the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Apr 2024
The influence of the Atlantic Multidecadal Variability on Storm Babet-like events
Abstract. In October 2023 Storm Babet led to extreme flooding and strong winds across the United Kingdom. We use atmospheric flow analogues to assess multidecadal variability in events similar to Storm Babet. We show that comparing analogues for timeslices results are highly sensitive to the chosen periods, thus we instead assess analogue trends through time. We identify a possible link between Storm Babet-like events and the Atlantic Multidecadal Variability (AMV), supporting the hypothesis that positive AMV leads to stormier western European weather. Events similar to Storm Babet are 7.5 times more likely during positive AMV. The method presented could be developed for use in the attribution of extreme weather events, allowing identification of possible causes of changes in the similarity of analogues to an extreme event through time. Increasing our understanding of the causes of extreme weather events can allow us to better predict future changes in such events, allowing society to prepare and adapt for the future.
- Preprint
(872 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-1136', Anonymous Referee #1, 06 Jun 2024
-
AC2: 'Reply on RC1', Vikki Thompson, 10 Sep 2024
We thank the reviewer for the detailed comments, and are pleased we have conveyed the importance of the research. We have revised the manuscript based on the comments, and those of another reviewer, and hope the reviewer finds it improved.
Please find attached point-by-point responses, with the changes to the manuscript highlighted. We have made many changes, including further statistical testing to show the robustness of the results, and a more detailed comparison with other scientific literature on the AMV and European storms to help support the case for causality.
On behalf of all authors,
Vikki Thompson
-
AC2: 'Reply on RC1', Vikki Thompson, 10 Sep 2024
-
RC2: 'Comment on egusphere-2024-1136', Anonymous Referee #2, 10 Jun 2024
The draft focuses on storm Babet and its analogs in terms of sea-level-pressure patterns. Storm Babet occurred in October 2023, bringing significant rainfall and wind speeds that affected many part of the British Isles. Analogs of storm Babet are calculated from October to November and from 1950 to 2023. It is found that the pattern of such analogs has a time evolution consistent with the influence of the Atlantic Multidecadal Variability (AMV). A warm Atlantic corresponds to a more frequent occurrence of daily sea-level-pressure patterns similar to storm Babet.
The draft is concise but remains somewhat superficial regarding methods. It lacks discussion and analysis on the role of AMV. Additionally, a discussion about the existing literature and the limitations of the analyses conducted is missing. I recommend some major revisions.
Major comments
- The definition of the AMV (region used, smoothing applied or not) is lacking. There is a legend in Fig. 4 specifying that what is called the AMV is, in fact, the SST averaged in 25°N-60°N in September-October-November. However, what is usually called the AMV is a yearly SST average from 0° to 60°N, including the tropics from 0°N to 20°N. The dataset used to calculate the AMV is not clear (HadISST at L68, but HadSST 4.0.1.0 mentioned at L168). The smoothing applied and the dataset used for the GMST is lacking. Lastly, I suggest that the authors remove the external forcing using other methods, as the linear regression used might induce spurious connections with the Indo-Pacific (Deser and Philips, 2023). This is relevant as ENSO can be a driver of fall European climate, which is excluded in the present draft (King et al., 2018).
- The links between the analogs and the AMV is not clear from the results shown. The links are presently deduced from correlations discussed at lines 155-163. However, the tests applied to deduce the p_value are not presented. The correlation of 0.53 given at line 60 is associated with a p-value of almost zero. However, given the few degrees of freedom (roughly 5, considering that there are 7 independent data points from the 7 decades), one can expect such a correlation is not significant at the 5% level. The physical process is also missing in the discussion. What is the process explaining that a warm Atlantic leads to more significant storms? This should be discussed more carefully, as such a link can be a statistical artifact.
- L162-163 : the other potential driver of the change in analogs are not investigated. Perhaps a regression of the sea surface temperature on the Similarity time series would help show which region is the most important: the Pacific Ocean, the subtropical Atlantic or the subpolar Atlantic. The choice of looking at the AMV time series only seems otherwise arbitrary.
- The method of adjusting a GEV with the AMV is not explained at all. Please expand the paragraph L107-112. Did the authors try to include the GMST into the estimations of the return period as well?
Minor comments:
L23-25: Can the authors describe more the previous methods and results that led to the conclusion that the AMV or PDO may have an influence on events like storm Babet?
L30-32: The other potential drivers of the AMV need to be discussed as it is a controversial topic. The atmospheric forcing plays a role. The external forcing also explains a large part of the AMV (Klavans et al., 2022).
L34-36: Provide more details on how the AMV affects the NAO. The references given all presents different mechanism that can be further presented to the reader.
L36-37: Provide more details on the influence of AMV on the stormtrack, and its link with the NAO. See also Varino et al. (2019)
L42-43: ‘’it is often inferred that the change in analogues is the effect of climate change’’ Please provide references.
L60: Can the authors describe Figure A1 and explain which aspects of the SLP-analogs are better than the 500 hPa-analogs.
L75: Is the Euclidian distance the root mean squared error of the fields using area weighting?
Fig 1 : the rainfall is strongly different in Fig. 1b and 1e. Was it expected? Can an analog built using only sea-level-pressure expected to capture high-precipitation events? Maybe part of the key dynamics of the event is missing in the analog, such an atmospheric rivers. Maybe the precipitation is not well captured in ERA5. Did the authors try other precipitation observations?
Fig. 1: please explain in the methods the test implemented to show statistical significance. Only comparing the composite to the standard deviation of the field is too arbitrary.
Fig. 2: the test used for the statistical significance is not presented or explained.
L154: I am not convinced that the Sx time series based on 500-hPa would have similar results as it is correlated with the Sx time series based on sea-level-pressure. The authors should apply their analyses to the two time series (i.e. Sx based on sea-level-pressure and Sx based on 500-hPa geopotential).
L163: ‘’other drivers likely play a role’’ I do not understand why the authors did not investigate maps of SST anomalies associated with their time series. The choice of only investigating the AMV looks like cherry-picking.
L175-180 : the analysis shown in Fig. 5 is not really about the impacts of different AMV phases (see name of section 3.3); it is about the impacts of the similarity variations.
L177: ‘’is 3x greater’’
L178 : ‘’1.2x windier’’
L188: ‘’climameter’’ Can the authors explain this word?
L189: ‘’Thompson et al., in review’’ Can the authors make this paper available?
L201-202 : ‘’Multiple modes of variability may be considered at once, using multiple regression, but this would increase uncertainty in the results’’ I do not understand why investigating the role of other modes of variability through the use of multiple regression would increase the uncertainty.
Fig A2 (a). Can the author explain in the legend what the green and orange lines represent?
References:
Deser, C., & Phillips, A. S. (2023). Spurious Indo-Pacific connections to internal Atlantic Multidecadal Variability introduced by the global temperature residual method. Geophysical Research Letters, 50, e2022GL100574. https://doi.org/10.1029/2022GL100574
King, M. P., Herceg-Bulić, I., Bladé, I., García-Serrano, J., Keenlyside, N., Kucharski, F., ... & Sobolowski, S. (2018). Importance of late fall ENSO teleconnection in the Euro-Atlantic sector. Bulletin of the American Meteorological Society, 99(7), 1337-1343.
Klavans, J. M., A. C. Clement, M. A. Cane, and L. N. Murphy, 2022: The Evolving Role of External Forcing in North Atlantic SST Variability over the Last Millennium. J. Climate, 35, 2741–2754, https://doi.org/10.1175/JCLI-D-21-0338.1.
Varino, F., Arbogast, P., Joly, B., Riviere, G., Fandeur, M. L., Bovy, H., & Granier, J. B. (2019). Northern Hemisphere extratropical winter cyclones variability over the 20th century derived from ERA-20C reanalysis. Climate dynamics, 52, 1027-1048.
Citation: https://doi.org/10.5194/egusphere-2024-1136-RC2 -
AC1: 'Reply on RC2', Vikki Thompson, 10 Sep 2024
We thank the reviewer for the detailed comments. We have revised the manuscript based on the comments, and those of another reviewer, and hope the reviewer finds it improved.
Please find attached a point-by-point response highlighting relevant changes to the manuscript. We have made many changes, including further statistical testing to show the robustness of the results, and a more detailed comparison with other scientific literature on the AMV and European storms.
On behalf of all authors,
Vikki Thompson
Status: closed
-
RC1: 'Comment on egusphere-2024-1136', Anonymous Referee #1, 06 Jun 2024
-
AC2: 'Reply on RC1', Vikki Thompson, 10 Sep 2024
We thank the reviewer for the detailed comments, and are pleased we have conveyed the importance of the research. We have revised the manuscript based on the comments, and those of another reviewer, and hope the reviewer finds it improved.
Please find attached point-by-point responses, with the changes to the manuscript highlighted. We have made many changes, including further statistical testing to show the robustness of the results, and a more detailed comparison with other scientific literature on the AMV and European storms to help support the case for causality.
On behalf of all authors,
Vikki Thompson
-
AC2: 'Reply on RC1', Vikki Thompson, 10 Sep 2024
-
RC2: 'Comment on egusphere-2024-1136', Anonymous Referee #2, 10 Jun 2024
The draft focuses on storm Babet and its analogs in terms of sea-level-pressure patterns. Storm Babet occurred in October 2023, bringing significant rainfall and wind speeds that affected many part of the British Isles. Analogs of storm Babet are calculated from October to November and from 1950 to 2023. It is found that the pattern of such analogs has a time evolution consistent with the influence of the Atlantic Multidecadal Variability (AMV). A warm Atlantic corresponds to a more frequent occurrence of daily sea-level-pressure patterns similar to storm Babet.
The draft is concise but remains somewhat superficial regarding methods. It lacks discussion and analysis on the role of AMV. Additionally, a discussion about the existing literature and the limitations of the analyses conducted is missing. I recommend some major revisions.
Major comments
- The definition of the AMV (region used, smoothing applied or not) is lacking. There is a legend in Fig. 4 specifying that what is called the AMV is, in fact, the SST averaged in 25°N-60°N in September-October-November. However, what is usually called the AMV is a yearly SST average from 0° to 60°N, including the tropics from 0°N to 20°N. The dataset used to calculate the AMV is not clear (HadISST at L68, but HadSST 4.0.1.0 mentioned at L168). The smoothing applied and the dataset used for the GMST is lacking. Lastly, I suggest that the authors remove the external forcing using other methods, as the linear regression used might induce spurious connections with the Indo-Pacific (Deser and Philips, 2023). This is relevant as ENSO can be a driver of fall European climate, which is excluded in the present draft (King et al., 2018).
- The links between the analogs and the AMV is not clear from the results shown. The links are presently deduced from correlations discussed at lines 155-163. However, the tests applied to deduce the p_value are not presented. The correlation of 0.53 given at line 60 is associated with a p-value of almost zero. However, given the few degrees of freedom (roughly 5, considering that there are 7 independent data points from the 7 decades), one can expect such a correlation is not significant at the 5% level. The physical process is also missing in the discussion. What is the process explaining that a warm Atlantic leads to more significant storms? This should be discussed more carefully, as such a link can be a statistical artifact.
- L162-163 : the other potential driver of the change in analogs are not investigated. Perhaps a regression of the sea surface temperature on the Similarity time series would help show which region is the most important: the Pacific Ocean, the subtropical Atlantic or the subpolar Atlantic. The choice of looking at the AMV time series only seems otherwise arbitrary.
- The method of adjusting a GEV with the AMV is not explained at all. Please expand the paragraph L107-112. Did the authors try to include the GMST into the estimations of the return period as well?
Minor comments:
L23-25: Can the authors describe more the previous methods and results that led to the conclusion that the AMV or PDO may have an influence on events like storm Babet?
L30-32: The other potential drivers of the AMV need to be discussed as it is a controversial topic. The atmospheric forcing plays a role. The external forcing also explains a large part of the AMV (Klavans et al., 2022).
L34-36: Provide more details on how the AMV affects the NAO. The references given all presents different mechanism that can be further presented to the reader.
L36-37: Provide more details on the influence of AMV on the stormtrack, and its link with the NAO. See also Varino et al. (2019)
L42-43: ‘’it is often inferred that the change in analogues is the effect of climate change’’ Please provide references.
L60: Can the authors describe Figure A1 and explain which aspects of the SLP-analogs are better than the 500 hPa-analogs.
L75: Is the Euclidian distance the root mean squared error of the fields using area weighting?
Fig 1 : the rainfall is strongly different in Fig. 1b and 1e. Was it expected? Can an analog built using only sea-level-pressure expected to capture high-precipitation events? Maybe part of the key dynamics of the event is missing in the analog, such an atmospheric rivers. Maybe the precipitation is not well captured in ERA5. Did the authors try other precipitation observations?
Fig. 1: please explain in the methods the test implemented to show statistical significance. Only comparing the composite to the standard deviation of the field is too arbitrary.
Fig. 2: the test used for the statistical significance is not presented or explained.
L154: I am not convinced that the Sx time series based on 500-hPa would have similar results as it is correlated with the Sx time series based on sea-level-pressure. The authors should apply their analyses to the two time series (i.e. Sx based on sea-level-pressure and Sx based on 500-hPa geopotential).
L163: ‘’other drivers likely play a role’’ I do not understand why the authors did not investigate maps of SST anomalies associated with their time series. The choice of only investigating the AMV looks like cherry-picking.
L175-180 : the analysis shown in Fig. 5 is not really about the impacts of different AMV phases (see name of section 3.3); it is about the impacts of the similarity variations.
L177: ‘’is 3x greater’’
L178 : ‘’1.2x windier’’
L188: ‘’climameter’’ Can the authors explain this word?
L189: ‘’Thompson et al., in review’’ Can the authors make this paper available?
L201-202 : ‘’Multiple modes of variability may be considered at once, using multiple regression, but this would increase uncertainty in the results’’ I do not understand why investigating the role of other modes of variability through the use of multiple regression would increase the uncertainty.
Fig A2 (a). Can the author explain in the legend what the green and orange lines represent?
References:
Deser, C., & Phillips, A. S. (2023). Spurious Indo-Pacific connections to internal Atlantic Multidecadal Variability introduced by the global temperature residual method. Geophysical Research Letters, 50, e2022GL100574. https://doi.org/10.1029/2022GL100574
King, M. P., Herceg-Bulić, I., Bladé, I., García-Serrano, J., Keenlyside, N., Kucharski, F., ... & Sobolowski, S. (2018). Importance of late fall ENSO teleconnection in the Euro-Atlantic sector. Bulletin of the American Meteorological Society, 99(7), 1337-1343.
Klavans, J. M., A. C. Clement, M. A. Cane, and L. N. Murphy, 2022: The Evolving Role of External Forcing in North Atlantic SST Variability over the Last Millennium. J. Climate, 35, 2741–2754, https://doi.org/10.1175/JCLI-D-21-0338.1.
Varino, F., Arbogast, P., Joly, B., Riviere, G., Fandeur, M. L., Bovy, H., & Granier, J. B. (2019). Northern Hemisphere extratropical winter cyclones variability over the 20th century derived from ERA-20C reanalysis. Climate dynamics, 52, 1027-1048.
Citation: https://doi.org/10.5194/egusphere-2024-1136-RC2 -
AC1: 'Reply on RC2', Vikki Thompson, 10 Sep 2024
We thank the reviewer for the detailed comments. We have revised the manuscript based on the comments, and those of another reviewer, and hope the reviewer finds it improved.
Please find attached a point-by-point response highlighting relevant changes to the manuscript. We have made many changes, including further statistical testing to show the robustness of the results, and a more detailed comparison with other scientific literature on the AMV and European storms.
On behalf of all authors,
Vikki Thompson
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 594 | 154 | 63 | 811 | 29 | 50 |
- HTML: 594
- PDF: 154
- XML: 63
- Total: 811
- BibTeX: 29
- EndNote: 50
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
