelet analysis section from the manuscript and revised the text accordingly to improve clarity and focus on the model-simulated QBO.
(5) Fig. 6: There is almost no difference between sulfate burden in VAL and VAL0 between 2001 and 2005, however the vertical velocity increases right from the start of the simulation and doesn’t seem to increase any further after 2005 or after larger eruptions. The relationship between sulfate and vertical velocity is not that convincing to me.
Reply (5): As mentioned in reply 3. We have removed a figure description of vertical velocity since it was creating confusion.
(6) Fig. 7 is not intuitive to understand. It may benefit from showing VAL, VAL0 and the anomaly, since the anomaly is hard to interpret on its own. Panels c and d are also hard to understand. In Panel d, the anomaly and VAL0 look identical with a small shift, suggesting VAL0 is slightly out of sync with VAL from the start but not necessarily suggesting an increased QBO period over the course of the simulation. To indicate an impact of volcanic eruptions, I would expect the anomaly to have a different period to VAL0 and maybe even to change over time in response to different eruptions. I am happy to be proven wrong, but I think more work needs to be done to confirm the impact of the eruptions.
Reply (6): Thank you for your insightful comment. The high vertical resolution model simulations (T65L95), six-member ensemble simulations, with internally generated QBO, have changed the results. In the revised analysis, we have included only VAL0 and the anomalies (VAL–VAL0) to better highlight perturbations due to volcanic activity. These simulations show anomalies are out of phase with VAL0. This indicates a strong influence of volcanic eruptions. We have added this in the manuscript at L383-L393.
(7) The section describing Fig. 8 needs some work. What are the consequences of your finding that the zonal mean cross section changes? Is it important? Is it significant? Why explain the thermal-wind balance principles but not discuss or show figures of temperature or wind anomalies?
Reply (7): We thank the reviewer for highlighting this important point. In the revised manuscript, we have substantially improved the section discussing Fig. 8. As suggested, we have included a discussion of the associated temperature anomalies in the revised manuscript, ensuring that the explanation of the thermal-wind balance principles is well supported (L417-L469).
(8) The thermal wind relationship and the height-latitude structure of the QBO is related to the meridional temperature gradient. To explain the changes in Fig. 8, I would guess the sulfate aerosol has caused a change in the temperature gradient (Brown et al., 2023 has a detailed explanation). In general, I find the lat-height changes to the QBO of great interest but I am not sure of what message you are trying to convey here.
Reply (8): As explained in reply 7 this section is modified.
Minor comments
(9) L221: Why do you do the wavelet analysis on the anomaly? Surely this is not certain to have an amplitude or period.
Reply (9): In the revised manuscript, we have removed the wavelet analysis section. Since new model simulations with higher vertical resolution (T63L95) are able to generate a QBO internally.
(10) The model evaluation section 3.1 needs a slightly more nuanced analysis in my opinion. The baseline SAOD is underestimated by the model, but it simulates increases in response to volcanic eruptions that are much larger than observed. This could lead to overestimation of the volcanic effects. Adding a 1:1 line to Fig. b would demonstrate this.
Reply (10): Thank you for your helpful comment. In the revised manuscript, we have added a 1:1 line to Fig. 2b to better illustrate the relationship between observed and simulated SAOD. We acknowledge that the model slightly underestimates the baseline SAOD while showing stronger responses to volcanic eruptions. However, we would like to emphasize that our analysis is based on a six-member ensemble and uses a higher vertical resolution configuration (L95), which improves the representation of stratospheric processes. Therefore, we believe the simulated volcanic response is more realistic and not necessarily underestimated and overestimated. We have revised the discussion in Section 3.1 accordingly to reflect this nuance. (L256-L260).
(11) L286: Why would output on different levels create a bias?
Reply (11): Our Six-member ensemble simulations show good agreement with observations. Hence the above text is changed.
(12) L288: How do you know transport and resolution processes can produce the difference in your model vs GloSSAC?
Reply (12): Differences between the model and GloSSAC can arise because GloSSAC is based on assimilated satellite observations with relatively higher vertical resolution, whereas the model has a coarser vertical grid and simulates aerosol transport based on its own dynamics. Coarser resolution can smooth out fine vertical structures and gradients, while model transport processes can differ from the real atmosphere, especially after a volcanic injection.
(13) L298-306: Why do you just compare the year 2008? To me this is not the best way to decide if the QBO can be well represented. I would think the higher latitude zonal winds are not so important and may even compare a zonal mean zonal wind at all pressure levels, similar to Fig. 8.
Reply (13): We thank the reviewer for this valuable suggestion. In the revised manuscript, we have improved the QBO evaluation by comparing the model-simulated QBO from the VAL simulation against the ERA5 reanalysis over an extended period, 2001-2013, rather than focusing on a single year (2008). The discussion has been updated accordingly to highlight the similarities and differences between the model and reanalysis, focusing particularly on the tropical region where the QBO is strongest (L267-L275).
(14) L432: A disruption of 12 months between which time period? Or every cycle of the QBO is extended?
Reply (14): In the revised manuscript, we have updated the text to make this point clearer and avoid any ambiguity. (L386-L388).
(15) L434: Is this an observed phenomenon or just in your simulation?
Reply (15): We have removed the original sentence and revised the text for clarity
(16) L492: sentence doesn’t read well
Reply (16): We have corrected the sentence for clarity and improved the wording in the revised manuscript (L439-L441).
(17) L493: higher atmospheric pressure but lower atmospheric level
Reply (17): We have revised the text accordingly.
(18) L497-489: How does Fig. 8 add to this background knowledge? Is this paragraph relevant or is it better suited to the introduction?
Reply (18): We have kept this paragraph in the results section, as it directly supports the interpretation of Fig. 8. This background information is necessary here to help explain the observed double-peak structure in the zonal mean cross-section and the changes in the QBO-related circulation. Including this explanation in the results section ensures that readers can immediately understand the physical basis for the patterns seen in Fig. 8, rather than having to recall this information from the introduction (L424-L438).
Conclusion section:
(19) I think point 1 is fine as it is, but to be a great point perhaps you could add what you have done that is different to existing results on this topic.
Reply (19): In the revised manuscript, we have strengthened point 1 by emphasizing what is new in our study compared to previous work (L483-L485).
(20) Point 3, similar to an earlier comment: is 12 – 24 months an accumulated delay over the course of the simulation? There isn’t any evidence of responses from specific eruptions in this work.
Reply (20): In the revised manuscript, we have revised the conclusion and the related discussion (L488-L489). While specific eruption-related signals are difficult to isolate clearly in the current analysis.
(21) Point 4 is not a new finding of your study as far as I am aware. It is just a description of the zonal mean zonal wind at the equator when you average over a long period of time.
Reply (21): We have revised point-4 in the conclusion (L490-L491).
(22) For point 5, is ‘large’ impacts on weather perhaps an overstatement.
Reply (22): We have removed it.
-
AC1: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
-
RC2: 'Comment on egusphere-2024-3825', Anonymous Referee #2, 17 Feb 2025
The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Citation: https://doi.org/10.5194/egusphere-2024-3825-RC2 -
AC3: 'Reply on RC2', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-II
(1) The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Reply (1): Thank you for your valuable comment. We have now performed six-member ensemble simulations at higher vertical resolution T63L95 so that QBO is simulated by the model. We have repeated the complete analysis using the new simulations. Details of the six-member simulations, their configuration, and the updated analysis have been included in the revised manuscript. The use of an ensemble helps to better account for natural interannual variability and improves the robustness of the results. (L179-L186)
Yes, we have used a double call radiative scheme to obtain Radiative forcing estimates. After using a six member-ensemble, the magnitude of radiative forcing is comparable with previous studies (e.g., Brühl et al., 2015; Schmidt et al., 2018).
(2) Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Reply (2): Previously, we used the ECHAM version with resolution T63L47, which did not have internally simulated QBO. Since the reviewer pointed out, we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results. We have repeated the complete analysis.
Citation: https://doi.org/10.5194/egusphere-2024-3825-AC3
-
AC3: 'Reply on RC2', Suvarna Fadnavis, 09 May 2025
Status: closed
-
RC1: 'Comment on egusphere-2024-3825', Flossie Brown, 16 Feb 2025
-
AC1: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-I
The concept of this study is very interesting, but as it stands, I am not convinced of the effect of recent volcanic activity on the QBO. The difference between the QBO in simulations with and without volcanic eruptions seems mostly to happen at the onset of the simulation, rather than in response to large volcanic activity happening after 2005. In my opinion, more analysis is needed to understand what is happening in the simulations and why. Here, I don’t find the conclusions sufficiently supported by the results. Furthermore, the final section of the study (Fig. 8) is a topic I am very interested in, but it is not clear what the findings of this work are or if any new conclusions have been reached. I think to meaningfully add to existing understanding, more than a simple lat-height anomaly is required. Beyond this, some statements would benefit from additional citations, and the supplementary could easily be extended to include the analysis suggested above. With these additions, I would find the study fits in the scope of the journal, and is well structured with a sufficient description of methods.
Reply: We thank the reviewer for careful reading and valuable suggestions. We have included additional analysis suggested by the reviewer in the revised manuscript. The changes are indicated at the line numbers mentioned in each reply and indicated in blue colour.
Major comments:
(1) I remain a bit confused with the statement ‘QBO is not included’ in the ECHAM model and whether that is an issue. Why did you choose to use this model over a model that has an internally generated QBO? What are the possible issues with not having the QBO and deriving it from wavelet analysis?
Reply (1): Thank you for the comment. The previously used ECHAM version with resolution T63L47 did not have internally simulated QBO. Since the reviewer pointed out we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results (L179-L186). This allows us to analyse QBO characteristics directly from the model output without depending on wavelet analysis.
(2) Could you compare observations to confirm your model results? The present day QBO reanalysis and observations are available.
Reply (2): Thank you for the suggestion. We compared the model-simulated QBO with ERA5 reanalysis for the period 2001–2013. A detailed description of this comparison is provided in Section 3.1 (L267-L275).
(3) I am unconvinced of the impact of the eruptions on the QBO and vertical winds. Why do larger eruptions not seem to have a larger effect on vertical velocity or the QBO? In addition to the vertical velocity anomaly, I would be interested to see the temperature anomaly and perhaps even the latitudinal temperature gradient.
Reply (3): This text is removed since it was creating confusion. In the previous version, we wanted to mention that heating caused by volcanoes will increase vertical velocity and therefore will increase the transport of sulfate aerosols in the UTLS region.
As suggested by the reviewer, we have plotted the temperature anomaly and the latitudinal temperature gradient (attached herewith). These figures show the influence of volcanoes on QBO in temperature and a double-peak structure at the equator. However, influence on temperature will divert the focus of the manuscript, hence not included in the manuscript. We plan to report it in the follow-up paper.
(4) The wavelet analysis and how it gets a QBO from the zonal wind is difficult for me to follow. Some of the mathematical terms are not defined and the purpose of each equation is not clearly explained.
Reply (4): In the revised manuscript, we performed the model simulation using a six-member ensemble mean approach and higher vertical resolution (L95), which is able to internally generate the QBO through resolved dynamics. Therefore, we have removed the wavelet analysis section from the manuscript and revised the text accordingly to improve clarity and focus on the model-simulated QBO.
(5) Fig. 6: There is almost no difference between sulfate burden in VAL and VAL0 between 2001 and 2005, however the vertical velocity increases right from the start of the simulation and doesn’t seem to increase any further after 2005 or after larger eruptions. The relationship between sulfate and vertical velocity is not that convincing to me.
Reply (5): As mentioned in reply 3. We have removed a figure description of vertical velocity since it was creating confusion.
(6) Fig. 7 is not intuitive to understand. It may benefit from showing VAL, VAL0 and the anomaly, since the anomaly is hard to interpret on its own. Panels c and d are also hard to understand. In Panel d, the anomaly and VAL0 look identical with a small shift, suggesting VAL0 is slightly out of sync with VAL from the start but not necessarily suggesting an increased QBO period over the course of the simulation. To indicate an impact of volcanic eruptions, I would expect the anomaly to have a different period to VAL0 and maybe even to change over time in response to different eruptions. I am happy to be proven wrong, but I think more work needs to be done to confirm the impact of the eruptions.
Reply (6): Thank you for your insightful comment. The high vertical resolution model simulations (T65L95), six-member ensemble simulations, with internally generated QBO, have changed the results. In the revised analysis, we have included only VAL0 and the anomalies (VAL–VAL0) to better highlight perturbations due to volcanic activity. These simulations show anomalies are out of phase with VAL0. This indicates a strong influence of volcanic eruptions. We have added this in the manuscript at L383-L393.
(7) The section describing Fig. 8 needs some work. What are the consequences of your finding that the zonal mean cross section changes? Is it important? Is it significant? Why explain the thermal-wind balance principles but not discuss or show figures of temperature or wind anomalies?
Reply (7): We thank the reviewer for highlighting this important point. In the revised manuscript, we have substantially improved the section discussing Fig. 8. As suggested, we have included a discussion of the associated temperature anomalies in the revised manuscript, ensuring that the explanation of the thermal-wind balance principles is well supported (L417-L469).
(8) The thermal wind relationship and the height-latitude structure of the QBO is related to the meridional temperature gradient. To explain the changes in Fig. 8, I would guess the sulfate aerosol has caused a change in the temperature gradient (Brown et al., 2023 has a detailed explanation). In general, I find the lat-height changes to the QBO of great interest but I am not sure of what message you are trying to convey here.
Reply (8): As explained in reply 7 this section is modified.
Minor comments
(9) L221: Why do you do the wavelet analysis on the anomaly? Surely this is not certain to have an amplitude or period.
Reply (9): In the revised manuscript, we have removed the wavelet analysis section. Since new model simulations with higher vertical resolution (T63L95) are able to generate a QBO internally.
(10) The model evaluation section 3.1 needs a slightly more nuanced analysis in my opinion. The baseline SAOD is underestimated by the model, but it simulates increases in response to volcanic eruptions that are much larger than observed. This could lead to overestimation of the volcanic effects. Adding a 1:1 line to Fig. b would demonstrate this.
Reply (10): Thank you for your helpful comment. In the revised manuscript, we have added a 1:1 line to Fig. 2b to better illustrate the relationship between observed and simulated SAOD. We acknowledge that the model slightly underestimates the baseline SAOD while showing stronger responses to volcanic eruptions. However, we would like to emphasize that our analysis is based on a six-member ensemble and uses a higher vertical resolution configuration (L95), which improves the representation of stratospheric processes. Therefore, we believe the simulated volcanic response is more realistic and not necessarily underestimated and overestimated. We have revised the discussion in Section 3.1 accordingly to reflect this nuance. (L256-L260).
(11) L286: Why would output on different levels create a bias?
Reply (11): Our Six-member ensemble simulations show good agreement with observations. Hence the above text is changed.
(12) L288: How do you know transport and resolution processes can produce the difference in your model vs GloSSAC?
Reply (12): Differences between the model and GloSSAC can arise because GloSSAC is based on assimilated satellite observations with relatively higher vertical resolution, whereas the model has a coarser vertical grid and simulates aerosol transport based on its own dynamics. Coarser resolution can smooth out fine vertical structures and gradients, while model transport processes can differ from the real atmosphere, especially after a volcanic injection.
(13) L298-306: Why do you just compare the year 2008? To me this is not the best way to decide if the QBO can be well represented. I would think the higher latitude zonal winds are not so important and may even compare a zonal mean zonal wind at all pressure levels, similar to Fig. 8.
Reply (13): We thank the reviewer for this valuable suggestion. In the revised manuscript, we have improved the QBO evaluation by comparing the model-simulated QBO from the VAL simulation against the ERA5 reanalysis over an extended period, 2001-2013, rather than focusing on a single year (2008). The discussion has been updated accordingly to highlight the similarities and differences between the model and reanalysis, focusing particularly on the tropical region where the QBO is strongest (L267-L275).
(14) L432: A disruption of 12 months between which time period? Or every cycle of the QBO is extended?
Reply (14): In the revised manuscript, we have updated the text to make this point clearer and avoid any ambiguity. (L386-L388).
(15) L434: Is this an observed phenomenon or just in your simulation?
Reply (15): We have removed the original sentence and revised the text for clarity
(16) L492: sentence doesn’t read well
Reply (16): We have corrected the sentence for clarity and improved the wording in the revised manuscript (L439-L441).
(17) L493: higher atmospheric pressure but lower atmospheric level
Reply (17): We have revised the text accordingly.
(18) L497-489: How does Fig. 8 add to this background knowledge? Is this paragraph relevant or is it better suited to the introduction?
Reply (18): We have kept this paragraph in the results section, as it directly supports the interpretation of Fig. 8. This background information is necessary here to help explain the observed double-peak structure in the zonal mean cross-section and the changes in the QBO-related circulation. Including this explanation in the results section ensures that readers can immediately understand the physical basis for the patterns seen in Fig. 8, rather than having to recall this information from the introduction (L424-L438).
Conclusion section:
(19) I think point 1 is fine as it is, but to be a great point perhaps you could add what you have done that is different to existing results on this topic.
Reply (19): In the revised manuscript, we have strengthened point 1 by emphasizing what is new in our study compared to previous work (L483-L485).
(20) Point 3, similar to an earlier comment: is 12 – 24 months an accumulated delay over the course of the simulation? There isn’t any evidence of responses from specific eruptions in this work.
Reply (20): In the revised manuscript, we have revised the conclusion and the related discussion (L488-L489). While specific eruption-related signals are difficult to isolate clearly in the current analysis.
(21) Point 4 is not a new finding of your study as far as I am aware. It is just a description of the zonal mean zonal wind at the equator when you average over a long period of time.
Reply (21): We have revised point-4 in the conclusion (L490-L491).
(22) For point 5, is ‘large’ impacts on weather perhaps an overstatement.
Reply (22): We have removed it.
Replies to Reviewer-II
(1) The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Reply (1): Thank you for your valuable comment. We have now performed six-member ensemble simulations at higher vertical resolution T63L95 so that QBO is simulated by the model. We have repeated the complete analysis using the new simulations. Details of the six-member simulations, their configuration, and the updated analysis have been included in the revised manuscript. The use of an ensemble helps to better account for natural interannual variability and improves the robustness of the results. (L179-L186)
Yes, we have used a double call radiative scheme to obtain Radiative forcing estimates. After using a six member-ensemble, the magnitude of radiative forcing is comparable with previous studies (e.g., Brühl et al., 2015; Schmidt et al., 2018).
(2) Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Reply (2): Previously, we used the ECHAM version with resolution T63L47, which did not have internally simulated QBO. Since the reviewer pointed out, we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results. We have repeated the complete analysis.
-
AC2: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-I
The concept of this study is very interesting, but as it stands, I am not convinced of the effect of recent volcanic activity on the QBO. The difference between the QBO in simulations with and without volcanic eruptions seems mostly to happen at the onset of the simulation, rather than in response to large volcanic activity happening after 2005. In my opinion, more analysis is needed to understand what is happening in the simulations and why. Here, I don’t find the conclusions sufficiently supported by the results. Furthermore, the final section of the study (Fig. 8) is a topic I am very interested in, but it is not clear what the findings of this work are or if any new conclusions have been reached. I think to meaningfully add to existing understanding, more than a simple lat-height anomaly is required. Beyond this, some statements would benefit from additional citations, and the supplementary could easily be extended to include the analysis suggested above. With these additions, I would find the study fits in the scope of the journal, and is well structured with a sufficient description of methods.
Reply: We thank the reviewer for careful reading and valuable suggestions. We have included additional analysis suggested by the reviewer in the revised manuscript. The changes are indicated at the line numbers mentioned in each reply and indicated in blue colour.
Major comments:
(1) I remain a bit confused with the statement ‘QBO is not included’ in the ECHAM model and whether that is an issue. Why did you choose to use this model over a model that has an internally generated QBO? What are the possible issues with not having the QBO and deriving it from wavelet analysis?
Reply (1): Thank you for the comment. The previously used ECHAM version with resolution T63L47 did not have internally simulated QBO. Since the reviewer pointed out we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results (L179-L186). This allows us to analyse QBO characteristics directly from the model output without depending on wavelet analysis.
(2) Could you compare observations to confirm your model results? The present day QBO reanalysis and observations are available.
Reply (2): Thank you for the suggestion. We compared the model-simulated QBO with ERA5 reanalysis for the period 2001–2013. A detailed description of this comparison is provided in Section 3.1 (L267-L275).
(3) I am unconvinced of the impact of the eruptions on the QBO and vertical winds. Why do larger eruptions not seem to have a larger effect on vertical velocity or the QBO? In addition to the vertical velocity anomaly, I would be interested to see the temperature anomaly and perhaps even the latitudinal temperature gradient.
Reply (3): This text is removed since it was creating confusion. In the previous version, we wanted to mention that heating caused by volcanoes will increase vertical velocity and therefore will increase the transport of sulfate aerosols in the UTLS region.
As suggested by the reviewer, we have plotted the temperature anomaly and the latitudinal temperature gradient (attached herewith). These figures show the influence of volcanoes on QBO in temperature and a double-peak structure at the equator. However, influence on temperature will divert the focus of the manuscript, hence not included in the manuscript. We plan to report it in the follow-up paper.
(4) The wavelet analysis and how it gets a QBO from the zonal wind is difficult for me to follow. Some of the mathematical terms are not defined and the purpose of each equation is not clearly explained.
Reply (4): In the revised manuscript, we performed the model simulation using a six-member ensemble mean approach and higher vertical resolution (L95), which is able to internally generate the QBO through resolved dynamics. Therefore, we have removed the wavelet analysis section from the manuscript and revised the text accordingly to improve clarity and focus on the model-simulated QBO.
(5) Fig. 6: There is almost no difference between sulfate burden in VAL and VAL0 between 2001 and 2005, however the vertical velocity increases right from the start of the simulation and doesn’t seem to increase any further after 2005 or after larger eruptions. The relationship between sulfate and vertical velocity is not that convincing to me.
Reply (5): As mentioned in reply 3. We have removed a figure description of vertical velocity since it was creating confusion.
(6) Fig. 7 is not intuitive to understand. It may benefit from showing VAL, VAL0 and the anomaly, since the anomaly is hard to interpret on its own. Panels c and d are also hard to understand. In Panel d, the anomaly and VAL0 look identical with a small shift, suggesting VAL0 is slightly out of sync with VAL from the start but not necessarily suggesting an increased QBO period over the course of the simulation. To indicate an impact of volcanic eruptions, I would expect the anomaly to have a different period to VAL0 and maybe even to change over time in response to different eruptions. I am happy to be proven wrong, but I think more work needs to be done to confirm the impact of the eruptions.
Reply (6): Thank you for your insightful comment. The high vertical resolution model simulations (T65L95), six-member ensemble simulations, with internally generated QBO, have changed the results. In the revised analysis, we have included only VAL0 and the anomalies (VAL–VAL0) to better highlight perturbations due to volcanic activity. These simulations show anomalies are out of phase with VAL0. This indicates a strong influence of volcanic eruptions. We have added this in the manuscript at L383-L393.
(7) The section describing Fig. 8 needs some work. What are the consequences of your finding that the zonal mean cross section changes? Is it important? Is it significant? Why explain the thermal-wind balance principles but not discuss or show figures of temperature or wind anomalies?
Reply (7): We thank the reviewer for highlighting this important point. In the revised manuscript, we have substantially improved the section discussing Fig. 8. As suggested, we have included a discussion of the associated temperature anomalies in the revised manuscript, ensuring that the explanation of the thermal-wind balance principles is well supported (L417-L469).
(8) The thermal wind relationship and the height-latitude structure of the QBO is related to the meridional temperature gradient. To explain the changes in Fig. 8, I would guess the sulfate aerosol has caused a change in the temperature gradient (Brown et al., 2023 has a detailed explanation). In general, I find the lat-height changes to the QBO of great interest but I am not sure of what message you are trying to convey here.
Reply (8): As explained in reply 7 this section is modified.
Minor comments
(9) L221: Why do you do the wavelet analysis on the anomaly? Surely this is not certain to have an amplitude or period.
Reply (9): In the revised manuscript, we have removed the wavelet analysis section. Since new model simulations with higher vertical resolution (T63L95) are able to generate a QBO internally.
(10) The model evaluation section 3.1 needs a slightly more nuanced analysis in my opinion. The baseline SAOD is underestimated by the model, but it simulates increases in response to volcanic eruptions that are much larger than observed. This could lead to overestimation of the volcanic effects. Adding a 1:1 line to Fig. b would demonstrate this.
Reply (10): Thank you for your helpful comment. In the revised manuscript, we have added a 1:1 line to Fig. 2b to better illustrate the relationship between observed and simulated SAOD. We acknowledge that the model slightly underestimates the baseline SAOD while showing stronger responses to volcanic eruptions. However, we would like to emphasize that our analysis is based on a six-member ensemble and uses a higher vertical resolution configuration (L95), which improves the representation of stratospheric processes. Therefore, we believe the simulated volcanic response is more realistic and not necessarily underestimated and overestimated. We have revised the discussion in Section 3.1 accordingly to reflect this nuance. (L256-L260).
(11) L286: Why would output on different levels create a bias?
Reply (11): Our Six-member ensemble simulations show good agreement with observations. Hence the above text is changed.
(12) L288: How do you know transport and resolution processes can produce the difference in your model vs GloSSAC?
Reply (12): Differences between the model and GloSSAC can arise because GloSSAC is based on assimilated satellite observations with relatively higher vertical resolution, whereas the model has a coarser vertical grid and simulates aerosol transport based on its own dynamics. Coarser resolution can smooth out fine vertical structures and gradients, while model transport processes can differ from the real atmosphere, especially after a volcanic injection.
(13) L298-306: Why do you just compare the year 2008? To me this is not the best way to decide if the QBO can be well represented. I would think the higher latitude zonal winds are not so important and may even compare a zonal mean zonal wind at all pressure levels, similar to Fig. 8.
Reply (13): We thank the reviewer for this valuable suggestion. In the revised manuscript, we have improved the QBO evaluation by comparing the model-simulated QBO from the VAL simulation against the ERA5 reanalysis over an extended period, 2001-2013, rather than focusing on a single year (2008). The discussion has been updated accordingly to highlight the similarities and differences between the model and reanalysis, focusing particularly on the tropical region where the QBO is strongest (L267-L275).
(14) L432: A disruption of 12 months between which time period? Or every cycle of the QBO is extended?
Reply (14): In the revised manuscript, we have updated the text to make this point clearer and avoid any ambiguity. (L386-L388).
(15) L434: Is this an observed phenomenon or just in your simulation?
Reply (15): We have removed the original sentence and revised the text for clarity
(16) L492: sentence doesn’t read well
Reply (16): We have corrected the sentence for clarity and improved the wording in the revised manuscript (L439-L441).
(17) L493: higher atmospheric pressure but lower atmospheric level
Reply (17): We have revised the text accordingly.
(18) L497-489: How does Fig. 8 add to this background knowledge? Is this paragraph relevant or is it better suited to the introduction?
Reply (18): We have kept this paragraph in the results section, as it directly supports the interpretation of Fig. 8. This background information is necessary here to help explain the observed double-peak structure in the zonal mean cross-section and the changes in the QBO-related circulation. Including this explanation in the results section ensures that readers can immediately understand the physical basis for the patterns seen in Fig. 8, rather than having to recall this information from the introduction (L424-L438).
Conclusion section:
(19) I think point 1 is fine as it is, but to be a great point perhaps you could add what you have done that is different to existing results on this topic.
Reply (19): In the revised manuscript, we have strengthened point 1 by emphasizing what is new in our study compared to previous work (L483-L485).
(20) Point 3, similar to an earlier comment: is 12 – 24 months an accumulated delay over the course of the simulation? There isn’t any evidence of responses from specific eruptions in this work.
Reply (20): In the revised manuscript, we have revised the conclusion and the related discussion (L488-L489). While specific eruption-related signals are difficult to isolate clearly in the current analysis.
(21) Point 4 is not a new finding of your study as far as I am aware. It is just a description of the zonal mean zonal wind at the equator when you average over a long period of time.
Reply (21): We have revised point-4 in the conclusion (L490-L491).
(22) For point 5, is ‘large’ impacts on weather perhaps an overstatement.
Reply (22): We have removed it.
-
AC1: 'Reply on RC1', Suvarna Fadnavis, 09 May 2025
-
RC2: 'Comment on egusphere-2024-3825', Anonymous Referee #2, 17 Feb 2025
The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Citation: https://doi.org/10.5194/egusphere-2024-3825-RC2 -
AC3: 'Reply on RC2', Suvarna Fadnavis, 09 May 2025
Replies to Reviewer-II
(1) The authors investigate the impact of small to moderate volcanic eruptions between 2001 – 2013 on the stratospheric aerosol layer, radiative forcing and QBO. While the topic is important and the paper relatively well written, I’m afraid that the approach the authors chose does not allow to gain meaningful results that warrant publication in the current state.
In particular, if I understand correctly, the authors performed one single simulation with volcanic eruptions and a second simulation without them, and inferred the role of aerosols by looking at differences between the two cases. While such an approach might just about be enough to obtain changes in stratospheric AOD and aerosols, it is nowhere near sufficient to infer impacts on the radiative forcing or stratospheric circulation / QBO, where the results reflect mostly changes caused by natural interannual variability (in meteorological conditions, clouds.. etc) and not a true impact of volcanic activity. And so, their results are very misleading, as they suggest relatively small changes in stratospheric aerosols can have an unproportionally large impact on the radiative budget and atmospheric circulation. E.g. their radiative forcing results are almost a factor of 10 larger than previously estimated (Bruhl et al., 2015; Schmidt et al., 2018). And while the other studies also performed only one simulation, they either estimated radiative effects from a double call to the radiative scheme that isolates the impacts of aerosols (Bruhl et al., 2015) or performed simulations using specified dynamics setup, i.e. where meteorology is nudged to observed conditions (Schmidt et al., 2018). Given the focus here was to, among other, look at impacts on QBO, I understand why the authors decided to use free-running simulations (and not nudged ones). However, a single simulation over a relatively short period (~14 years) is not sufficient to obtain that result with any confidence, where instead a much larger ensemble size would be needed.
Therefore, I’m afraid I cannot recommend a publication in the current form, at least not until the authors perform a number of additional ensemble members (5? 10?) and redo the analysis using a much larger data set. I’d also recommend the use of double call to radiative scheme to obtain values of radiative forcing changes (if possible).
Reply (1): Thank you for your valuable comment. We have now performed six-member ensemble simulations at higher vertical resolution T63L95 so that QBO is simulated by the model. We have repeated the complete analysis using the new simulations. Details of the six-member simulations, their configuration, and the updated analysis have been included in the revised manuscript. The use of an ensemble helps to better account for natural interannual variability and improves the robustness of the results. (L179-L186)
Yes, we have used a double call radiative scheme to obtain Radiative forcing estimates. After using a six member-ensemble, the magnitude of radiative forcing is comparable with previous studies (e.g., Brühl et al., 2015; Schmidt et al., 2018).
(2) Finally, on line 222-223, the authors state that “QBO is not included in the ECHAM6 HAMMOZ model used in this study”. I’m assuming it’s a typo or incorrect phrasing, because if the model doesn’t simulate QBO, then I’m not sure what is analyzed in this study. Please clarify.
Reply (2): Previously, we used the ECHAM version with resolution T63L47, which did not have internally simulated QBO. Since the reviewer pointed out, we made simulations at T63L95 resolution wherein QBO is internally simulated in the model. We performed six-member simulations to obtain the ensemble mean for robust results. We have repeated the complete analysis.
Citation: https://doi.org/10.5194/egusphere-2024-3825-AC3
-
AC3: 'Reply on RC2', Suvarna Fadnavis, 09 May 2025
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 440 | 91 | 26 | 557 | 38 | 25 | 34 |
- HTML: 440
- PDF: 91
- XML: 26
- Total: 557
- Supplement: 38
- BibTeX: 25
- EndNote: 34
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Review PDF attached.