the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Process-based classification of Mediterranean cyclones using potential vorticity
Yonatan Givon
Or Hess
Emmanouil Flaounas
Jennifer L. Catto
Michael Sprenger
Shira Raveh-Rubin
Abstract. Mediterranean cyclones govern extreme weather events across the Euro-African basin, affecting the lives of hundreds of millions. Despite many studies addressing Mediterranean cyclones (MCs) in the last decades, their correct simulation and prediction remain a significant challenge to the present day, which may be attributed to the large variability among MCs. Past classifications of MCs are primarily based on geographical and/or seasonal separations, however, here we focus on cyclone genesis and deepening mechanisms. A variety of processes combine to govern MC genesis and evolution, including adiabatic and diabatic processes, topographic influences, land-sea contrasts, and local temperature anomalies. As each process bears a distinct signature on the potential vorticity (PV) field, a PV approach is used to distinguish among different “types” of MCs. Here, a combined cyclone tracking algorithm is used to detect 3190 Mediterranean cyclone tracks in ECMWF ERA5 from 1979–2020. Cyclone-centered, upper-level isentropic PV structures in the peak time of each cyclone track are classified using the Self Organizing Map (SOM). The SOM analysis reveals 9 classes of Mediterranean cyclones, with distinct Rossby wave-breaking patterns as discernible in corresponding PV structures. Though classified by upper-level PV structures, each class shows different contributions of lower-tropospheric PV and flow structures down to the surface. Unique cyclone life cycle characteristics, associated hazards (precipitation, winds, and temperature anomalies), and long-term trends, as well as synoptic, thermal, dynamical, seasonal, and geographical features of each cyclone class, indicate dominant processes in their evolution. Among others, the classification reveals the importance of topographically-induced Rossby wave breaking to the generation of the most extreme Mediterranean cyclones. These results enhance our understanding of MC predictability, by linking the large-scale Rossby wave formations and life cycles to coherent classes of under-predicted cyclone aspects.
- Preprint
(3402 KB) - Metadata XML
- BibTeX
- EndNote
Yonatan Givon et al.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-1247', Anonymous Referee #1, 28 Jul 2023
This study deals with the clustering of Mediterranean cyclones based on dynamic classification (upper-level PV). I think the results are relevant for the entire scientific community working on cyclones in the Mediterranean, but also outside the area, considering that the methodology can be applied anywhere. In my opinion, two results deserve particular attention: the observation of different trends for different clusters, which were only marginally visible in historical data until now; the fact that the cluster category remains the same since cyclogenesis, thus providing additional predictability. Therefore, my suggestion is for minor revision, mainly focused on improving the readability of the article and its structure.
MINOR POINTS:
L93: I think it is not appropriate to consider tropical-like cyclones as thermal lows. They are also not scattered in the Mediterranean (they are mainly concentrated near the Balearic Islands and the Ionian Sea).
L98: explosive development can also occur for extratropical cyclones, therefore the mixing of information with medicanes is not appropriate in this case; therefore, I would keep the two sentences separate.
L110: Portmann et al. (2020)
L113: I cannot find Buzzy et al. (2022) among the references.
L177: The reason for choosing 9 clusters is not clearly addressed even in the Appendix.
L200: I suggest moving these details about statistical methods to an Appendix.
L219: “contour points with small great circle distance (anchor points)”: please clarify what you mean, adding some additional information to better explain the methodology.
L234: therefore, the fronts are identified along the inflections of the gradient ( )? Should not it be along the maxima of the gradient ( )? Please explain.
L260: How do you explain the difference between large-scale and convective precipitation distribution? Should not it be the convective precipitation more localized along the cold front?
L280: I suggest naming the different clusters only in Section 4, after analyzing the different properties.
L280: Are cyclonically (8) and anticyclonically (2, 5) breaking PV streamers named from the analysis of the similarities to the theoretical configuration given in Thorncroft's study?
L281: I suggest turning to "The frequency of objectively identified PV streamers".
L290 (L430): what does S1 (S3) mean?
L368: To overcome these interpretation issues, could you also look at the 12-hour interval intensification rates?
L424: "weakest for clusters 6 and 9": it does not seem to me that they are the weakest in terms of wind anomaly, they are the weakest in terms of PV anomaly.
L427: “anticyclonic breaking PV streamers”: this creates some confusion since you classify Cluster 8 as “CWB low” in Table 1.
L535: most instead of more.
L538: windward side instead of wind side?
L455: “downward movement into the lower levels” is more appropriate here than downdrafts.
L451-L471: the narration here does not work smoothly, since some statements (L458-459; L463: “Anti-cyclonic followed by cyclonic wave breaking”) refer to points that will be better discussed in the following Section 4 and in Appendix B (Fig. B1), but they are not adequately supported here.
L463: a new paragraph should start from “The presence of…”.
L535: most instead of more.
L538: windward side instead of wind side?
L607: Appendix B: Considering this part is central to the discussion, I ask the authors to consider moving it into the main text.
L635: Figure B2: No PV towers can be detected here: is this due to the coarse resolution of the reanalyses?
Citation: https://doi.org/10.5194/egusphere-2023-1247-RC1 -
RC2: 'Comment on egusphere-2023-1247', Anonymous Referee #2, 31 Aug 2023
This study classifies nine types of cyclones in the Mediterranean by applying the Self Organizing Map analysis to the upper-level potential vorticity fields around cyclones. In spite of a simple classification based on a single variable, it effectively identifies different types of cyclones including lee-cyclones, Rossby wave breakings, and heat lows with different characteristics such as three-dimensional structures and seasonality. These results are nicely summarized in Table 1. I agree that the process-based classification transcends the geographical-seasonal classification because a similar process can occur in different regions and different processes can occur in the same region. This study further examines the predictability and long-term trends. Overall, this study clearly demonstrates the variability of Mediterranean cyclones with appropriate, sophisticated, and straightforward analyses. While the manuscript is almost acceptable for the publication in Weather and Climate Dynamics in the current form, I would like to make three suggestions (major comments 1-3) which may improve the manuscript. Therefore, I recommend minor revisions of this manuscript at this stage.
Major comments
1. In section 4, is it possible to discuss whether individual classes are similar to or different from cyclone types in other regions? Although the authors discuss it for some classes (e.g., a description of clusters 2 and 8 in Lines 557-558), it would be helpful to summarize such discussion after Table 1. For example, are some clusters in the present study similar to clusters for the cyclones in the Southwest Pacific (Catto 2018, https://doi.org/10.1175/JCLI-D-17-0746.1)? Readers may also want to know whether some classes are linked to medicanes because the authors explain it in detail in section 1.
2. The authors should provide more information on precipitation analysis. Regarding the large-scale and convective precipitation, are they outputs from a global model used for the ERA-5 reanalysis? Regarding the warm conveyor belt analysis, did the authors apply the trajectory analysis introduced by Madonna et al. (2014) to the hourly wind fields of the ERA-5 reanalysis? In particular, I wonder to what extent the relevant results are changed if the analyses are applied to a higher resolution dataset which can resolve convection more explicitly. While I think these analyses in the present manuscript are still informative (the authors do not need to remove them), the authors should discuss caveats on these analyses. Another minor suggestion is that the authors show the sum of large-scale and convective precipitation by shading or contour in Fig. 2d.
3. Lines 420-434: This paragraph describes 10-m wind speed and upper-level PV. I wonder why the upper-level PV is discussed in section 3.4 which focuses on surface impact. I think that the analysis of the upper-level PV anomaly is more closely linked to Figs. 3 and C1 which are explained in section 3.2. On the other hand, this paragraph focuses less on 10-m wind speed. From the perspective of surface impact, I think that the total wind speed is also important. Fig. 2d and 10 only show the anomaly concerning monthly local climatology. Does it make sense to examine the total wind speed?
Minor comments
4. Lines 422-426: The wind anomaly for cluster 5 looks weaker than that for clusters 6 and 9.
5. Line 445: Give the cluster numbers for the summer clusters.
6. Lines 463- 465: The explanation of warm conveyor belt and Fig. C3 should be moved before the sentence that refer to Fig. C2 first in Lines 456-457.
7. Lines 544-545: Give the section of the Appendix.
8. Fig. B2: is this the y-z plane through the cyclone center?
Citation: https://doi.org/10.5194/egusphere-2023-1247-RC2
Yonatan Givon et al.
Data sets
Cyclone labels, associated cluster, and location at classification time Yonatan Givon https://www.dropbox.com/s/0bqdhv0hus4oekp/Label_Cluster.pdf?dl=0
Yonatan Givon et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 281 | 139 | 11 | 431 | 6 | 4 |
- HTML: 281
- PDF: 139
- XML: 11
- Total: 431
- BibTeX: 6
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1