Extreme Mediterranean cyclones and associated variables in an atmosphere-only vs an ocean-coupled regional model

Chericoni, Marco; Fosser, Giorgia; Flaounas, Emmanouil; Sannino, Gianmaria; Anav, Alessandro

doi:https://doi.org/10.5194/egusphere-2024-2829

Preprints

https://doi.org/10.5194/egusphere-2024-2829

Preprints

27 Sep 2024

| 27 Sep 2024

Extreme Mediterranean cyclones and associated variables in an atmosphere-only vs an ocean-coupled regional model

Marco Chericoni, Giorgia Fosser, Emmanouil Flaounas, Gianmaria Sannino, and Alessandro Anav

Abstract. Complex air-sea interactions play a major role in both the variability and the extremes of the Mediterranean climate. This study investigates the differences between an atmosphere-only and an ocean-coupled model in reproducing Mediterranean cyclones and their associated atmospheric fields. To this end, two simulations are performed using the ENEA-REG regional Earth system model at 12 km atmospheric horizontal resolution over the Med-CORDEX domain, both driven by ERA5 reanalysis, for a common 33-year period (1982–2014). The atmosphere stand-alone simulation uses the WRF model with prescribed ERA5 Sea Surface Temperature (SST), while in the second WRF is coupled to the MITgcm ocean model at horizontal resolution of 1/12°. A cyclone track method, based on sea level pressure, is applied to both simulations and to the ERA5 reanalysis to assess the model capability to reproduce the climatology of intense, potentially most impactful, cyclones. Results show that the seasonal and spatial distribution of the 500 most intense cyclones is similarly reproduced between WRF and ERA5, regardless the use of the coupling. The two simulations are then compared in terms of sub-daily fields at the cyclones' maximum intensity. Differences in SST distribution between the models primarily control variations in atmospheric variables, not only at the surface, but throughout the planet boundary layer, due to the mixing of the turbulent processes, enhanced during intense cyclones. Additionally, the research investigates the cyclone effects on ocean properties in the coupled simulation, revealing that strong winds enhance surface heat fluxes and upper ocean mixing, while lowering SST. The analysis shows the effectiveness of the coupled model in representing dynamic and thermodynamic processes associated with extreme cyclones across both the atmosphere and the ocean.

Received: 10 Sep 2024 – Discussion started: 27 Sep 2024

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 preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Preprint (PDF, 4155 KB)

Supplement (4302 KB)

Download & links

Marco Chericoni, Giorgia Fosser, Emmanouil Flaounas, Gianmaria Sannino, and Alessandro Anav

Status: final response (author comments only)

RC1:
'Comment on egusphere-2024-2829', Anonymous Referee #1, 14 Oct 2024

This manuscript uses a cyclone tracking algorithm to track the most intense Mediterranean cyclones in atmosphere-only and coupled atmosphere-ocean simulations. It is shown that both STD and CPL simulations represent the climatology of storms in the Mediterranean with no notable advantage to CPL. Also, it is shown that CPL has an SST bias relative to STD, which affects various fields in the PBL. CPL can be used to understand dynamical mechanisms in the ocean mixed layer in the presence of atmospheric cyclones.
Overall, the work performed for this study is impressive – a combination of coupled model simulations with cyclone tracking algorithms and the subsequent analysis. The presentation of the results and the discussions are interesting and well-structured. Yet, I struggle to see the innovative part of this paper. Instead, I see a nice comparison between two simulations – coupled and uncoupled. Ultimately, the primary distinction between simulations is the SST bias in CPL, which imprints on various fields in the PBL. The relevancy of the coupling vs. non-coupling is only demonstrated in Figure 10, Panels a and b. There, the advantage of coupling is clear.
I think it would be helpful to use more careful language that does not attribute the difference of various fields to the coupling (or the explicitly resolved SST). These suggested changes perhaps mean that the conclusions are more relaxed. Still, considering the large effort made by the authors, after addressing the above critics and the more specific comments below, I would recommend the paper for publication.
Specific comments:
Lines 103-104: The authors try to answer the question, “To which extent in the vertical column, and through which physical mechanisms, the explicitly resolved SST distribution and sea surface fluxes impact the precipitation, and the wind speed during extreme cyclones?”, but the designed simulations can’t really separate the effect of explicitly resolved SST when the SST in the Western Mediterranean is about 1.5 degrees warmer. In that case, I think the only thing that can be done is to downgrade the question to something that fits the analysis in the paper.
Line 167-168: “Two cyclones are considered the same event if their minimum of SLP is within a 500 km distance and within a time range of 12 hours.” – this stage removes from the analysis all cases in which CPL is different from STD (~30%). This difference by itself sounds very large, suggesting that many cyclones are represented very differently in the CPL simulation. The authors do not compare their tracks with ERA5-based tracks (e.g., by calculating the RMSE of the distance between observed and simulated cyclone location at maximum intensity), and it is hard to tell which simulation is better. Therefore, it may lead to an unverified conclusion that CPL represents the cyclones well (although it has a large SST bias) and that CPL does not have an added value.
Line 172-173: does this mean that out of 341, 199 cyclones occur in DJF and SON? Please clarify this.
Line 176: can you explain or provide a reference for why this field at the specific level was chosen?
Figure 3: It is not clear how this figure was made. Is the percentage calculated from all days in the specific grid cell or is it a percentage from all cyclones in the region? Could you please clarify this? Also, while the spatial variability is well represented in the models, there is still a very large difference in the percentage relative to observations. The authors should discuss this, at least by providing some information about the source of this large difference.
Line 307: “This is explained by the higher Θ gradient of the CPL (Fig. 7d), that makes the PBL less stratified and higher” – what is exactly explained by the higher Θ gradient, and does this gradient is the reason why PBL is less stratified and higher? I would say that this is because of the higher SST, as mentioned in the previous sentence. This reasoning is not clear to me. I would say that, in general, the PBL should be well mixed, and differences between STD and CPL should be pretty small in terms of the temperature gradients inside the PBL. I would attribute the difference only to the SST difference.
Figure 10: It is unclear which region is considered when calculating the SST difference. Is it one grid cell where maximum cyclone intensity occurred, or is it a regional average?
Minor:
Line 21: “plant” -> “planetary”
Line 27: “because is” -> “because it is”?
Line 240: “(Fig. 3a and b)” -> “(Fig. 3b and c)”?
Line 286: “CLP” -> “CPL”
Line 288: “Fig. S6” - > “Fig. 6”
Line 291: “CLP” -> “CPL”
Figure S2: need to correct the reference to the panels in the caption
Line 351: “CLP” -> “CPL”
Figures: can you explain what the deltas at the top of the panels mean? is it a simple domain average?

Citation: https://doi.org/10.5194/egusphere-2024-2829-RC1
- AC1: 'Reply on RC1', Marco Chericoni, 06 Dec 2024
  
  We thank the Reviewer for dedicating time to review our manuscript and for providing detailed observations that have improved the quality of the paper. To address all concerns, we have prepared a PDF document (attached as a supplement) where we answer to each point individually.
  
  Citation: https://doi.org/10.5194/egusphere-2024-2829-AC1
RC2:
'Comment on egusphere-2024-2829', Anonymous Referee #2, 29 Oct 2024
General comments
The present study investigates the Mediterranean cyclones using regional climate models (RCMs) as well as the ERA5 reanalysis. They assess the reproducibility of intense Mediterranean cyclones in their atmosphere-ocean coupled RCM simulation and compare the simulation with an atmosphere stand-alone simulation to examine the effect of air-sea coupling in RCMs. They also investigate the impacts of intense Mediterranean cyclones on the ocean using the coupled simulation in comparison to the ocean observations.

Overall, this study potentially provides materials with implications to improve our understanding of the importance of air-sea interactions for Mediterranean cyclones, which hold significant socio-economic relevance. The influence of air-sea coupling assessed through their RCM simulations is clear and interesting

I am afraid, however, that a major revision is needed before this study can be published, for the specific reasons shown below. I am particularly concerned about the interpretation of the results.

Specific comments
Turbulent heat fluxes, precipitation (and associated diabatic heating), and low-level temperature distribution are important factors for extratropical cyclone development. This study argues that the different SST distribution between the models, which is due to the air-sea coupling, is the dominant factor in shaping anomalies of those variables. Yet, this study also concludes that the coupling between the atmosphere and ocean exerts a limited influence on their statistics such as frequency, lifetime, speed, and intensity. To me, the two conclusions appear inconsistent. The authors should provide a detailed discussion on why these results emerged, addressing both the model's role and potential underlying mechanisms.

The spatial composite maps (Figs. 4-8) seem to be composites of time steps when one or more “common” selected intense cyclones are located within the domain of interest (Fig. 1). If this is the case, signals in these composite maps do not necessarily occur in the vicinity of intense cyclones of interest. This discrepancy could partly explain the study’s apparently inconsistent conclusions (as noted above). While signals in composite maps during intense cyclone time-steps are associated with cyclone dynamics, some may not be directly relevant to cyclones.

The reasoning behind the differences in cyclone frequencies between RCMs and ERA5 (Figs. 2 and 3) is not quite convincing. The authors argue that the differences are due to the different native resolutions of the RCMs (higher) and the ERA5 model (lower). I am afraid, however, that ERA5 is a reanalysis assimilated with a lot of observations. Those observations are particularly abundant in the land and expected to substantially improve the reproducibility of synoptic and larger-scale features, including intense cyclones near the coast. It would make more sense to attribute differences in the open ocean mainly to models’ native resolutions, but this is not the case for cyclones, for example, in the Adriatic Sea, the Ligurian Sea, and the Aegean Sea, and off the Gulf of Antalya. For synoptic-scale features, I think that reanalyses generally to a considerable extent reflect the reality, whose “native resolution” is much higher than any simulations.

The interpretation and discussion of SST differences (or biases) between STD and CPL are insufficient, and the authors should discuss the SST bias in relation to cyclone frequency and other properties. Is it related to the difference in the frequency and other properties of intense cyclones? Because there is no corresponding map of cyclone frequency difference to Figs. 5 and 6, it is hard to consider the relationship between the cyclone frequency and SST. To me, the SST difference (or bias) of nearly 2K is large enough to be expected to have impacts on the overlying atmosphere and cyclones (as in the following subsection). If SST differences indeed influence the atmosphere, it would be beneficial to clarify (or at least suggest) the mechanisms within the model that generate these differences.

Related comment: Evaporation (or latent + sensible heat fluxes) are influenced by wind speed, surface temperature and specific humidity, and SST, following the bulk formula. Thus the evaporation difference (Fig. 7a) can result from the other atmospheric differences (e.g., Figs. 7b-c), rather than a one-way causation as in this study. While there is spatial alignment between SST bias and evaporation differences, this is not adequately addressed in the manuscript.

Section 2.1 mentions that “STD” is an atmosphere-only WRF simulation with prescribed SST, while “CPL” is the ENEA-REG regional Earth system model including ocean, land, and freshwater fluxes and river discharge model. However, the authors argue that the only difference between STD and CPL resides in the SST over the Mediterranean Sea. Is this accurate? I am wondering if there is any significant difference on the land between the two simulations.

L172: Given that cyclones are more intense in winter and the role of the SST and the air-sea fluxes on extreme events is expected to be stronger in autumn (LL106-107), it is counterintuitive to see the number of the selected most intense cyclones in winter and autumn is 199, while 142 (341-199) in summer and spring, corresponding to 71 cyclones in three months (JJA or MAM) on average. Does this mean there are fewer intense cyclones around the Mediterranean in autumn than in spring or summer? If so, I wonder why the focus of this study is only on autumn and winter Mediterranean cyclones.

Technical corrections and minor comments
L15: “in the second WRF” -> “in the second simulation the WRF” ?

L18: similarly reproduced -> “similar” ?

L19: regardless -> regardless of

L21: planet -> planetary

L32: “Sea Surface Temperature” -> “sea surface temperature”

L35: “midlatitude cyclones, entering the Mediterranean basin” -> “midlatitude cyclones entering the Mediterranean basin” ?

L64: long ago -> long

L67 “boundary conditions, that becomes” -> “boundary conditions, which becomes”

L69: challenged -> attempted?

L87: “improves the track length” -> “improves the reproducibility of the track length” ?

L95: investigating -> by investigating

L95: affects -> affect

L103: “the explicitly resolved SST” -> “do the explicitly resolved SST”?

L107: I am not sure why the role of the SST and the air-sea fluxes on extreme events is expected to be stronger in autumn than in winter.

L108: “next” -> “the next”

L118: “extensively used” -> “which is extensively used”

L123: I think that 1/12deg is substantially less than approximately 10km, especially in the zonal direction.

L124 (the Med-CORDEX region): It would be helpful to refer to Fig. 1 here.

L125: The resolution of the SST prescribed to ERA5 depends on the period. In particular, the resolution is substantially different between HadISST2 (~2007) and OSTIA (2007~). The description of ERA5 SST as Δx~0.25deg might be misleading.

L146: "cyclone tracking algorithm" might be preferable to "storm track method."

L147: “CYCLOYTRACK” -> “CycloTRACK” (see Flaounas et al. 2014)

L147: “Mean Sea Level Pressure” -> “mean sea level pressure”

L157: Referring to terrain > 800m as “high mountain environment” sounds a bit weird.

L163: The minimum SLP during a cyclone’s lifetime depends on the background, larger-scale SLP distribution than the synoptic scale. Is the climatological-mean SLP nearly uniform in the Mediterranean Sea?

L172: Please specify what “DJF” and “SON” stand for here.

L179: STD e CLP -> STD and CPL

L182-183: The authors argue that “the influence of cyclones on the atmospheric state is independent of the location of the cyclones in the Mediterranean Sea”. Why?

L229: The present study utilizes the same cyclone tracking algorithm as Flaounas et al. (2023). Thus the authors should specify references for “different cyclone tracking methods” here.

3: Consider using a more realistic mask shape (e.g., circles) over 3degx3deg square domains. Additionally, which season(s) is this figure showing? DJF+SON?

L256-259: I consider that the explanation of the structure of section 3.2 (subsection) here is not needed.

L264: It does not make sense to me that cyclones do not turn into precipitation. (Additionally, the sentence here seems grammatically incorrect.)

“PBL is higher” -> “PBL height is larger”?

4: The panel labels are hard to find. Please improve their visibility.

If possible, the results in Fig. 4 should be compared with observations (e.g., ERA5).

L277: The title “SST analysis” is obscured and needs to be clarified.

The figures' sequence (including supplementary ones) could be improved to follow numerical order.

L288 “All the outcomes on DJF are also valid for the analysis of the SST bias in SON (Fig. S6)”: It seems contradictory with the following statement. Please clarify which outcomes in DJF are valid for SST bias analysis in SON.

L288 Fig. S6 -> Fig. 6

5-8: White contours for significance are difficult to identify. Consider improving visibility. (Perhaps There are a lot of white contours to white out insignificant differences with colors?)

L311: What correlation is “the high correlation” (and Fig. 9)? Spatial correlation over the sea?

L315 Clarify if the observed link between warmer SST and higher 10 m wind speed is related to the "vertical mixing effect" from Wallace et al. (1989) and Hayes et al. (1989).

L319: I could not follow the argument that “the stronger horizontal winds in CPL lead to a mismatch between areas of high vertical moisture flux and total precipitation”. What is “high vertical moisture flux” and how is the mismatch related to the stronger horizontal winds? This paragraph (L319-326) should be improved so as to be understood more easily.

L 389: Please clarify the term "mean cooling."

Related comment to major comment (B): I think that showing the composited wind (Fig. 4d) is misleading, because the wind is the superposition of all the events considered, which come mainly from those around Italy but include those occurring in distant regions (Fig. 3). In other words, the precipitation distribution (Figs. 4a-b) does not necessarily occur associated with the wind pattern in Fig. 4d.
Citation: https://doi.org/10.5194/egusphere-2024-2829-RC2
- AC2: 'Reply on RC2', Marco Chericoni, 06 Dec 2024
  
  We thank the Reviewer for dedicating time to review our manuscript and for providing detailed observations that have improved the quality of the paper. To address all concerns, we have prepared a PDF document (attached as a supplement) where we answer to each point individually.
  
  Citation: https://doi.org/10.5194/egusphere-2024-2829-AC2
RC3:
'Comment on egusphere-2024-2829', Anonymous Referee #3, 01 Nov 2024

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-2829/egusphere-2024-2829-RC3-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2024-2829-RC3
- AC3: 'Reply on RC3', Marco Chericoni, 06 Dec 2024
  
  We thank the Reviewer for dedicating time to review our manuscript and for providing detailed observations that have improved the quality of the paper. To address all concerns, we have prepared a PDF document (attached as a supplement) where we answer to each point individually.
  
  Citation: https://doi.org/10.5194/egusphere-2024-2829-AC3

Marco Chericoni, Giorgia Fosser, Emmanouil Flaounas, Gianmaria Sannino, and Alessandro Anav

Supplement

https://doi.org/10.5194/egusphere-2024-2829-supplement

Marco Chericoni, Giorgia Fosser, Emmanouil Flaounas, Gianmaria Sannino, and Alessandro Anav

Viewed

Total article views: 337 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
188	54	95	337	17	3	4

HTML: 188
PDF: 54
XML: 95
Total: 337
Supplement: 17
BibTeX: 3
EndNote: 4

Views and downloads (calculated since 27 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	55	8	3	66
Oct 2024	68	22	35	125
Nov 2024	45	10	49	104
Dec 2024	20	14	8	42

Cumulative views and downloads (calculated since 27 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	55	8	3	66
Oct 2024	68	22	35	125
Nov 2024	45	10	49	104
Dec 2024	20	14	8	42

Viewed (geographical distribution)

Total article views: 329 (including HTML, PDF, and XML) Thereof 329 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Dec 2024

Short summary

This study explores how sea surface energy influences both the atmosphere and ocean at various vertical levels during extreme Mediterranean cyclones. It focuses on cyclones' precipitation and wind speed response, as well as on ocean temperature variation. The analysis shows the effectiveness of the Regional Coupled Model in coherently representing the thermodynamic processes associated with extreme cyclones across both the atmosphere and the ocean.


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