the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jan 2026
Attribution of the 2025 Mediterranean Marine Heatwave to Climate Change Using Analogues
Abstract. The Mediterranean Sea experienced a record-breaking marine heatwave in 2025, raising questions about the influence of human-induced climate change. Using an analogue-based attribution approach, we compared historical sea surface temperature (SST) patterns from ERA5 reanalysis (1950–2024) to those observed during the event. By identifying the most similar SST anomaly patterns in a past period (1950–1986) and a more recent one (1987–2024), we assessed changes in SST and related atmospheric variables. We find that, under present-day climate conditions, analogous patterns result in significantly higher SST anomalies and stronger atmospheric responses – such as warmer near-surface air temperatures and intensified radiative fluxes – compared to the past. Statistical tests confirm that long-term warming trend has amplified Mediterranean SST extremes by up to 1.5 °C and associated heat exchange processes, though shifts in large-scale natural climate variability may also influence these outcomes, complicating attribution. Nonetheless, the dominant contribution to the 2025 marine heatwave severity is attributable to anthropogenic forcing. This study proves the effectiveness of the analogue method for assessing extreme events, also including marine heatwaves, in a warming Mediterranean context.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(0 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-5055', Anonymous Referee #1, 18 Jan 2026
-
AC1: 'Reply on RC1', Davide Faranda, 27 Mar 2026
Response to Referee #1 (comments in bold, answers in normal text)
We thank the referee for the careful reading of the manuscript and for the constructive and encouraging comments. We appreciate the recognition that the study proposes a perspective that differs from the standard use of analogue based attribution approaches. The suggestions provided by the referee will help us clarify the contribution of the manuscript and improve the presentation of the results.
In the revised manuscript we will better highlight the methodological perspective adopted here. In particular, we will make clearer that the study does not follow the classical analogue attribution framework in which atmospheric circulation analogues are first identified and then associated extremes are analysed. Instead, we start from the observed extreme SST anomaly pattern and search for analogous SST patterns in the historical record, and only afterwards analyse the atmospheric and thermodynamic context associated with those events. We will clarify this conceptual point more explicitly in the title, abstract, and introduction.
Below we provide a point-by-point reply to each comment. Referee’s comments are provided in bold font, while our replies are in normal font.
“Summary: The manuscript presents a novel application of the analogue method to attribute a Mediterranean heat wave to anthropogenic climate change. Although the application of the analog method for attribution is not new, this study adopts a different perspective: instead of searching for analogs of the atmospheric circulation and then comparing the observed extreme with those analog dates, this study searches for analogs dates of the observed extremes and then compares the atmospheric circulation of those dates with that observed at the dates of the extreme. The conclusion reached for this particular Mediterranean heat wave is that the atmospheric dynamics in the past, while similar to that observed in the extreme, were linked to higher radiative and thermodynamic forcings.”
Response:
We thank the referee for recognizing the different methodological perspective adopted in the manuscript. As suggested, we will emphasize this distinction more clearly in the revised manuscript. In particular, we will clarify that the analogue search is performed in the SST anomaly space rather than in circulation space, and that the subsequent analysis compares the atmospheric and thermodynamic context of those SST analogue events across different climate backgrounds. This clarification will be introduced more explicitly in the title, abstract, and early part of the introduction.“Recommendation: I found the study interesting and, as I wrote above, considering this novel perspective that complements the standard application of the analog method for attribution. I have a recommendation that the authors may want to consider in a revised version”
Response:
We thank the referee for the positive assessment of the study and for the constructive recommendations.Main points
“1) The title and the abstract do not suggest that the perspective adopted in this study is different from other applications of the analog method. So the authors could better highlight this novel aspect. When reading the title, I first thought that this would be 'just another attribution study'. However, it is not, but this becomes clear when reading the complete manuscript. So I would recommend amending the title and the abstract.”
Response:
This is a very useful suggestion. In the revised manuscript we will modify both the title and the abstract to make clearer that the study adopts a different analogue based attribution perspective. Specifically, we will emphasize that the analogue search is conducted in SST anomaly space and that the atmospheric circulation and thermodynamic conditions are analysed afterwards for those analogue events. This methodological point will be highlighted earlier in the abstract and in the introduction so that the novelty of the approach is immediately visible.“2) In the conclusions section, the authors fall into their own trap. and write some sentences as if the conclusions came from a standard analog-attribution study. For instance, in line 389 they write 'This approach yields a physically transparent storyline: given the same atmospheric/oceanic setup that lead to the 2025 heatwave, a mid-20th-century climate would have produced a far less severe event'.However, this is not what the study has done. It has not explored the outcome of the same (or similar) atmospheric/oceanic set up - it has explored the differences in the predictors of analogous heat waves”
Response:
We agree with the referee. The original wording was misleading. The analysis does not reconstruct identical atmospheric or oceanic states and therefore does not support a counterfactual statement based on the same dynamical setup. We will revise the text accordingly. The conclusions will be reformulated to state that we compare atmospheric and thermodynamic conditions associated with similar SST anomaly patterns in different climate backgrounds. The attribution will be framed as evidence consistent with thermodynamic amplification, not as a causal statement based on identical dynamics.“3) This different attribution scope, though interesting, makes it more difficult to quantify the impact of climate change. In the traditional approach, the difference between, say, the mean of past analogs and the observed extreme would be the climate change signal. But how can one quantify, with the approach presented in this manuscript, the contribution of climate change? Actually, the authors do not provide a number, perhaps because it is indeed more difficult to obtain. This should be discussed.”
Response:
This is a helpful comment. The approach used here does not provide a single scalar estimate of the climate change contribution, as is often done in traditional analogue-based attribution or model-based frameworks. Instead, the contribution of climate change is assessed through systematic differences in the distributions and composite properties of key variables associated with SST analogue events in different climate backgrounds. In particular, we analyse shifts in thermodynamic indicators and radiative fluxes between past and present analogues, which provide quantitative diagnostics of how the background state has changed. We will clarify this point in the revised manuscript. The method should be understood as providing quantitative, process-based diagnostics rather than a single aggregated attribution metric. This perspective is complementary to more classical approaches and contributes an additional line of evidence within a multi-method attribution framework. We will also make explicit that the method does not fully disentangle all sources of low-frequency variability from externally forced trends, which remains a limitation of the observational analogue framework.Minor points
“4) 'we compared historical sea surface temperature (SST) patterns from ERA5 reanalysis (1950–2024) to those observed during the event.' The SST from the ER5 reanalysis come from two different sources, as the manuscript clearly states, depending on the period. However, in neither of these two periods are the SSTs a 'reanalysis' product in the classical sense. The OSTIA SSTs are indeed the result of a data assimilation methodology, but the 'model' is very simple, just persistence of the previous time step. A reader may think that SSts are also the result of data assimilation of observations and a proper ocean model.”
Response:
We thank the referee for pointing out this important nuance. In the revised manuscript we will clarify the wording to avoid implying that the SST field corresponds to a fully coupled ocean reanalysis. We will explicitly state that the SST data in ERA5 are based on observationally constrained products that differ between periods and that they do not result from assimilation into a full ocean circulation model. The text will be revised to describe this more precisely.“5) Figure 1. 'Time series of daily sea surface temperature (SST) anomalies averaged (standard deviation shaded in grey)' I cannot see the shaded area in the downloaded pdf file. Perhaps the rendering of the figure is not totally correct”
Response:
We thank the referee for noting this issue. We will check the figure rendering and ensure that the shading is clearly visible in the revised version of the figure.“6) Section 5 consists of just one big paragraph (!). I would recommend splitting the paragraph, perhaps by the variables being described, to help the reader.”
Response:
We will revise this section and split the paragraph into shorter subsections organized by the variables being discussed. This will improve readability and make the structure of the analysis clearer.
“7) 'STR (q–t) presents positive Present–Past differences, implying reduced net longwave loss in the present climate, consistent with a warmer, moister atmosphere returning more longwave radiation to the surface (Trenberth et al., 2015; Vautardet al., 2023)' I think this paragraph and the corresponding figure warrant a more detailed discussion or description. In principle, attributing it to climate change would entail stronger long-wave forcing from greenhouse gases. Feedbacks increase specific humidity, thereby amplifying the greenhouse effect. But in theory, another scenario is possible: solar radiation increases, warming the ocean surface and increasing specific humidity. In this scenario, anthropogenic climate change would not be ultimately responsible, or only indirectly through reduced cloudiness. So there are subtle questions to discuss based on this figure. A clean attribution would find that the same MHW is linked to a similar or smaller solar radiation and to increased longwave radiation. I am aware that this reduces the number of possible analogs, but , again, this seems to me to be a complication of this novel attribution approach.”
Response:
In the revised manuscript we will expand the discussion of the radiative components associated with the analogue composites. In particular, we will clarify that the interpretation of the longwave radiation signal should be considered in combination with the shortwave component and with the thermodynamic background conditions. We will also explicitly acknowledge that different mechanisms can potentially produce similar surface responses, including increased solar input or thermodynamic feedbacks associated with a warmer and moister atmosphere. In addition, we will also mention that reduced cloudiness is less relevant in the Mediterranean Sea compared to others basins (e.g. North Atlantic) due to very low cloud cover in summer. The revised discussion will therefore present the radiative signals as evidence consistent with thermodynamic amplification in a warmer climate, while acknowledging that the observational framework used here cannot isolate every contributing process. This clarification will make the interpretation more balanced and transparent.Citation: https://doi.org/10.5194/egusphere-2025-5055-AC1
-
AC1: 'Reply on RC1', Davide Faranda, 27 Mar 2026
-
RC2: 'Review on the Attribution of the 2025 Mediterranean Marine Heatwave to Climate Change Using Analogues', Anonymous Referee #2, 15 Mar 2026
Dear editor,
I apologize for the delay in my response. I have taken the time to carefully review the manuscript titled "Attribution of the 2025 Mediterranean Marine Heatwave to Climate Change Using Analogues." My detailed comments and suggestions are provided in the attached PDF document.-
AC2: 'Reply on RC2', Davide Faranda, 27 Mar 2026
Response to Referee #2
We thank the referee for the careful and constructive review. We agree that the original manuscript did not always distinguish clearly enough between figure-based results, physical interpretation, and the scope of the attribution claims. In the revised version, we will substantially revise the structure and framing of the manuscript. In particular, we will separate Results and Discussion more clearly, restrict the Results section to what is directly supported by the figures, and reformulate statements that were too strong or not sufficiently supported.
We will also clarify the scope of the attribution perspective. The analogue framework used here provides an observational and conditional line of attribution evidence, based on comparing similar SST anomaly patterns across different climate backgrounds. It does not constitute a complete formal attribution assessment on its own, and we will make this limitation explicit. At the same time, analogue-based attribution is an established approach in the recent literature, and our intention is therefore not to remove the attribution framing, but to formulate it more precisely and position it clearly within a broader multi-line evidence framework.
Finally, we will strengthen the discussion of limitations, in particular regarding the role of internal variability and the omission of ocean dynamical processes such as advection, mixing, and stratification.
Below we provide a point-by-point response to the referee’s comments.
The authors present an analysis of the 2025 Mediterranean marine heatwave (MHW). Using an analogue-based approach, they examine SST anomalies and synoptic conditions in a past (1950-2024) and a present (1987-2024) period, in order to assess the atmospheric conditions, conducive to the development of similar events in these two periods, in comparison to the conditions associated with the 2025 event. While the results are broadly consistent with the growing body of literature documenting the intensification of Mediterranean MHWs and their atmospheric drivers, the current manuscript does not yet provide sufficiently robust evidence to support its main attribution-related claims.
Response:
In the revised manuscript we will clarify more carefully what kind of attribution claim is being made. We will state explicitly that our approach provides an observational line of evidence for attribution, based on (conditional) analogue comparison across different climate backgrounds, and not a full formal attribution assessment by itself. We will therefore tone down several formulations in the Abstract, Results, and Discussion, while maintaining the attribution framing of the manuscript. We will also explain more clearly what the analogue approach adds here, namely a process oriented and event specific comparison of similar SST patterns under different climate backgrounds.A key concern is the methodological interpretation. Although the analogue method is a useful tool for identifying historical precursors and exploring variability of atmospheric drivers, it cannot on its own provide formal attribution of events to anthropogenic climate change. At several points in the manuscript, the results derived from the analogue framework appear to be treated as a direct proxy for the influence of anthropogenic climate change on the occurrence of the event. In addition, many of the results rely primarily on qualitative comparisons and visual inspection rather than on a clearly defined quantitative framework, which leaves parts of the analysis more descriptive than substantiated. Also, the study focuses on the atmospheric drivers of the 2025 event, with the role of oceanic processes being largely neglected in the analysis and only briefly acknowledged in the limitations of the study.
Response:
We will revise the manuscript to avoid formulations that could be read as equating analogue results with a complete attribution proof. Instead, we will frame the method as one observational line of evidence within a broader attribution framework. We will also make the quantitative parts of the analysis more explicit in the Methods section and reduce descriptive language where it is not directly supported by the figures. In addition, we will expand the discussion of limitations related to ocean processes. In the summer Mediterranean, strong stratification, low mixed layer variability, and weak memory effects suggest a limited role for ocean preconditioning beyond the long term warming background, while atmospheric forcing is likely to play a major role in the development of these events. We will nevertheless make clear that ocean dynamical contributions such as advection and mixing are not explicitly resolved in our framework and cannot be fully excluded.
To strengthen the manuscript, I suggest reconsidering the framing of the conclusions. In particular, the authors may wish to avoid drawing definitive inferences regarding the causality or likelihood of the event under climate change. Such questions are typically addressed through formal attribution studies using climate model ensembles, and the influence of anthropogenic climate change on the increasing likelihood and intensity of Mediterranean MHWs is already well established in the literature. As a result, attempting to infer this within the current framework, without a formal attribution approach, does not substantially advance existing understanding of the topic. Instead, the study could focus more on the specific insights that the analogue approach may provide in comparison to existing methods. Clarifying whether and in what ways, this approach may offer additional physical understanding or diagnostic value specifically for the 2025 event, beyond established analyses would, perhaps, help better define the contribution of the study. At present, however, this added value (if any) is not yet clear.
Response:
We will revise the conclusions along these lines. In particular, we will avoid overly definitive wording on causality and likelihood, and we will better explain the specific contribution of the analogue approach. Our intention is not to duplicate model based attribution studies, but to provide an event based observational (conditional) attribution line of evidence and a process oriented diagnostic interpretation under comparable large scale conditions. We will make this contribution clearer in the Introduction and Discussion, and we will explicitly position the manuscript relative to existing literature showing that Mediterranean MHWs are intensifying in response to climate change.Overall, a major revision that addresses the detailed comments below and refocuses the manuscript on the methodological and physical insights provided by the analogue analysis (rather than on causal attribution) would substantially improve the clarity and impact of the study. In addition, the manuscript would benefit from improvements in structure and clarity. For instance, the introduction remains quite general and is not sufficiently aligned with the specific objectives of the study, while the Results section at times blends results with discussion. A clearer organization and a more cohesive, logically structured narrative would strengthen the presentation of the work.
Response:
We will substantially revise the manuscript structure. The Introduction will be shortened and made more focused on the scientific objective and on the role of analogue based attribution within the broader attribution framework. The Methods and data descriptions will be streamlined and better organized. The Results section will be restricted to what is directly shown. Interpretive and literature based statements will be moved to the Discussion. We will preserve the attribution perspective of the manuscript, but express it more carefully and more consistently.Abstract
Line 14:
“..raising questions about the influence of human-induced climate change.” This is not a novel question as the relationship between climate change and the increasing occurrence and intensity of Mediterranean MHWs is already well documented in the literature.Response:
We will revise this sentence. In the revised Abstract we will not present this as a novel question in itself. Instead, we will state that the 2025 event is analysed in the context of an already established literature on climate change and Mediterranean MHWs, and that our contribution is to provide an event specific analogue based line of evidence and process oriented interpretation.Line 19:
It is unclear how this result provides new insights compared to existing studies that have addressed similar questions using other methods.Response:
We will revise the Abstract to state more clearly what is specific to the present study, namely the use of an SST analogue framework to compare physically similar events across different climate backgrounds, and to diagnose how the thermodynamic context differs between past and present analogue sets. We will avoid implying novelty where the result mainly confirms existing understanding.Line 25:
"...method for assessing extreme events, also including marine heatwaves, in a warming Mediterranean context"Response:
We will revise this wording for clarity and precision.Introduction
Line 29:
A MHW is not an extreme weather event, it is an extreme marine event.Response:
We will correct this terminology throughout the manuscript. Indeed, although MHWs are extreme marine events, they are closely linked to atmospheric drivers especially in the Mediterranean Sea both in their onset and intensification. In Sen et al (20220) the authors state that “many of the extreme subtropical MHWs were associated with anomalous high-pressure systems during their build-up”. With these atmospheric settings, reduced cloudiness and increased solar radiation come together with weak winds that may weaken heat losses from the ocean and strengthen upper ocean level stratification which, in turn, weakens vertical mixing. This process has been documented in the Mediterranean by Paredes-Fortuny et al (2025) in the case of atmospheric and marine heatwaves concurrence.Sen Gupta, A., Thomsen, M., Benthuysen, J.A. et al. Drivers and impacts of the most extreme marine heatwave events. Sci Rep 10, 19359 (2020). https://doi.org/10.1038/s41598-020-75445-3
Paredes-Fortuny, L., Pastor, F. & Khodayar, S. Concurrent atmospheric heatwaves intensify marine heatwaves through air-sea heat flux change in the Mediterranean Sea. Commun Earth Environ 6, (2025).
Lines 40-45:
This literature background on global MHW impacts is already known and less relevant to the 2025 Mediterranean event. The paragraph could be condensed to focus on Mediterranean MHW impacts and the specific insights the study seeks to gain using the analogue method.Response:
We will condense this paragraph and focus the Introduction more tightly on Mediterranean MHWs, the 2025 event, and the specific role of the analogue approach.Line 51:
...The Mediterranean Sea...Response:
We will correct this.Lines 51-53:
How is the series of record-breaking marine heatwaves (why not MHW by the way?) in the past two decades is different from the immediately previous sentence on frequency, intensity and duration of MHWs? These two sentences convey the same information.Response:
We will merge and simplify these sentences and use consistent terminology. In addition to the list of record-breaking events we will assess the annual number of MHW days. As stated in Pastor et al. "The annual mean value (black line inside the boxplot) had small variation from 1982 to 2000, except for the years 1990 and 1994 with the highest values in the period but experienced a remarkable increase since 2000 and especially in the last decade with the linear trend for the annual mean increasing from 1 day/year in 1982–1999 to 5.5 days/year in the 2000–2021 period”Pastor, F. & Khodayar, S. Marine heat waves: Characterizing a major climate impact in the Mediterranean. Science of the Total Environment 861, (2023).
Lines 56-60:
This is not a new regime. MHWs are already known to be becoming more frequent and intense as a results of the mean warming. Especially in the Mediterranean Sea. It is unclear what is novel here. The sentence is also too long: terms like “conditioning synoptic circulation” need clarification. What exactly is meant by these phrases?Response:
We will rewrite this passage. We will avoid presenting the general intensification of MHWs as novel and clarify the circulation-related wording. In the revised manuscript, we will state more precisely that MHW intensity can be enhanced when they co-occur with atmospheric heatwaves (AHWs). We will also clarify that trends indicate longer and more intense MHWs, together with an increase in the frequency of AHWs, leading to more frequent co-occurrences. Rather than referring to a “new regime”, we will frame this as an additional mechanism through which climate change may contribute to the intensification of MHWs.
Pastor, F., Paredes-Fortuny, L. & Khodayar, S. Mediterranean marine heatwaves intensify in the presence of concurrent atmospheric heatwaves. Commun Earth Environ 5, 797 (2024).
Lines 61-64:
These two sentences are redundant. The question raised here and the effect of anthropogenic climate change on individual extreme events is already well established globally and regionally. It would be more useful to focus on the analogue method itself and clarify how it complements or differs from previous studies addressing similar questions in the Mediterranean Sea.Response:
We will revise this part accordingly. We will reduce redundancy and place more emphasis on how the analogue framework complements other attribution approaches rather than restating the general attribution question.Line 75:
The role of atmospheric dynamics in Mediterranean MHWs has already been highlighted multple times in previous observational and modeling studies. It would be helpful to clarify what (or if) the analogue method adds beyond these earlier investigations of atmospheric forcing, for the specific 2025 event.Response:
We will clarify this point in the Introduction and Discussion. In particular, we will explain that the analogue framework allows us to compare events with similar SST pattern structure under different climate backgrounds, thereby providing event specific process oriented context rather than a wholly new general mechanism.Lines 86-90:
This reads more like a Methods section than an Introduction. Consider moving the detailed description of the analogue approach to Methods and in the Introduction focus on clearly stating the main goal of the paper: investigating the 2025 MHW using the analogue method, with only a brief descriptive mention of its core concept.Response:
We will move technical detail from the Introduction to the Methods section and keep only a concise conceptual description in the Introduction.Data: There should be a section describing the method and datasets together. Not separately.
Response:
We will reorganize these sections to improve readability and avoid repetition. The revised manuscript will present data and methods in a more integrated and coherent way.Line 109-110:
It is unclear what is meant by “average” here. Is this a spatial average over 20-30 June, a temporal average, or both? the daily SST anomalies calculated first and then averaged, or was the (temporal) mean SST over 20-30 June computed before subtracting the climatology ( in space)? Do these approaches yield similar results?Response:
The averaging corresponds to a temporal mean over the period 20–30 June. Daily SST anomalies are first computed for each calendar day by subtracting the corresponding climatological mean for that day over the historical period, and then averaged over the selected period. This will be clarified
Are these the actual dates (20-30 June) of the 2025 MHW? If so, the Introduction should include a brief description of the event, rather than just listing the dates, to explain why this period was selected.
Response:
We will clarify in the manuscript why this period was selected and provide a short event description in the relevant introductory part of the paper.Line 115:
This appears redundant with the previous sentence. Additionally, the term “distances” is introduced for the first time here and is unclear. Please clarify its meaning by describing the method where the method section is.Response:
We will remove redundancy and define the distance metric clearly in the Methods section before using the term elsewhere.Lines 126-132:
The dataset details here are excessive; providing just the link is sufficient. The term “refer to” is unclear and it should be explained why this observational dataset was chosen.Response:
We will simplify the dataset description and explain more clearly why the dataset was selected.Lines 133-135:
Do you mean the daily Mediterranean-mean SST from the satellite dataset? If so, please clarify. It is also unclear how the timing and magnitude of the 2025 MHW can be compared using a 1991-2020 baseline when the ERA5 anomalies discussed in the study are based on 1950-2025. Can similar anomalies be expected with different climatologies? Since the satellite observations cover 1991-2020, shouldn’t the “present-climate” period be the same to allow a proper comparison? Additionally, please clarify which years are considered as “previous years.”Response:
We will clarify these definitions and the role of the different datasets and climatological reference periods. In particular, anomalies are evaluated for calendar days while baseline is averaged over the period of interests. It is not the same approach and in attribution studies it is used to compare anomalies rather than baselines. We will also make explicit which years are referred to. We do not plan an additional sensitivity analysis with alternative climatologies, but we will improve the text so that the comparison is not overstated and the limits of comparing across reference periods are acknowledged.Lines 136-137:
I do not understand how OISST is being used. What is meant by “descriptive purposes”? If Level-4 AVHRR data are already used, what is the added value of including another observational dataset? Please clarify how this dataset is applied and which reference period was used to compute anomalies.Response:
Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June.Methodology Line 143: maybe you meant analogue events? Also the description feels a lot like repeated from the introduction.
Response:
We will correct the wording and reduce repetition.Line 144:
The phrase “dynamical situation” may be unclear. Here, it would be more precise to refer specifically to the relevant atmosphere state or circulation pattern that drives the observed SST anomalies, rather than implying the entire coupled system, since no ocean variables are included.Response:
We will adopt more precise wording and avoid implying that the full coupled system is represented.Line 148:
Since this refers to an SST anomaly pattern, it should pertain only to the sea, including coastal regions, unless another aspect is intended, which should be clarified.Response:
We will clarify this point.Lines 152-155:
In this description It is unclear whether the closest analogues at each grid point occur on the same days, or how analogue days are selected. Is a spatial average of the distances across grid points computed for each day to identify candidate analogue dates? Please clarify how analogue days are determined consistently.Response:
We will rewrite this methodological description so that the analogue selection procedure is explicit and reproducible.Lines 155-159:
What is meant by “fix the number of analogues,” ? You also mention that earlier sensitivity analyses provided robust results. Please specify which analyses were performed and how they support this conclusion.Response:
We will clarify what is meant by fixing the number of analogues and specify more clearly what earlier sensitivity tests refer to. We will not add new analyses, but we will describe more precisely the basis for the methodological choice already made.Lines 162-163:
You have already stated that in the beginning of the paragraph. Maybe merge?Response:
We will merge these sentences.Lines 167-168:
what do you mean by “avoid introducing biases from model simulations”?
ERA5 is a reanalysis product that assimilates observations and therefore inherently includes biases and errors, so the purpose of this statement is unclear.Response:
We will revise this sentence. Our intent was to distinguish the present observationally constrained framework from a purely climate-model generated ensemble framework, not to imply that reanalysis is bias free. We will correct the wording accordingly.Lines 170-171:
It is still unclear why 42 events were selected, why this number and not another? Additionally, the manuscript has already emphasized multiple times that the analysis compares different background climate conditions.Response:
We select the top 1% of analogue dates, corresponding here to 42 events, and we will state this criterion explicitly in the Methods.Line 178:
main drivers of MHWs during the summer perhaps?Response:
We will revise accordingly.Results Lines 215-216:
Before presenting the Mediterranean-averaged SST anomaly timeseries, it would be more informative to show a map of the event’s spatial pattern. This would justify focusing on the Western Mediterranean rather than the entire basin, as the choice is presumably due to the event being more intense there? Currently, the paragraph implies the 2025 MHW spanned the entire basin, but no explanation is provided for narrowing the analysis in the west.Response:
As suggested, Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June. This will also clarify the choice of the period we have used.Line 217:
It is unclear why the 1991-2020 climatology was chosen. According to the Methods, the “present climate” period is 1987-2024, so it would be more consistent to calculate the MHW SST anomalies relative to the same period. If there is a specific reason for using 1991-2020, it should be explained; otherwise, using 1987-2024 would be more appropriate.Response:
To avoid this ambiguity, we will revise Figure 1 to show absolute temperatures and climatology rather than anomalies. This will avoid problems with using different baseline.Lines 221-223:
would these results still hold if a climatology of 1987-2024 was used?Response:
We will revise the wording so that no statement depends on an untested alternative climatology. We will also clarify the descriptive role of the anomaly reference period.Lines 225-227:
What is meant by “in the context of recent warming”? Which period does “recent warming” refer to? Also, would this statement still hold if the climatology were shifted to 1987-2024?Response:
We will define this wording more clearly and avoid any formulation that suggests greater precision than provided by the current setup and include a brief discussion of this paper that confirms the acceleration of MHW intensification (intensity, duration, frequency…) in the 2010-2020 decade with respect to the trend in 1982-2020.Juza, M., Fernández-Mora, A. & Tintoré, J. Sub-Regional Marine Heat Waves in the Mediterranean Sea From Observations: Long-Term Surface Changes, Sub-Surface and Coastal Responses. Front Mar Sci 9, (2022).
Lines 228-230:
The reader does not see a map of SST anomalies for the entire basin. Fig. 2a shows only the Western Mediterranean. Later, a west-east gradient is mentioned, but it is unclear what this refers to. Presumably it is a basin-wide gradient across the Mediterranean rather than confined to the Western part. This should be clarified.Response:
We will add a map of the entire basin in the new Figure 1. This new map will make it clear why then we focus on this region. We will remove sentences referring to domains not shown in the figure.Lines 233:
The phrase “embedded in a basin-wide warming” is misleading for two reasons: (1) the 2003 and 2022 events exhibited basin-wide warming except in the Aegean Sea, which was cooler than normal (Darmaraki et al., 2024); (2) a MHW cannot be “embedded” within another MHW, as each event is defined by its own spatial extent.Response:
We will rephrase this sentence to avoid ambiguity. We will describe more precisely the spatial structure of the event without using the term “embedded” and avoid implying a uniform basin-wide signal. We also note that MHWs can exhibit spatial heterogeneity and evolve in time, which will be reflected more clearly in the revised wording.
Line 237:
This belongs in the Discussion rather than the Results. Additionally, the cited paper discusses co-occurrence of atmospheric and marine heatwaves in general and does not specifically address the 2025 event.Response:
We will move this type of interpretive comparison to the Discussion.Lines 237-240:
The figures label the variable as MSL, not MSLP, this should be made consistent throughout. Also, the description of the event characteristics is somewhat disorganized, starting with Fig. 2i and then providing separate comparisons with the literature for each atmospheric variable. It would be clearer to first describe all atmospheric variables for the 2025 event and then discuss past, present and Δ differences for each field. Comparisons with previous studies should be moved to the Discussion section, as the current order makes the narrative confusing.Response:
We will ensure consistent notation and reorganize this section so that variables are first described clearly from the figures, while comparisons with literature are transferred to the Discussion.Line 242:
We are still in the Results section, so it is premature to compare with the literature. In particular, there is no evidence supporting the claim that this synoptic pattern corresponds to a northward shift of the subtropical ridge. Such a statement should be framed as a hypothesis or possibility and addressed in the Discussion section.Response:
We will revise such statements accordingly. Where not directly demonstrated by the present analysis, they will be framed as hypotheses or interpretations and moved to the Discussion.Line 244:
Again, this is Results section, the focus should be on presenting the findings. Discussion of how these results relate to other studies belongs in the Discussion section.Response:
We will revise the structure accordingly.Line 245:
According to the caption of Figure 2.1, the data represent the average over 20-30 June, so no figure shows the specific conditions on 30 June as mentioned in the text.Response:
We will correct this inconsistency.Line 246:
The analysis does not include cloud cover, only wind is shown. Please be precise about which variable is being discussed and refer to the specific figure. Currently, results are mixed with general theory about synoptic conditions associated with Mediterranean MHWs, which reduces clarity.Response:
We will remove references to cloud cover, since this variable is not shown. We will also revise the text to refer only to displayed variables and to separate direct results from general interpretation.Line 248:
If this is the case, it should be addressed in the Discussion section. The Results section on the 2025 atmospheric conditions should first present a clear, coherent description of all relevant variables and only afterward should the writers discuss the likely interpretations or implications of each, in the Discussion section.Response:
We will revise the structure accordingly.Lines 253-254:
This result is not novel and has already been documented in previous Mediterranean studies. This discussion belongs in the Discussion section. Currently, theoretical knowledge about the co-occurrence of atmospheric and marine heatwaves is mixed with the presentation of your results. While the observed 6 °C air-temperature anomaly may suggest a local atmospheric heatwave, this has not been formally demonstrated, as no climatology or definition of atmospheric heatwaves was applied here. As such, the statement remains speculative.Response:
We will move this interpretation to the Discussion and tone down the wording. We will avoid implying a formally identified atmospheric heatwave where no dedicated definition has been applied.Line 255:
Again, I did not see anywhere the cloud variable been presented.Response:
We will delete this reference.Line 259-261:
What is the purpose of Figure 1 without a map showing the full spatial pattern of the event across the entire basin? The figure shows anomalies relative to other years, but the reader cannot see how the 2025 MHW was distributed across the Mediterranean.Response:
Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June.Line 263:
..record-breakingResponse:
We will revise this wording as needed for precision.Line 265:
Did the studies you compare your results with use the same climatology to define the 2022 events? If different climatologies were used, it is unclear how these comparisons are valid. In any case this belongs to the Discussion section.Response:
We will move such comparisons to the Discussion and phrase them more cautiously.Lines 266-267:
There is no information on the spatial extent of the event or how it differs from previous events. From Figure 2a, it is not clear that >2 °C anomalies cover a larger area than in prior events. The statement that the spatial extent is “larger” (larger than what?) is unclear and unsupported. The figure only shows a concentration of warm anomalies in the Western Mediterranean, which has been observed in past events in this region. The claim, as currently stated, is not substantiated. The authors should clearly demonstrate the validity of their results.Response:
We will compute the surface >2°C in the data so that the statement will be more preciseLines 268-269;
Again, comparisons with previous studies belong in the Discussion section. Moreover, have you verified whether the anomalies in those studies were calculated using the same climatology as yours? Without this, the comparison is not meaningful.Response:
We will move these statements to the Discussion and temper the comparisons.Line 269:
which event? the 2003 or 2025? clarify.Response:
We will clarify this.Line 270:
It is unclear which figure this description refers to. Please clarify where the reader is expected to look.Response:
We will revise the figure references carefully throughout.Lines 273-275:
This statement is expected given the mean warming of the basin. it is unclear what is novel about this argument.Response:
We will revise the wording to avoid overstating novelty.Line 278:
...described in Section 2..Response:
We will correct this cross reference.Lines 281-287:
This content does not present results. It corresponds to a figure caption and is not relevant to the main text.Response:
We will remove or condense this passage.Lines 289-292:
Aside from the warmer background and increased MHW frequency, these results are largely expected: the 2025 event falls right after the present period, so SST patterns and atmospheric conditions are naturally more aligned with present-day climatology than with past conditions. Again, comparisons with the literature belong in the Discussion section.Response:
We will move literature based interpretation to the Discussion and reduce wording that presents expected findings as more novel than they are. At the same time, we will keep the attribution relevance of the comparison, but express it more carefully as an event based observational line of evidence.Lines 292-294:
Which figure does this refer to? If it refers to SST in Fig. 2d, it only shows the Western Mediterranean and the Adriatic. Why do the event figures focus solely on the Western Mediterranean, while the atmospheric variables cover the entire basin? For consistency, the full basin should be shown for all variables.Response:
We will clarify the figure references and domain descriptions. We do not plan new figures, but the revised text will make the domain choices explicit and avoid confusion.Line 294-295:
Co-variability of what? Clarify.Response:
We will clarify this.Lines 295-297:
In Figure 2h, only parts of southern Europe and North Africa are shown, so it makes sense that the text refers to values in these regions. Additionally, the statement that the “actual event exceeds both composites” is unclear. Which composites are being referred to, and what type of exceedance is meant, given that Fig. 2a-f shows SST over the Mediterranean, while the other variables cover land and atmosphere? This needs to be clearly explained.Response:
We will rewrite this passage to specify exactly which composites are being compared and in what sense.Lines 297-301:
These belong to the Discussion section again. Why do the MSL patterns suggest a subtropical-ridge warming? The basis for this statement is unclear and not sufficiently substantiated here.Response:
We will move this interpretation to the Discussion and frame it as a possible interpretation rather than as a demonstrated result.Line 304:
Since the study does not demonstrate that the SSR patterns are actually linked to anticyclones, this should be framed as a hypothesis rather than a fact and discussed in the Discussion section.Response:
We will revise accordingly.Lines 308-309:
Has it been demonstrated that higher absolute humidity causes the weaker and more spatially mixed SLHF patterns? If this is a hypothesis rather than a proven result, it should be moved to the Discussion section.Response:
We will move this to the Discussion and make the wording more cautious.Line 309:
The differences in SSHF shown in Fig. 2a-b are actually negative around the Mediterranean basin and near zero over the Mediterranean Sea. It is unclear where the authors see positive sensible heat flux differences.Response:
We will correct the description so that it matches the figure exactly.Lines 313-314:
What is meant by “modest MSL differences”? Fig. 2l shows values around zero or negative, and ΔWSPD exhibits both positive and negative variability, so it is unclear how this is consistent with the MSL differences.
What does “indicating similar dynamics” also refer to? Similar to what? Please clarify. The description does not clearly explain how the thermodynamic background is shifted.
In general, the current Results section mixes figure description with speculative interpretations based on literature. Most claims about the background thermodynamics belong in the Discussion, and the Results should present a clearer, fact-based account of what actually occurred.Response:
We will revise this entire passage so that the Results remain strictly descriptive and figure based, while broader physical interpretation is moved to the Discussion.Line 325:
What defines an “anomalous circulation pattern”? Did the writers expect that the atmospheric thermodynamics would change the sign of their influence (positive or negative) on the ocean during a MHW? The rationale behind this statement is unclear and should be clarified.Response:
We will clarify or remove this wording.Lines 327-329:
Based on Fig. 2d,h,l,p,t,x,ab,af, the results show:- Higher SST in the Western Mediterranean, but not across the rest of the basin (Fig. 2a).
- Warmer 2m air temperatures over land surrounding the Mediterranean, while over the Sea the anomalies are near zero (Fig. 2h).
- Increased shortwave radiation input (Fig. 2p).
- Reduced net longwave loss only over North Africa. The Mediterranean Sea shows zero or positive STR changes (Fig. 2t).
- Smaller sensible heat flux mainly around the basin, with no clear change over the Mediterranean Sea itself (Fig. 2ab).
These observations indicate that the statement in the text is inconsistent with the figures. Additionally:
- What about latent heat flux and wind (Fig. 2x, af)? These variables are relevant in the analogue framework, and latent heat flux (decrease from the ocean to the atmosphere) has been shown to play an important role in previous studies. Here, only small increases are seen over limited Mediterranean regions. How do the authors explain this discrepancy?
Response:
We thank the referee for this close reading. We will revise the relevant text carefully so that every statement matches the figures exactly. We will also ensure that latent heat flux and wind are discussed where shown. Where an interpretation is only tentative, it will be clearly labelled as such and moved to the Discussion.Lines 329-331: I am not convinced by this analysis of “coherence across independent variables”, as it does not provide clear or robust evidence to support the attribution of the 2025 event’s severity to thermodynamic changes. Also the role of the ocean, such as advection and mixing processes, which are important for Mediterranean MHWs, has been entirely neglected.”
Response: We will soften this wording. Rather than presenting this convergence as definitive proof, we will present it as a consistent pattern across displayed variables within this observational line of evidence. We will also strengthen the discussion of the missing ocean dynamical processes and explicitly acknowledge that natural variability and omitted ocean mechanisms cannot be ruled out. Finally, we will discuss the relatively small contribution of mixing processes as the Mediterranean Sea is strongly stratified in summer.
Lines 335-343: This whole paragraph belongs to the Methods section. The methodology for using these large-scale climate modes is entirely absent from the manuscript. The authors do not specify which indices were used for each mode, their source, whether they are 2D fields or simple time series, or their resolution. As a result, the current analysis relies on datasets that the reader cannot trace or understand. The indices should be fully introduced in the Methods section before presenting the analysis. Otherwise, this section lacks clarity and context.”
Response:
We will move the methodological description of the climate mode indices to the Methods section and provide the missing technical information.
Lines 360-363: The discussion of the positive AMO phase describes its influence on the regional summer climate, but does not explain how these conditions specifically affect the Mediterranean basin. Since no evidence is provided for the mechanism through which these large-scale climate modes to Mediterranean MHWs, the claims remain hypotheses rather than substantiated results. The same applies to all other climate modes discussed in this section. Additionally, “pin-down” is not appropriate for a scientific paper and should be rephrased. Finally, the transition from AMO in the first sentence to ENSO in the next is unclear; the relationship between these two statements should be clarified.”
Response:
We will revise this section to clarify the transitions between climate modes, in particular the shift from AMO to ENSO, and to ensure that the proposed links are explicitly framed as hypotheses or possible interpretations rather than demonstrated mechanisms. We acknowledge that the processes through which large-scale climate modes influence Mediterranean MHWs are not fully constrained by the present analysis, and the text will be amended accordingly. The term "pin-down" will be replaced with more appropriate scientific language. To further support the discussion of SST variability in relation to AMO, we will integrate additional references (Pisano et al., 2020; Marullo et al., 2011), which provide observational evidence of multidecadal SST variability in the Atlantic–Mediterranean region and its connection to AMO.
Pisano, A., Marullo, S., Artale, V., Falcini, F., Yang, C., Leonelli, F. E., ... & Buongiorno Nardelli, B. (2020). New evidence of Mediterranean climate change and variability from sea surface temperature observations. Remote Sensing, 12(1), 132.
Marullo, S., Artale, V., & Santoleri, R. (2011). The SST multidecadal variability in the Atlantic–Mediterranean region and its relation to AMO. Journal of Climate, 24(16), 4385-4401.
“Lines 368-372: Figure 3 only compares the distributions of climate mode indices during analogue dates in past and present periods. There is no direct analysis linking these indices to the 2025 MHW, no correlation, regression, mechanism, or attribution is presented. Therefore, any claim that these large-scale modes “explain” or directly drive the 2025 event is unsubstantiated. Such statements should be framed as hypotheses and placed in the Discussion, where a potential mechanism connecting these modes to the event could be outlined, unless the authors provide evidence demonstrating a causal link.”
Response: We will revise these statements accordingly and avoid causal wording that goes beyond the presented analysis.
“Lines 373- 380: This paragraph is not clear or well explained. Rephrase.”
Response: We will rewrite this paragraph.
“Lines 382-383: This is expected, since 2025 falls within the present period rather than the past. It is unclear what is novel about this claim or why a different outcome would be anticipated.”
Response: We will revise the wording to avoid overstating novelty and make the intended interpretation clearer.
“Discussion:
Lines 386-387: The manuscript does not demonstrate that the 2025 MHW is amplified by anthropogenic climate change. Both the past and present periods used to identify analogues contain contributions from internal variability and climate change. What has been shown is simply that the 2025 MHW has closer analogues in the present period than in the past, which is
naturally anticipated and primarily reflects a warmer background climate rather than isolating the cause of the warming. The differences between the two periods as they are examined here based on ERA5, arise from both anthropogenic forcing and internal variability, or changes in observational/reanalysis datasets. The analogue method detects a shift but does not provide formal attribution to climate change. Even without the analogue method, by using observational datasets alone, (whether a single or multiple datasets), one cannot isolate the effect of climate change properly, because each provides only a single/the same realization of internal variability. Robust attribution of an event to anthropogenic climate change requires the
use of model ensembles designed to separate the influence of internal variability from external forcing.”Response: We will revise this discussion substantially. We will explicitly acknowledge that the analogue method does not by itself separate all externally forced and internally generated contributions, and that natural variability cannot be ruled out. However, we will not remove the attribution framing, because analogue based attribution is an established observational line of evidence in extreme event attribution. In the revised manuscript we will make clear that our results should be interpreted within a multi line evidence perspective, consistent with Otto et al. (2024) and Clarke et al. (2023), and not as a stand alone complete assessment. We will also reformulate the claims so they refer to attribution consistent evidence from an observational analogue framework, rather than to an exhaustive formal quantification.
“Lines 401-402: I do not agree with the statement that these changes are “purely from the climate shift.” The study provides no evidence for this, as it relies on a single observational- based dataset, which, as already underlined in previous comments, cannot isolate the climate signal from internal variability. Furthermore, the role of the ocean has not been considered, representing a major limitation in attributing the causes of the event or its intensity. Such statements should be made with caution.”
Response: We will remove the wording “purely from the climate shift” and replace it with a more cautious formulation that acknowledges the role of internal variability and the omission of some ocean processes.
“Lines 402-404: It is well established from previous studies using formal attribution methods that Mediterranean MHWs are increasingly likely due to climate change. However, the present study does not perform such an attribution. The method only compares synoptic conditions in the “present” versus “past” climate for summer, 2025-like events, which demonstrates thermodynamic amplification during the 2025 MHW, a result that is expected and already documented. Therefore, it is unclear what the novel contribution of this study is: the method, the results, or both.”
Response: We will clarify that the contribution of this study is not to replace formal model based attribution, but to provide an event specific observational analogue based attribution perspective and process oriented interpretation for the 2025 event. We will revise the text so that novelty is not overstated.
“Line 406: The statement “Here we can say the same for the marine heatwave of 2025” is not justified, as the study has not performed a formal climate change attribution. The analysis only shows thermodynamic amplification under present conditions, which makes the 2025 event appear more intense than past events, a result that has already been demonstrated in previous studies using other approaches. The authors should focus on how their method provides unique insights into the characterization of the event, rather than claiming to have quantified the effect of climate change, which has not been done.”
Response: We will revise this sentence. We will keep the attribution perspective but rephrase it more carefully, emphasizing that the analogue framework provides event specific evidence consistent with anthropogenic thermodynamic amplification under comparable pattern conditions, while not claiming a complete probabilistic attribution on its own.
“Lines 420-423: Is the discussion here still about MHWs or more generally about atmospheric or compound heatwaves? Please clarify. Also, what is meant by a “warmer ocean–atmosphere baseline”? Are you referring to sensible heat fluxes? These showed little difference between present and past periods, similar to MSL, so it is unclear which atmospheric field this claim is based on.”
Response: We will clarify the scope of the discussion and rephrase the “warmer ocean atmosphere baseline” statement more precisely.
“Line 426: The claim that a northward shift of the subtropical ridge influenced the 2025 event has not been demonstrated here. It appears to be inferred visually from MSLP patterns and is therefore unsubstantiated. The authors should either provide evidence to support this claim or reframe it as a hypothesis.”
Response: We will reframe this as a hypothesis or possible interpretation.
“Lines 436-438: The manuscript discusses the assumption of stationarity in the analogue approach and argues that pattern similarity in Mediterranean SST anomalies can be meaningfully compared across climates because the spatial structures are largely governed by stationary factors, such as basin geography and recurrent wind systems (e.g., Mistral-driven cooling in the western Mediterranean). While it is true that the Mediterranean basin geometry and dominant atmospheric forcings impose persistent large-scale spatial structures (e.g., west– east contrasts and sub-basin variability), the statement that this likely ensures the validity of the stationarity assumption may be somewhat too strong. A growing body of literature shows that Mediterranean SST trends and variability have evolved over recent decades, with spatially heterogeneous warming rates (often stronger in the eastern
basin) and shifts in basin-scale gradients associated with large-scale modes of variability and long-term warming (e.g. Pastor et al., 2018, Kubin et al 2023). These changes do not necessarily imply that entirely new SST spatial structures appear, but they can modify the magnitude and relative gradients of SST anomalies across the basin. As a result, analogue methods based on historical pattern similarity may still face limitations due to the non- stationarity of the background climate state, particularly when comparing early-period events (e.g., 1950–1986) with recent extremes. I suggest slightly moderating or clarifying this discussion by acknowledging that, although Mediterranean SST anomaly patterns are indeed constrained by basin geography and recurrent atmospheric forcing, climate change may still alter anomaly magnitudes and spatial gradients in ways that reduce the representativeness of historical analogues. This would provide a more balanced description of the stationarity assumption underlying the analogue framework”Response: We will revise this discussion in exactly that spirit. We will moderate the wording and state more clearly that stationarity is an assumption of the framework and a source of limitation, especially under evolving spatial gradients and background warming.
“Line 444: The argument that moderate analogue distances justify using analogues for attribution is not valid. Moderate distances only indicate that the 2025 event resembles present patterns; they do not provide evidence that the method can isolate or quantify the effect of climate change. Attribution requires separating the influence of external forcing from internal variability, which analogues alone cannot do. Using analogue similarity as a “justification for attribution” conflates pattern resemblance with causal attribution, so this claim is not substantiated. Why is this considered sufficient justification for using analogues for attribution, when other established methods for climate change attribution exist? Please clarify.”
Response: We will revise this point carefully. We will not use moderate analogue distances as a stand alone justification for attribution. Instead, we will state that analogue similarity underpins the physical comparability of events within this observational framework. The attribution relevance then comes from comparing similar patterns under different climate backgrounds, as one line of evidence. We will also clarify that this does not replace other established attribution methods and should be interpreted within a broader multi line evidence perspective.
“Lines 447-450: That claim may apply mainly to summer MHWs. In other seasons oceanic advection often dominates MHW development, and even in summer advection can be important in coastal areas while mixing frequently controls MHWs in regions with strong seasonal wind forcing. Thus the statement is incomplete. A more explicit discussion of the potential contribution of ocean dynamics (which are completlye omitted here) would provide a more balanced interpretation of the results.”
Response: We will qualify the statement so that it is clearly restricted to the summer context considered here, and we will expand the limitations discussion regarding ocean dynamics.
“Lines 471-472: This point is valid but already noted in the Introduction, so there is no need to repeat it here.”
Response: We will remove the repetition.
“Lines 473-474: It is unclear from the text which sources or studies support the statement that the 2025 event likely caused or will cause significant ecological stress and where. The authors should specify the evidence or references underlying this claim.”
Response: We will either document this statement more clearly or remove it if not sufficiently supported in the current scope.
“Lines 474-475: As with many previous events and studies, higher SSTs naturally lead to more intense MHWs. Defining each new event relative to a fixed past baseline means that increasingly extreme events are expected every year. Therefore, it is unclear what is novel in these results.”
Response: We will revise the wording to avoid overstating novelty and to focus on the event specific diagnostic contribution.
“Lines 476-477: Once again, this is not a climate change attribution. The study only applies a method showing an amplified thermodynamic background for SST anomalies. It is unclear whether the same result would hold if the climatology reference period were changed.”
Response: We will revise the conclusion so that it does not claim more than supported by the present line of evidence. At the same time, we will keep the attribution language by making explicit that the method provides observational attribution evidence based on analogue comparison, while not ruling out internal variability and while not claiming a full multi method attribution assessment.
“Lines 480-482: You have not demonstrated how the ocean or the atmospheric variables have responded specifically to climate change.The analysis only shows how atmospheric variables have changed between the past 30 years and the present 30 years, without isolating the effect of climate change.”
Response: We will make this limitation clearer and revise the wording to avoid implying full isolation of the climate change signal from all other factors.
“Lines 480-491: The authors’ claim that the 2025 event would have been “extremely unlikely” under pre-industrial conditions is not supported by the methods used here. The analogue approach identifies historical cases with similar synoptic patterns to characterise variability. However, it does not provide the framework required to isolate the climate change signal. Without a formal framework (e.g., model ensembles) the study cannot demonstrate the event’s probability in a pre-industrial climate. Inferring causation from the absence of historical analogues conflates rarity with attribution and is circular. I recommend removing or substantially rephrasing this claim unless a proper attribution analysis is performed; Alternatively, present it only as a hypothesis and move the discussion to the Discussion section. Using a single, or even multiple, observational datasets cannot isolate the climate change signal from internal variability. Proper separation of these effects typically requires the use of climate model ensembles. This should be mentioned somewhere as a limitation of this
study,”Response: We will substantially rephrase this statement. In the revised manuscript we will avoid claims about pre industrial probability that are too strong for the present framework. We will instead state that the analogue based evidence is consistent with a substantial anthropogenic contribution to the thermodynamic amplification of the 2025 event, while explicitly acknowledging that this observational line of evidence alone cannot fully separate forced change from internal variability and does not provide a complete probabilistic counterfactual assessment. This limitation will be stated clearly.
Citation: https://doi.org/10.5194/egusphere-2025-5055-AC2
-
AC2: 'Reply on RC2', Davide Faranda, 27 Mar 2026
Peer review completion
Post-review adjustments
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2025-5055', Anonymous Referee #1, 18 Jan 2026
Summary: The manuscript presents a novel application of the analogue method to attribute a Mediterranean heat wave to anthropogenic climate change. Although the application of the analog method for attribution is not new, this study adopts a different perspective: instead of searching for analogs of the atmospheric circulation and then comparing the observed extreme with those analog dates, this study searches for analogs dates of the observed extremes and then compares the atmospheric circulation of those dates with that observed at the dates of the extreme.
The conclusion reached for this particular Mediterranean heat wave is that the atmospheric dynamics in the past, while similar to that observed in the extreme, were linked to higher radiative and thermodynamic forcings.
Recommendation: I found the study interesting and, as I wrote above, considering this novel perspective that complements the standard application of the analog method for attribution. I have a recommendation that the authors may want to consider in a revised version
Main points:
1) The title and the abstract do not suggest that the perspective adopted in this study is different from other applications of the analog method. So the authors could better highlight this novel aspect. When reading the title, I first thought that this would be 'just another attribution study'. However, it is not, but this becomes clear when reading the complete manuscript. So I would recommend amending the title and the abstract.
2) In the conclusions section, the authors fall into their own trap. and write some sentences as if the conclusions came from a standard analog-attribution study. For instance, in line 389 they write 'This approach yields a physically transparent storyline: given the same atmospheric/oceanic setup that lead to the 2025 heatwave, a mid-20th-century climate would have produced a far less severe event'.
However, this is not what the study has done. It has not explored the outcome of the *same* (or similar) atmospheric/oceanic set up - it has explored the differences in the predictors of analogous heat waves
3) This different attribution scope, though interesting, makes it more difficult to quantify the impact of climate change. In the traditional approach, the difference between, say, the mean of past analogs and the observed extreme would be the climate change signal. But how can one quantify, with the approach presented in this manuscript, the contribution of climate change? Actually, the authors do not provide a number, perhaps because it is indeed more difficult to obtain.
This should be discussed.
Minor points
4) 'we compared historical sea surface temperature (SST) patterns from ERA5 reanalysis (1950–2024) to those observed during the event.'
The SST from the ER5 reanalysis come from two different sources, as the manuscript clearly states, depending on the period. However, in neither of these two periods are the SSTs a 'reanalysis' product in the classical sense. The OSTIA SSTs are indeed the result of a data assimilation methodology, but the 'model' is very simple, just persistence of the previous time step. A reader may think that SSts are also the result of data assimilation of observations and a proper ocean model.
5) Figure 1. 'Time series of daily sea surface temperature (SST) anomalies averaged (standard deviation
shaded in grey)'
I cannot see the shaded area in the downloaded pdf file. Perhaps the rendering of the figure is not totally correct
6) Section 5 consists of just one big paragraph (!). I would recommend splitting the paragraph, perhaps by the variables being described, to help the reader.
7) 'STR (q–t) presents positive Present–Past differences, implying reduced net longwave loss in the present climate, consistent with a warmer, moister atmosphere returning more longwave radiation to the surface (Trenberth et al., 2015; Vautardet al., 2023)'
I think this paragraph and the corresponding figure warrant a more detailed discussion or description. In principle, attributing it to climate change would entail stronger long-wave forcing from greenhouse gases. Feedbacks increase specific humidity, thereby amplifying the greenhouse effect. But in theory, another scenario is possible: solar radiation increases, warming the ocean surface and increasing specific humidity. In this scenario, anthropogenic climate change would not be ultimately responsible, or only indirectly through reduced cloudiness. So there are subtle questions to discuss based on this figure.
A clean attribution would find that the same MHW is linked to a similar or smaller solar radiation and to increased longwave radiation. I am aware that this reduces the number of possible analogs, but , again, this seems to me to be a complication of this novel attribution approach.
Citation: https://doi.org/10.5194/egusphere-2025-5055-RC1 -
AC1: 'Reply on RC1', Davide Faranda, 27 Mar 2026
Response to Referee #1 (comments in bold, answers in normal text)
We thank the referee for the careful reading of the manuscript and for the constructive and encouraging comments. We appreciate the recognition that the study proposes a perspective that differs from the standard use of analogue based attribution approaches. The suggestions provided by the referee will help us clarify the contribution of the manuscript and improve the presentation of the results.
In the revised manuscript we will better highlight the methodological perspective adopted here. In particular, we will make clearer that the study does not follow the classical analogue attribution framework in which atmospheric circulation analogues are first identified and then associated extremes are analysed. Instead, we start from the observed extreme SST anomaly pattern and search for analogous SST patterns in the historical record, and only afterwards analyse the atmospheric and thermodynamic context associated with those events. We will clarify this conceptual point more explicitly in the title, abstract, and introduction.
Below we provide a point-by-point reply to each comment. Referee’s comments are provided in bold font, while our replies are in normal font.
“Summary: The manuscript presents a novel application of the analogue method to attribute a Mediterranean heat wave to anthropogenic climate change. Although the application of the analog method for attribution is not new, this study adopts a different perspective: instead of searching for analogs of the atmospheric circulation and then comparing the observed extreme with those analog dates, this study searches for analogs dates of the observed extremes and then compares the atmospheric circulation of those dates with that observed at the dates of the extreme. The conclusion reached for this particular Mediterranean heat wave is that the atmospheric dynamics in the past, while similar to that observed in the extreme, were linked to higher radiative and thermodynamic forcings.”
Response:
We thank the referee for recognizing the different methodological perspective adopted in the manuscript. As suggested, we will emphasize this distinction more clearly in the revised manuscript. In particular, we will clarify that the analogue search is performed in the SST anomaly space rather than in circulation space, and that the subsequent analysis compares the atmospheric and thermodynamic context of those SST analogue events across different climate backgrounds. This clarification will be introduced more explicitly in the title, abstract, and early part of the introduction.“Recommendation: I found the study interesting and, as I wrote above, considering this novel perspective that complements the standard application of the analog method for attribution. I have a recommendation that the authors may want to consider in a revised version”
Response:
We thank the referee for the positive assessment of the study and for the constructive recommendations.Main points
“1) The title and the abstract do not suggest that the perspective adopted in this study is different from other applications of the analog method. So the authors could better highlight this novel aspect. When reading the title, I first thought that this would be 'just another attribution study'. However, it is not, but this becomes clear when reading the complete manuscript. So I would recommend amending the title and the abstract.”
Response:
This is a very useful suggestion. In the revised manuscript we will modify both the title and the abstract to make clearer that the study adopts a different analogue based attribution perspective. Specifically, we will emphasize that the analogue search is conducted in SST anomaly space and that the atmospheric circulation and thermodynamic conditions are analysed afterwards for those analogue events. This methodological point will be highlighted earlier in the abstract and in the introduction so that the novelty of the approach is immediately visible.“2) In the conclusions section, the authors fall into their own trap. and write some sentences as if the conclusions came from a standard analog-attribution study. For instance, in line 389 they write 'This approach yields a physically transparent storyline: given the same atmospheric/oceanic setup that lead to the 2025 heatwave, a mid-20th-century climate would have produced a far less severe event'.However, this is not what the study has done. It has not explored the outcome of the same (or similar) atmospheric/oceanic set up - it has explored the differences in the predictors of analogous heat waves”
Response:
We agree with the referee. The original wording was misleading. The analysis does not reconstruct identical atmospheric or oceanic states and therefore does not support a counterfactual statement based on the same dynamical setup. We will revise the text accordingly. The conclusions will be reformulated to state that we compare atmospheric and thermodynamic conditions associated with similar SST anomaly patterns in different climate backgrounds. The attribution will be framed as evidence consistent with thermodynamic amplification, not as a causal statement based on identical dynamics.“3) This different attribution scope, though interesting, makes it more difficult to quantify the impact of climate change. In the traditional approach, the difference between, say, the mean of past analogs and the observed extreme would be the climate change signal. But how can one quantify, with the approach presented in this manuscript, the contribution of climate change? Actually, the authors do not provide a number, perhaps because it is indeed more difficult to obtain. This should be discussed.”
Response:
This is a helpful comment. The approach used here does not provide a single scalar estimate of the climate change contribution, as is often done in traditional analogue-based attribution or model-based frameworks. Instead, the contribution of climate change is assessed through systematic differences in the distributions and composite properties of key variables associated with SST analogue events in different climate backgrounds. In particular, we analyse shifts in thermodynamic indicators and radiative fluxes between past and present analogues, which provide quantitative diagnostics of how the background state has changed. We will clarify this point in the revised manuscript. The method should be understood as providing quantitative, process-based diagnostics rather than a single aggregated attribution metric. This perspective is complementary to more classical approaches and contributes an additional line of evidence within a multi-method attribution framework. We will also make explicit that the method does not fully disentangle all sources of low-frequency variability from externally forced trends, which remains a limitation of the observational analogue framework.Minor points
“4) 'we compared historical sea surface temperature (SST) patterns from ERA5 reanalysis (1950–2024) to those observed during the event.' The SST from the ER5 reanalysis come from two different sources, as the manuscript clearly states, depending on the period. However, in neither of these two periods are the SSTs a 'reanalysis' product in the classical sense. The OSTIA SSTs are indeed the result of a data assimilation methodology, but the 'model' is very simple, just persistence of the previous time step. A reader may think that SSts are also the result of data assimilation of observations and a proper ocean model.”
Response:
We thank the referee for pointing out this important nuance. In the revised manuscript we will clarify the wording to avoid implying that the SST field corresponds to a fully coupled ocean reanalysis. We will explicitly state that the SST data in ERA5 are based on observationally constrained products that differ between periods and that they do not result from assimilation into a full ocean circulation model. The text will be revised to describe this more precisely.“5) Figure 1. 'Time series of daily sea surface temperature (SST) anomalies averaged (standard deviation shaded in grey)' I cannot see the shaded area in the downloaded pdf file. Perhaps the rendering of the figure is not totally correct”
Response:
We thank the referee for noting this issue. We will check the figure rendering and ensure that the shading is clearly visible in the revised version of the figure.“6) Section 5 consists of just one big paragraph (!). I would recommend splitting the paragraph, perhaps by the variables being described, to help the reader.”
Response:
We will revise this section and split the paragraph into shorter subsections organized by the variables being discussed. This will improve readability and make the structure of the analysis clearer.
“7) 'STR (q–t) presents positive Present–Past differences, implying reduced net longwave loss in the present climate, consistent with a warmer, moister atmosphere returning more longwave radiation to the surface (Trenberth et al., 2015; Vautardet al., 2023)' I think this paragraph and the corresponding figure warrant a more detailed discussion or description. In principle, attributing it to climate change would entail stronger long-wave forcing from greenhouse gases. Feedbacks increase specific humidity, thereby amplifying the greenhouse effect. But in theory, another scenario is possible: solar radiation increases, warming the ocean surface and increasing specific humidity. In this scenario, anthropogenic climate change would not be ultimately responsible, or only indirectly through reduced cloudiness. So there are subtle questions to discuss based on this figure. A clean attribution would find that the same MHW is linked to a similar or smaller solar radiation and to increased longwave radiation. I am aware that this reduces the number of possible analogs, but , again, this seems to me to be a complication of this novel attribution approach.”
Response:
In the revised manuscript we will expand the discussion of the radiative components associated with the analogue composites. In particular, we will clarify that the interpretation of the longwave radiation signal should be considered in combination with the shortwave component and with the thermodynamic background conditions. We will also explicitly acknowledge that different mechanisms can potentially produce similar surface responses, including increased solar input or thermodynamic feedbacks associated with a warmer and moister atmosphere. In addition, we will also mention that reduced cloudiness is less relevant in the Mediterranean Sea compared to others basins (e.g. North Atlantic) due to very low cloud cover in summer. The revised discussion will therefore present the radiative signals as evidence consistent with thermodynamic amplification in a warmer climate, while acknowledging that the observational framework used here cannot isolate every contributing process. This clarification will make the interpretation more balanced and transparent.Citation: https://doi.org/10.5194/egusphere-2025-5055-AC1
-
AC1: 'Reply on RC1', Davide Faranda, 27 Mar 2026
-
RC2: 'Review on the Attribution of the 2025 Mediterranean Marine Heatwave to Climate Change Using Analogues', Anonymous Referee #2, 15 Mar 2026
Dear editor,
I apologize for the delay in my response. I have taken the time to carefully review the manuscript titled "Attribution of the 2025 Mediterranean Marine Heatwave to Climate Change Using Analogues." My detailed comments and suggestions are provided in the attached PDF document.-
AC2: 'Reply on RC2', Davide Faranda, 27 Mar 2026
Response to Referee #2
We thank the referee for the careful and constructive review. We agree that the original manuscript did not always distinguish clearly enough between figure-based results, physical interpretation, and the scope of the attribution claims. In the revised version, we will substantially revise the structure and framing of the manuscript. In particular, we will separate Results and Discussion more clearly, restrict the Results section to what is directly supported by the figures, and reformulate statements that were too strong or not sufficiently supported.
We will also clarify the scope of the attribution perspective. The analogue framework used here provides an observational and conditional line of attribution evidence, based on comparing similar SST anomaly patterns across different climate backgrounds. It does not constitute a complete formal attribution assessment on its own, and we will make this limitation explicit. At the same time, analogue-based attribution is an established approach in the recent literature, and our intention is therefore not to remove the attribution framing, but to formulate it more precisely and position it clearly within a broader multi-line evidence framework.
Finally, we will strengthen the discussion of limitations, in particular regarding the role of internal variability and the omission of ocean dynamical processes such as advection, mixing, and stratification.
Below we provide a point-by-point response to the referee’s comments.
The authors present an analysis of the 2025 Mediterranean marine heatwave (MHW). Using an analogue-based approach, they examine SST anomalies and synoptic conditions in a past (1950-2024) and a present (1987-2024) period, in order to assess the atmospheric conditions, conducive to the development of similar events in these two periods, in comparison to the conditions associated with the 2025 event. While the results are broadly consistent with the growing body of literature documenting the intensification of Mediterranean MHWs and their atmospheric drivers, the current manuscript does not yet provide sufficiently robust evidence to support its main attribution-related claims.
Response:
In the revised manuscript we will clarify more carefully what kind of attribution claim is being made. We will state explicitly that our approach provides an observational line of evidence for attribution, based on (conditional) analogue comparison across different climate backgrounds, and not a full formal attribution assessment by itself. We will therefore tone down several formulations in the Abstract, Results, and Discussion, while maintaining the attribution framing of the manuscript. We will also explain more clearly what the analogue approach adds here, namely a process oriented and event specific comparison of similar SST patterns under different climate backgrounds.A key concern is the methodological interpretation. Although the analogue method is a useful tool for identifying historical precursors and exploring variability of atmospheric drivers, it cannot on its own provide formal attribution of events to anthropogenic climate change. At several points in the manuscript, the results derived from the analogue framework appear to be treated as a direct proxy for the influence of anthropogenic climate change on the occurrence of the event. In addition, many of the results rely primarily on qualitative comparisons and visual inspection rather than on a clearly defined quantitative framework, which leaves parts of the analysis more descriptive than substantiated. Also, the study focuses on the atmospheric drivers of the 2025 event, with the role of oceanic processes being largely neglected in the analysis and only briefly acknowledged in the limitations of the study.
Response:
We will revise the manuscript to avoid formulations that could be read as equating analogue results with a complete attribution proof. Instead, we will frame the method as one observational line of evidence within a broader attribution framework. We will also make the quantitative parts of the analysis more explicit in the Methods section and reduce descriptive language where it is not directly supported by the figures. In addition, we will expand the discussion of limitations related to ocean processes. In the summer Mediterranean, strong stratification, low mixed layer variability, and weak memory effects suggest a limited role for ocean preconditioning beyond the long term warming background, while atmospheric forcing is likely to play a major role in the development of these events. We will nevertheless make clear that ocean dynamical contributions such as advection and mixing are not explicitly resolved in our framework and cannot be fully excluded.
To strengthen the manuscript, I suggest reconsidering the framing of the conclusions. In particular, the authors may wish to avoid drawing definitive inferences regarding the causality or likelihood of the event under climate change. Such questions are typically addressed through formal attribution studies using climate model ensembles, and the influence of anthropogenic climate change on the increasing likelihood and intensity of Mediterranean MHWs is already well established in the literature. As a result, attempting to infer this within the current framework, without a formal attribution approach, does not substantially advance existing understanding of the topic. Instead, the study could focus more on the specific insights that the analogue approach may provide in comparison to existing methods. Clarifying whether and in what ways, this approach may offer additional physical understanding or diagnostic value specifically for the 2025 event, beyond established analyses would, perhaps, help better define the contribution of the study. At present, however, this added value (if any) is not yet clear.
Response:
We will revise the conclusions along these lines. In particular, we will avoid overly definitive wording on causality and likelihood, and we will better explain the specific contribution of the analogue approach. Our intention is not to duplicate model based attribution studies, but to provide an event based observational (conditional) attribution line of evidence and a process oriented diagnostic interpretation under comparable large scale conditions. We will make this contribution clearer in the Introduction and Discussion, and we will explicitly position the manuscript relative to existing literature showing that Mediterranean MHWs are intensifying in response to climate change.Overall, a major revision that addresses the detailed comments below and refocuses the manuscript on the methodological and physical insights provided by the analogue analysis (rather than on causal attribution) would substantially improve the clarity and impact of the study. In addition, the manuscript would benefit from improvements in structure and clarity. For instance, the introduction remains quite general and is not sufficiently aligned with the specific objectives of the study, while the Results section at times blends results with discussion. A clearer organization and a more cohesive, logically structured narrative would strengthen the presentation of the work.
Response:
We will substantially revise the manuscript structure. The Introduction will be shortened and made more focused on the scientific objective and on the role of analogue based attribution within the broader attribution framework. The Methods and data descriptions will be streamlined and better organized. The Results section will be restricted to what is directly shown. Interpretive and literature based statements will be moved to the Discussion. We will preserve the attribution perspective of the manuscript, but express it more carefully and more consistently.Abstract
Line 14:
“..raising questions about the influence of human-induced climate change.” This is not a novel question as the relationship between climate change and the increasing occurrence and intensity of Mediterranean MHWs is already well documented in the literature.Response:
We will revise this sentence. In the revised Abstract we will not present this as a novel question in itself. Instead, we will state that the 2025 event is analysed in the context of an already established literature on climate change and Mediterranean MHWs, and that our contribution is to provide an event specific analogue based line of evidence and process oriented interpretation.Line 19:
It is unclear how this result provides new insights compared to existing studies that have addressed similar questions using other methods.Response:
We will revise the Abstract to state more clearly what is specific to the present study, namely the use of an SST analogue framework to compare physically similar events across different climate backgrounds, and to diagnose how the thermodynamic context differs between past and present analogue sets. We will avoid implying novelty where the result mainly confirms existing understanding.Line 25:
"...method for assessing extreme events, also including marine heatwaves, in a warming Mediterranean context"Response:
We will revise this wording for clarity and precision.Introduction
Line 29:
A MHW is not an extreme weather event, it is an extreme marine event.Response:
We will correct this terminology throughout the manuscript. Indeed, although MHWs are extreme marine events, they are closely linked to atmospheric drivers especially in the Mediterranean Sea both in their onset and intensification. In Sen et al (20220) the authors state that “many of the extreme subtropical MHWs were associated with anomalous high-pressure systems during their build-up”. With these atmospheric settings, reduced cloudiness and increased solar radiation come together with weak winds that may weaken heat losses from the ocean and strengthen upper ocean level stratification which, in turn, weakens vertical mixing. This process has been documented in the Mediterranean by Paredes-Fortuny et al (2025) in the case of atmospheric and marine heatwaves concurrence.Sen Gupta, A., Thomsen, M., Benthuysen, J.A. et al. Drivers and impacts of the most extreme marine heatwave events. Sci Rep 10, 19359 (2020). https://doi.org/10.1038/s41598-020-75445-3
Paredes-Fortuny, L., Pastor, F. & Khodayar, S. Concurrent atmospheric heatwaves intensify marine heatwaves through air-sea heat flux change in the Mediterranean Sea. Commun Earth Environ 6, (2025).
Lines 40-45:
This literature background on global MHW impacts is already known and less relevant to the 2025 Mediterranean event. The paragraph could be condensed to focus on Mediterranean MHW impacts and the specific insights the study seeks to gain using the analogue method.Response:
We will condense this paragraph and focus the Introduction more tightly on Mediterranean MHWs, the 2025 event, and the specific role of the analogue approach.Line 51:
...The Mediterranean Sea...Response:
We will correct this.Lines 51-53:
How is the series of record-breaking marine heatwaves (why not MHW by the way?) in the past two decades is different from the immediately previous sentence on frequency, intensity and duration of MHWs? These two sentences convey the same information.Response:
We will merge and simplify these sentences and use consistent terminology. In addition to the list of record-breaking events we will assess the annual number of MHW days. As stated in Pastor et al. "The annual mean value (black line inside the boxplot) had small variation from 1982 to 2000, except for the years 1990 and 1994 with the highest values in the period but experienced a remarkable increase since 2000 and especially in the last decade with the linear trend for the annual mean increasing from 1 day/year in 1982–1999 to 5.5 days/year in the 2000–2021 period”Pastor, F. & Khodayar, S. Marine heat waves: Characterizing a major climate impact in the Mediterranean. Science of the Total Environment 861, (2023).
Lines 56-60:
This is not a new regime. MHWs are already known to be becoming more frequent and intense as a results of the mean warming. Especially in the Mediterranean Sea. It is unclear what is novel here. The sentence is also too long: terms like “conditioning synoptic circulation” need clarification. What exactly is meant by these phrases?Response:
We will rewrite this passage. We will avoid presenting the general intensification of MHWs as novel and clarify the circulation-related wording. In the revised manuscript, we will state more precisely that MHW intensity can be enhanced when they co-occur with atmospheric heatwaves (AHWs). We will also clarify that trends indicate longer and more intense MHWs, together with an increase in the frequency of AHWs, leading to more frequent co-occurrences. Rather than referring to a “new regime”, we will frame this as an additional mechanism through which climate change may contribute to the intensification of MHWs.
Pastor, F., Paredes-Fortuny, L. & Khodayar, S. Mediterranean marine heatwaves intensify in the presence of concurrent atmospheric heatwaves. Commun Earth Environ 5, 797 (2024).
Lines 61-64:
These two sentences are redundant. The question raised here and the effect of anthropogenic climate change on individual extreme events is already well established globally and regionally. It would be more useful to focus on the analogue method itself and clarify how it complements or differs from previous studies addressing similar questions in the Mediterranean Sea.Response:
We will revise this part accordingly. We will reduce redundancy and place more emphasis on how the analogue framework complements other attribution approaches rather than restating the general attribution question.Line 75:
The role of atmospheric dynamics in Mediterranean MHWs has already been highlighted multple times in previous observational and modeling studies. It would be helpful to clarify what (or if) the analogue method adds beyond these earlier investigations of atmospheric forcing, for the specific 2025 event.Response:
We will clarify this point in the Introduction and Discussion. In particular, we will explain that the analogue framework allows us to compare events with similar SST pattern structure under different climate backgrounds, thereby providing event specific process oriented context rather than a wholly new general mechanism.Lines 86-90:
This reads more like a Methods section than an Introduction. Consider moving the detailed description of the analogue approach to Methods and in the Introduction focus on clearly stating the main goal of the paper: investigating the 2025 MHW using the analogue method, with only a brief descriptive mention of its core concept.Response:
We will move technical detail from the Introduction to the Methods section and keep only a concise conceptual description in the Introduction.Data: There should be a section describing the method and datasets together. Not separately.
Response:
We will reorganize these sections to improve readability and avoid repetition. The revised manuscript will present data and methods in a more integrated and coherent way.Line 109-110:
It is unclear what is meant by “average” here. Is this a spatial average over 20-30 June, a temporal average, or both? the daily SST anomalies calculated first and then averaged, or was the (temporal) mean SST over 20-30 June computed before subtracting the climatology ( in space)? Do these approaches yield similar results?Response:
The averaging corresponds to a temporal mean over the period 20–30 June. Daily SST anomalies are first computed for each calendar day by subtracting the corresponding climatological mean for that day over the historical period, and then averaged over the selected period. This will be clarified
Are these the actual dates (20-30 June) of the 2025 MHW? If so, the Introduction should include a brief description of the event, rather than just listing the dates, to explain why this period was selected.
Response:
We will clarify in the manuscript why this period was selected and provide a short event description in the relevant introductory part of the paper.Line 115:
This appears redundant with the previous sentence. Additionally, the term “distances” is introduced for the first time here and is unclear. Please clarify its meaning by describing the method where the method section is.Response:
We will remove redundancy and define the distance metric clearly in the Methods section before using the term elsewhere.Lines 126-132:
The dataset details here are excessive; providing just the link is sufficient. The term “refer to” is unclear and it should be explained why this observational dataset was chosen.Response:
We will simplify the dataset description and explain more clearly why the dataset was selected.Lines 133-135:
Do you mean the daily Mediterranean-mean SST from the satellite dataset? If so, please clarify. It is also unclear how the timing and magnitude of the 2025 MHW can be compared using a 1991-2020 baseline when the ERA5 anomalies discussed in the study are based on 1950-2025. Can similar anomalies be expected with different climatologies? Since the satellite observations cover 1991-2020, shouldn’t the “present-climate” period be the same to allow a proper comparison? Additionally, please clarify which years are considered as “previous years.”Response:
We will clarify these definitions and the role of the different datasets and climatological reference periods. In particular, anomalies are evaluated for calendar days while baseline is averaged over the period of interests. It is not the same approach and in attribution studies it is used to compare anomalies rather than baselines. We will also make explicit which years are referred to. We do not plan an additional sensitivity analysis with alternative climatologies, but we will improve the text so that the comparison is not overstated and the limits of comparing across reference periods are acknowledged.Lines 136-137:
I do not understand how OISST is being used. What is meant by “descriptive purposes”? If Level-4 AVHRR data are already used, what is the added value of including another observational dataset? Please clarify how this dataset is applied and which reference period was used to compute anomalies.Response:
Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June.Methodology Line 143: maybe you meant analogue events? Also the description feels a lot like repeated from the introduction.
Response:
We will correct the wording and reduce repetition.Line 144:
The phrase “dynamical situation” may be unclear. Here, it would be more precise to refer specifically to the relevant atmosphere state or circulation pattern that drives the observed SST anomalies, rather than implying the entire coupled system, since no ocean variables are included.Response:
We will adopt more precise wording and avoid implying that the full coupled system is represented.Line 148:
Since this refers to an SST anomaly pattern, it should pertain only to the sea, including coastal regions, unless another aspect is intended, which should be clarified.Response:
We will clarify this point.Lines 152-155:
In this description It is unclear whether the closest analogues at each grid point occur on the same days, or how analogue days are selected. Is a spatial average of the distances across grid points computed for each day to identify candidate analogue dates? Please clarify how analogue days are determined consistently.Response:
We will rewrite this methodological description so that the analogue selection procedure is explicit and reproducible.Lines 155-159:
What is meant by “fix the number of analogues,” ? You also mention that earlier sensitivity analyses provided robust results. Please specify which analyses were performed and how they support this conclusion.Response:
We will clarify what is meant by fixing the number of analogues and specify more clearly what earlier sensitivity tests refer to. We will not add new analyses, but we will describe more precisely the basis for the methodological choice already made.Lines 162-163:
You have already stated that in the beginning of the paragraph. Maybe merge?Response:
We will merge these sentences.Lines 167-168:
what do you mean by “avoid introducing biases from model simulations”?
ERA5 is a reanalysis product that assimilates observations and therefore inherently includes biases and errors, so the purpose of this statement is unclear.Response:
We will revise this sentence. Our intent was to distinguish the present observationally constrained framework from a purely climate-model generated ensemble framework, not to imply that reanalysis is bias free. We will correct the wording accordingly.Lines 170-171:
It is still unclear why 42 events were selected, why this number and not another? Additionally, the manuscript has already emphasized multiple times that the analysis compares different background climate conditions.Response:
We select the top 1% of analogue dates, corresponding here to 42 events, and we will state this criterion explicitly in the Methods.Line 178:
main drivers of MHWs during the summer perhaps?Response:
We will revise accordingly.Results Lines 215-216:
Before presenting the Mediterranean-averaged SST anomaly timeseries, it would be more informative to show a map of the event’s spatial pattern. This would justify focusing on the Western Mediterranean rather than the entire basin, as the choice is presumably due to the event being more intense there? Currently, the paragraph implies the 2025 MHW spanned the entire basin, but no explanation is provided for narrowing the analysis in the west.Response:
As suggested, Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June. This will also clarify the choice of the period we have used.Line 217:
It is unclear why the 1991-2020 climatology was chosen. According to the Methods, the “present climate” period is 1987-2024, so it would be more consistent to calculate the MHW SST anomalies relative to the same period. If there is a specific reason for using 1991-2020, it should be explained; otherwise, using 1987-2024 would be more appropriate.Response:
To avoid this ambiguity, we will revise Figure 1 to show absolute temperatures and climatology rather than anomalies. This will avoid problems with using different baseline.Lines 221-223:
would these results still hold if a climatology of 1987-2024 was used?Response:
We will revise the wording so that no statement depends on an untested alternative climatology. We will also clarify the descriptive role of the anomaly reference period.Lines 225-227:
What is meant by “in the context of recent warming”? Which period does “recent warming” refer to? Also, would this statement still hold if the climatology were shifted to 1987-2024?Response:
We will define this wording more clearly and avoid any formulation that suggests greater precision than provided by the current setup and include a brief discussion of this paper that confirms the acceleration of MHW intensification (intensity, duration, frequency…) in the 2010-2020 decade with respect to the trend in 1982-2020.Juza, M., Fernández-Mora, A. & Tintoré, J. Sub-Regional Marine Heat Waves in the Mediterranean Sea From Observations: Long-Term Surface Changes, Sub-Surface and Coastal Responses. Front Mar Sci 9, (2022).
Lines 228-230:
The reader does not see a map of SST anomalies for the entire basin. Fig. 2a shows only the Western Mediterranean. Later, a west-east gradient is mentioned, but it is unclear what this refers to. Presumably it is a basin-wide gradient across the Mediterranean rather than confined to the Western part. This should be clarified.Response:
We will add a map of the entire basin in the new Figure 1. This new map will make it clear why then we focus on this region. We will remove sentences referring to domains not shown in the figure.Lines 233:
The phrase “embedded in a basin-wide warming” is misleading for two reasons: (1) the 2003 and 2022 events exhibited basin-wide warming except in the Aegean Sea, which was cooler than normal (Darmaraki et al., 2024); (2) a MHW cannot be “embedded” within another MHW, as each event is defined by its own spatial extent.Response:
We will rephrase this sentence to avoid ambiguity. We will describe more precisely the spatial structure of the event without using the term “embedded” and avoid implying a uniform basin-wide signal. We also note that MHWs can exhibit spatial heterogeneity and evolve in time, which will be reflected more clearly in the revised wording.
Line 237:
This belongs in the Discussion rather than the Results. Additionally, the cited paper discusses co-occurrence of atmospheric and marine heatwaves in general and does not specifically address the 2025 event.Response:
We will move this type of interpretive comparison to the Discussion.Lines 237-240:
The figures label the variable as MSL, not MSLP, this should be made consistent throughout. Also, the description of the event characteristics is somewhat disorganized, starting with Fig. 2i and then providing separate comparisons with the literature for each atmospheric variable. It would be clearer to first describe all atmospheric variables for the 2025 event and then discuss past, present and Δ differences for each field. Comparisons with previous studies should be moved to the Discussion section, as the current order makes the narrative confusing.Response:
We will ensure consistent notation and reorganize this section so that variables are first described clearly from the figures, while comparisons with literature are transferred to the Discussion.Line 242:
We are still in the Results section, so it is premature to compare with the literature. In particular, there is no evidence supporting the claim that this synoptic pattern corresponds to a northward shift of the subtropical ridge. Such a statement should be framed as a hypothesis or possibility and addressed in the Discussion section.Response:
We will revise such statements accordingly. Where not directly demonstrated by the present analysis, they will be framed as hypotheses or interpretations and moved to the Discussion.Line 244:
Again, this is Results section, the focus should be on presenting the findings. Discussion of how these results relate to other studies belongs in the Discussion section.Response:
We will revise the structure accordingly.Line 245:
According to the caption of Figure 2.1, the data represent the average over 20-30 June, so no figure shows the specific conditions on 30 June as mentioned in the text.Response:
We will correct this inconsistency.Line 246:
The analysis does not include cloud cover, only wind is shown. Please be precise about which variable is being discussed and refer to the specific figure. Currently, results are mixed with general theory about synoptic conditions associated with Mediterranean MHWs, which reduces clarity.Response:
We will remove references to cloud cover, since this variable is not shown. We will also revise the text to refer only to displayed variables and to separate direct results from general interpretation.Line 248:
If this is the case, it should be addressed in the Discussion section. The Results section on the 2025 atmospheric conditions should first present a clear, coherent description of all relevant variables and only afterward should the writers discuss the likely interpretations or implications of each, in the Discussion section.Response:
We will revise the structure accordingly.Lines 253-254:
This result is not novel and has already been documented in previous Mediterranean studies. This discussion belongs in the Discussion section. Currently, theoretical knowledge about the co-occurrence of atmospheric and marine heatwaves is mixed with the presentation of your results. While the observed 6 °C air-temperature anomaly may suggest a local atmospheric heatwave, this has not been formally demonstrated, as no climatology or definition of atmospheric heatwaves was applied here. As such, the statement remains speculative.Response:
We will move this interpretation to the Discussion and tone down the wording. We will avoid implying a formally identified atmospheric heatwave where no dedicated definition has been applied.Line 255:
Again, I did not see anywhere the cloud variable been presented.Response:
We will delete this reference.Line 259-261:
What is the purpose of Figure 1 without a map showing the full spatial pattern of the event across the entire basin? The figure shows anomalies relative to other years, but the reader cannot see how the 2025 MHW was distributed across the Mediterranean.Response:
Figure 1 will be replaced by a new version displaying absolute temperatures rather than anomalies for both datasets. It will include their corresponding time series, along with a spatial map of the mean sea surface temperature over the basin for the period 20 to 30 June.Line 263:
..record-breakingResponse:
We will revise this wording as needed for precision.Line 265:
Did the studies you compare your results with use the same climatology to define the 2022 events? If different climatologies were used, it is unclear how these comparisons are valid. In any case this belongs to the Discussion section.Response:
We will move such comparisons to the Discussion and phrase them more cautiously.Lines 266-267:
There is no information on the spatial extent of the event or how it differs from previous events. From Figure 2a, it is not clear that >2 °C anomalies cover a larger area than in prior events. The statement that the spatial extent is “larger” (larger than what?) is unclear and unsupported. The figure only shows a concentration of warm anomalies in the Western Mediterranean, which has been observed in past events in this region. The claim, as currently stated, is not substantiated. The authors should clearly demonstrate the validity of their results.Response:
We will compute the surface >2°C in the data so that the statement will be more preciseLines 268-269;
Again, comparisons with previous studies belong in the Discussion section. Moreover, have you verified whether the anomalies in those studies were calculated using the same climatology as yours? Without this, the comparison is not meaningful.Response:
We will move these statements to the Discussion and temper the comparisons.Line 269:
which event? the 2003 or 2025? clarify.Response:
We will clarify this.Line 270:
It is unclear which figure this description refers to. Please clarify where the reader is expected to look.Response:
We will revise the figure references carefully throughout.Lines 273-275:
This statement is expected given the mean warming of the basin. it is unclear what is novel about this argument.Response:
We will revise the wording to avoid overstating novelty.Line 278:
...described in Section 2..Response:
We will correct this cross reference.Lines 281-287:
This content does not present results. It corresponds to a figure caption and is not relevant to the main text.Response:
We will remove or condense this passage.Lines 289-292:
Aside from the warmer background and increased MHW frequency, these results are largely expected: the 2025 event falls right after the present period, so SST patterns and atmospheric conditions are naturally more aligned with present-day climatology than with past conditions. Again, comparisons with the literature belong in the Discussion section.Response:
We will move literature based interpretation to the Discussion and reduce wording that presents expected findings as more novel than they are. At the same time, we will keep the attribution relevance of the comparison, but express it more carefully as an event based observational line of evidence.Lines 292-294:
Which figure does this refer to? If it refers to SST in Fig. 2d, it only shows the Western Mediterranean and the Adriatic. Why do the event figures focus solely on the Western Mediterranean, while the atmospheric variables cover the entire basin? For consistency, the full basin should be shown for all variables.Response:
We will clarify the figure references and domain descriptions. We do not plan new figures, but the revised text will make the domain choices explicit and avoid confusion.Line 294-295:
Co-variability of what? Clarify.Response:
We will clarify this.Lines 295-297:
In Figure 2h, only parts of southern Europe and North Africa are shown, so it makes sense that the text refers to values in these regions. Additionally, the statement that the “actual event exceeds both composites” is unclear. Which composites are being referred to, and what type of exceedance is meant, given that Fig. 2a-f shows SST over the Mediterranean, while the other variables cover land and atmosphere? This needs to be clearly explained.Response:
We will rewrite this passage to specify exactly which composites are being compared and in what sense.Lines 297-301:
These belong to the Discussion section again. Why do the MSL patterns suggest a subtropical-ridge warming? The basis for this statement is unclear and not sufficiently substantiated here.Response:
We will move this interpretation to the Discussion and frame it as a possible interpretation rather than as a demonstrated result.Line 304:
Since the study does not demonstrate that the SSR patterns are actually linked to anticyclones, this should be framed as a hypothesis rather than a fact and discussed in the Discussion section.Response:
We will revise accordingly.Lines 308-309:
Has it been demonstrated that higher absolute humidity causes the weaker and more spatially mixed SLHF patterns? If this is a hypothesis rather than a proven result, it should be moved to the Discussion section.Response:
We will move this to the Discussion and make the wording more cautious.Line 309:
The differences in SSHF shown in Fig. 2a-b are actually negative around the Mediterranean basin and near zero over the Mediterranean Sea. It is unclear where the authors see positive sensible heat flux differences.Response:
We will correct the description so that it matches the figure exactly.Lines 313-314:
What is meant by “modest MSL differences”? Fig. 2l shows values around zero or negative, and ΔWSPD exhibits both positive and negative variability, so it is unclear how this is consistent with the MSL differences.
What does “indicating similar dynamics” also refer to? Similar to what? Please clarify. The description does not clearly explain how the thermodynamic background is shifted.
In general, the current Results section mixes figure description with speculative interpretations based on literature. Most claims about the background thermodynamics belong in the Discussion, and the Results should present a clearer, fact-based account of what actually occurred.Response:
We will revise this entire passage so that the Results remain strictly descriptive and figure based, while broader physical interpretation is moved to the Discussion.Line 325:
What defines an “anomalous circulation pattern”? Did the writers expect that the atmospheric thermodynamics would change the sign of their influence (positive or negative) on the ocean during a MHW? The rationale behind this statement is unclear and should be clarified.Response:
We will clarify or remove this wording.Lines 327-329:
Based on Fig. 2d,h,l,p,t,x,ab,af, the results show:- Higher SST in the Western Mediterranean, but not across the rest of the basin (Fig. 2a).
- Warmer 2m air temperatures over land surrounding the Mediterranean, while over the Sea the anomalies are near zero (Fig. 2h).
- Increased shortwave radiation input (Fig. 2p).
- Reduced net longwave loss only over North Africa. The Mediterranean Sea shows zero or positive STR changes (Fig. 2t).
- Smaller sensible heat flux mainly around the basin, with no clear change over the Mediterranean Sea itself (Fig. 2ab).
These observations indicate that the statement in the text is inconsistent with the figures. Additionally:
- What about latent heat flux and wind (Fig. 2x, af)? These variables are relevant in the analogue framework, and latent heat flux (decrease from the ocean to the atmosphere) has been shown to play an important role in previous studies. Here, only small increases are seen over limited Mediterranean regions. How do the authors explain this discrepancy?
Response:
We thank the referee for this close reading. We will revise the relevant text carefully so that every statement matches the figures exactly. We will also ensure that latent heat flux and wind are discussed where shown. Where an interpretation is only tentative, it will be clearly labelled as such and moved to the Discussion.Lines 329-331: I am not convinced by this analysis of “coherence across independent variables”, as it does not provide clear or robust evidence to support the attribution of the 2025 event’s severity to thermodynamic changes. Also the role of the ocean, such as advection and mixing processes, which are important for Mediterranean MHWs, has been entirely neglected.”
Response: We will soften this wording. Rather than presenting this convergence as definitive proof, we will present it as a consistent pattern across displayed variables within this observational line of evidence. We will also strengthen the discussion of the missing ocean dynamical processes and explicitly acknowledge that natural variability and omitted ocean mechanisms cannot be ruled out. Finally, we will discuss the relatively small contribution of mixing processes as the Mediterranean Sea is strongly stratified in summer.
Lines 335-343: This whole paragraph belongs to the Methods section. The methodology for using these large-scale climate modes is entirely absent from the manuscript. The authors do not specify which indices were used for each mode, their source, whether they are 2D fields or simple time series, or their resolution. As a result, the current analysis relies on datasets that the reader cannot trace or understand. The indices should be fully introduced in the Methods section before presenting the analysis. Otherwise, this section lacks clarity and context.”
Response:
We will move the methodological description of the climate mode indices to the Methods section and provide the missing technical information.
Lines 360-363: The discussion of the positive AMO phase describes its influence on the regional summer climate, but does not explain how these conditions specifically affect the Mediterranean basin. Since no evidence is provided for the mechanism through which these large-scale climate modes to Mediterranean MHWs, the claims remain hypotheses rather than substantiated results. The same applies to all other climate modes discussed in this section. Additionally, “pin-down” is not appropriate for a scientific paper and should be rephrased. Finally, the transition from AMO in the first sentence to ENSO in the next is unclear; the relationship between these two statements should be clarified.”
Response:
We will revise this section to clarify the transitions between climate modes, in particular the shift from AMO to ENSO, and to ensure that the proposed links are explicitly framed as hypotheses or possible interpretations rather than demonstrated mechanisms. We acknowledge that the processes through which large-scale climate modes influence Mediterranean MHWs are not fully constrained by the present analysis, and the text will be amended accordingly. The term "pin-down" will be replaced with more appropriate scientific language. To further support the discussion of SST variability in relation to AMO, we will integrate additional references (Pisano et al., 2020; Marullo et al., 2011), which provide observational evidence of multidecadal SST variability in the Atlantic–Mediterranean region and its connection to AMO.
Pisano, A., Marullo, S., Artale, V., Falcini, F., Yang, C., Leonelli, F. E., ... & Buongiorno Nardelli, B. (2020). New evidence of Mediterranean climate change and variability from sea surface temperature observations. Remote Sensing, 12(1), 132.
Marullo, S., Artale, V., & Santoleri, R. (2011). The SST multidecadal variability in the Atlantic–Mediterranean region and its relation to AMO. Journal of Climate, 24(16), 4385-4401.
“Lines 368-372: Figure 3 only compares the distributions of climate mode indices during analogue dates in past and present periods. There is no direct analysis linking these indices to the 2025 MHW, no correlation, regression, mechanism, or attribution is presented. Therefore, any claim that these large-scale modes “explain” or directly drive the 2025 event is unsubstantiated. Such statements should be framed as hypotheses and placed in the Discussion, where a potential mechanism connecting these modes to the event could be outlined, unless the authors provide evidence demonstrating a causal link.”
Response: We will revise these statements accordingly and avoid causal wording that goes beyond the presented analysis.
“Lines 373- 380: This paragraph is not clear or well explained. Rephrase.”
Response: We will rewrite this paragraph.
“Lines 382-383: This is expected, since 2025 falls within the present period rather than the past. It is unclear what is novel about this claim or why a different outcome would be anticipated.”
Response: We will revise the wording to avoid overstating novelty and make the intended interpretation clearer.
“Discussion:
Lines 386-387: The manuscript does not demonstrate that the 2025 MHW is amplified by anthropogenic climate change. Both the past and present periods used to identify analogues contain contributions from internal variability and climate change. What has been shown is simply that the 2025 MHW has closer analogues in the present period than in the past, which is
naturally anticipated and primarily reflects a warmer background climate rather than isolating the cause of the warming. The differences between the two periods as they are examined here based on ERA5, arise from both anthropogenic forcing and internal variability, or changes in observational/reanalysis datasets. The analogue method detects a shift but does not provide formal attribution to climate change. Even without the analogue method, by using observational datasets alone, (whether a single or multiple datasets), one cannot isolate the effect of climate change properly, because each provides only a single/the same realization of internal variability. Robust attribution of an event to anthropogenic climate change requires the
use of model ensembles designed to separate the influence of internal variability from external forcing.”Response: We will revise this discussion substantially. We will explicitly acknowledge that the analogue method does not by itself separate all externally forced and internally generated contributions, and that natural variability cannot be ruled out. However, we will not remove the attribution framing, because analogue based attribution is an established observational line of evidence in extreme event attribution. In the revised manuscript we will make clear that our results should be interpreted within a multi line evidence perspective, consistent with Otto et al. (2024) and Clarke et al. (2023), and not as a stand alone complete assessment. We will also reformulate the claims so they refer to attribution consistent evidence from an observational analogue framework, rather than to an exhaustive formal quantification.
“Lines 401-402: I do not agree with the statement that these changes are “purely from the climate shift.” The study provides no evidence for this, as it relies on a single observational- based dataset, which, as already underlined in previous comments, cannot isolate the climate signal from internal variability. Furthermore, the role of the ocean has not been considered, representing a major limitation in attributing the causes of the event or its intensity. Such statements should be made with caution.”
Response: We will remove the wording “purely from the climate shift” and replace it with a more cautious formulation that acknowledges the role of internal variability and the omission of some ocean processes.
“Lines 402-404: It is well established from previous studies using formal attribution methods that Mediterranean MHWs are increasingly likely due to climate change. However, the present study does not perform such an attribution. The method only compares synoptic conditions in the “present” versus “past” climate for summer, 2025-like events, which demonstrates thermodynamic amplification during the 2025 MHW, a result that is expected and already documented. Therefore, it is unclear what the novel contribution of this study is: the method, the results, or both.”
Response: We will clarify that the contribution of this study is not to replace formal model based attribution, but to provide an event specific observational analogue based attribution perspective and process oriented interpretation for the 2025 event. We will revise the text so that novelty is not overstated.
“Line 406: The statement “Here we can say the same for the marine heatwave of 2025” is not justified, as the study has not performed a formal climate change attribution. The analysis only shows thermodynamic amplification under present conditions, which makes the 2025 event appear more intense than past events, a result that has already been demonstrated in previous studies using other approaches. The authors should focus on how their method provides unique insights into the characterization of the event, rather than claiming to have quantified the effect of climate change, which has not been done.”
Response: We will revise this sentence. We will keep the attribution perspective but rephrase it more carefully, emphasizing that the analogue framework provides event specific evidence consistent with anthropogenic thermodynamic amplification under comparable pattern conditions, while not claiming a complete probabilistic attribution on its own.
“Lines 420-423: Is the discussion here still about MHWs or more generally about atmospheric or compound heatwaves? Please clarify. Also, what is meant by a “warmer ocean–atmosphere baseline”? Are you referring to sensible heat fluxes? These showed little difference between present and past periods, similar to MSL, so it is unclear which atmospheric field this claim is based on.”
Response: We will clarify the scope of the discussion and rephrase the “warmer ocean atmosphere baseline” statement more precisely.
“Line 426: The claim that a northward shift of the subtropical ridge influenced the 2025 event has not been demonstrated here. It appears to be inferred visually from MSLP patterns and is therefore unsubstantiated. The authors should either provide evidence to support this claim or reframe it as a hypothesis.”
Response: We will reframe this as a hypothesis or possible interpretation.
“Lines 436-438: The manuscript discusses the assumption of stationarity in the analogue approach and argues that pattern similarity in Mediterranean SST anomalies can be meaningfully compared across climates because the spatial structures are largely governed by stationary factors, such as basin geography and recurrent wind systems (e.g., Mistral-driven cooling in the western Mediterranean). While it is true that the Mediterranean basin geometry and dominant atmospheric forcings impose persistent large-scale spatial structures (e.g., west– east contrasts and sub-basin variability), the statement that this likely ensures the validity of the stationarity assumption may be somewhat too strong. A growing body of literature shows that Mediterranean SST trends and variability have evolved over recent decades, with spatially heterogeneous warming rates (often stronger in the eastern
basin) and shifts in basin-scale gradients associated with large-scale modes of variability and long-term warming (e.g. Pastor et al., 2018, Kubin et al 2023). These changes do not necessarily imply that entirely new SST spatial structures appear, but they can modify the magnitude and relative gradients of SST anomalies across the basin. As a result, analogue methods based on historical pattern similarity may still face limitations due to the non- stationarity of the background climate state, particularly when comparing early-period events (e.g., 1950–1986) with recent extremes. I suggest slightly moderating or clarifying this discussion by acknowledging that, although Mediterranean SST anomaly patterns are indeed constrained by basin geography and recurrent atmospheric forcing, climate change may still alter anomaly magnitudes and spatial gradients in ways that reduce the representativeness of historical analogues. This would provide a more balanced description of the stationarity assumption underlying the analogue framework”Response: We will revise this discussion in exactly that spirit. We will moderate the wording and state more clearly that stationarity is an assumption of the framework and a source of limitation, especially under evolving spatial gradients and background warming.
“Line 444: The argument that moderate analogue distances justify using analogues for attribution is not valid. Moderate distances only indicate that the 2025 event resembles present patterns; they do not provide evidence that the method can isolate or quantify the effect of climate change. Attribution requires separating the influence of external forcing from internal variability, which analogues alone cannot do. Using analogue similarity as a “justification for attribution” conflates pattern resemblance with causal attribution, so this claim is not substantiated. Why is this considered sufficient justification for using analogues for attribution, when other established methods for climate change attribution exist? Please clarify.”
Response: We will revise this point carefully. We will not use moderate analogue distances as a stand alone justification for attribution. Instead, we will state that analogue similarity underpins the physical comparability of events within this observational framework. The attribution relevance then comes from comparing similar patterns under different climate backgrounds, as one line of evidence. We will also clarify that this does not replace other established attribution methods and should be interpreted within a broader multi line evidence perspective.
“Lines 447-450: That claim may apply mainly to summer MHWs. In other seasons oceanic advection often dominates MHW development, and even in summer advection can be important in coastal areas while mixing frequently controls MHWs in regions with strong seasonal wind forcing. Thus the statement is incomplete. A more explicit discussion of the potential contribution of ocean dynamics (which are completlye omitted here) would provide a more balanced interpretation of the results.”
Response: We will qualify the statement so that it is clearly restricted to the summer context considered here, and we will expand the limitations discussion regarding ocean dynamics.
“Lines 471-472: This point is valid but already noted in the Introduction, so there is no need to repeat it here.”
Response: We will remove the repetition.
“Lines 473-474: It is unclear from the text which sources or studies support the statement that the 2025 event likely caused or will cause significant ecological stress and where. The authors should specify the evidence or references underlying this claim.”
Response: We will either document this statement more clearly or remove it if not sufficiently supported in the current scope.
“Lines 474-475: As with many previous events and studies, higher SSTs naturally lead to more intense MHWs. Defining each new event relative to a fixed past baseline means that increasingly extreme events are expected every year. Therefore, it is unclear what is novel in these results.”
Response: We will revise the wording to avoid overstating novelty and to focus on the event specific diagnostic contribution.
“Lines 476-477: Once again, this is not a climate change attribution. The study only applies a method showing an amplified thermodynamic background for SST anomalies. It is unclear whether the same result would hold if the climatology reference period were changed.”
Response: We will revise the conclusion so that it does not claim more than supported by the present line of evidence. At the same time, we will keep the attribution language by making explicit that the method provides observational attribution evidence based on analogue comparison, while not ruling out internal variability and while not claiming a full multi method attribution assessment.
“Lines 480-482: You have not demonstrated how the ocean or the atmospheric variables have responded specifically to climate change.The analysis only shows how atmospheric variables have changed between the past 30 years and the present 30 years, without isolating the effect of climate change.”
Response: We will make this limitation clearer and revise the wording to avoid implying full isolation of the climate change signal from all other factors.
“Lines 480-491: The authors’ claim that the 2025 event would have been “extremely unlikely” under pre-industrial conditions is not supported by the methods used here. The analogue approach identifies historical cases with similar synoptic patterns to characterise variability. However, it does not provide the framework required to isolate the climate change signal. Without a formal framework (e.g., model ensembles) the study cannot demonstrate the event’s probability in a pre-industrial climate. Inferring causation from the absence of historical analogues conflates rarity with attribution and is circular. I recommend removing or substantially rephrasing this claim unless a proper attribution analysis is performed; Alternatively, present it only as a hypothesis and move the discussion to the Discussion section. Using a single, or even multiple, observational datasets cannot isolate the climate change signal from internal variability. Proper separation of these effects typically requires the use of climate model ensembles. This should be mentioned somewhere as a limitation of this
study,”Response: We will substantially rephrase this statement. In the revised manuscript we will avoid claims about pre industrial probability that are too strong for the present framework. We will instead state that the analogue based evidence is consistent with a substantial anthropogenic contribution to the thermodynamic amplification of the 2025 event, while explicitly acknowledging that this observational line of evidence alone cannot fully separate forced change from internal variability and does not provide a complete probabilistic counterfactual assessment. This limitation will be stated clearly.
Citation: https://doi.org/10.5194/egusphere-2025-5055-AC2
-
AC2: 'Reply on RC2', Davide Faranda, 27 Mar 2026
Peer review completion
Post-review adjustments
Journal article(s) based on this preprint
Viewed
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 2,494 | 0 | 19 | 2,513 | 0 | 0 |
- HTML: 2,494
- PDF: 0
- XML: 19
- Total: 2,513
- BibTeX: 0
- EndNote: 0
Viewed (geographical distribution)
Since the preprint corresponding to this journal article was posted outside of Copernicus Publications, the preprint-related metrics are limited to HTML views.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Davide Faranda
Thibault Guinaldo
Francisco Pastor
Tommaso Alberti
Samira Khodayar
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
Summary: The manuscript presents a novel application of the analogue method to attribute a Mediterranean heat wave to anthropogenic climate change. Although the application of the analog method for attribution is not new, this study adopts a different perspective: instead of searching for analogs of the atmospheric circulation and then comparing the observed extreme with those analog dates, this study searches for analogs dates of the observed extremes and then compares the atmospheric circulation of those dates with that observed at the dates of the extreme.
The conclusion reached for this particular Mediterranean heat wave is that the atmospheric dynamics in the past, while similar to that observed in the extreme, were linked to higher radiative and thermodynamic forcings.
Recommendation: I found the study interesting and, as I wrote above, considering this novel perspective that complements the standard application of the analog method for attribution. I have a recommendation that the authors may want to consider in a revised version
Main points:
1) The title and the abstract do not suggest that the perspective adopted in this study is different from other applications of the analog method. So the authors could better highlight this novel aspect. When reading the title, I first thought that this would be 'just another attribution study'. However, it is not, but this becomes clear when reading the complete manuscript. So I would recommend amending the title and the abstract.
2) In the conclusions section, the authors fall into their own trap. and write some sentences as if the conclusions came from a standard analog-attribution study. For instance, in line 389 they write 'This approach yields a physically transparent storyline: given the same atmospheric/oceanic setup that lead to the 2025 heatwave, a mid-20th-century climate would have produced a far less severe event'.
However, this is not what the study has done. It has not explored the outcome of the *same* (or similar) atmospheric/oceanic set up - it has explored the differences in the predictors of analogous heat waves
3) This different attribution scope, though interesting, makes it more difficult to quantify the impact of climate change. In the traditional approach, the difference between, say, the mean of past analogs and the observed extreme would be the climate change signal. But how can one quantify, with the approach presented in this manuscript, the contribution of climate change? Actually, the authors do not provide a number, perhaps because it is indeed more difficult to obtain.
This should be discussed.
Minor points
4) 'we compared historical sea surface temperature (SST) patterns from ERA5 reanalysis (1950–2024) to those observed during the event.'
The SST from the ER5 reanalysis come from two different sources, as the manuscript clearly states, depending on the period. However, in neither of these two periods are the SSTs a 'reanalysis' product in the classical sense. The OSTIA SSTs are indeed the result of a data assimilation methodology, but the 'model' is very simple, just persistence of the previous time step. A reader may think that SSts are also the result of data assimilation of observations and a proper ocean model.
5) Figure 1. 'Time series of daily sea surface temperature (SST) anomalies averaged (standard deviation
shaded in grey)'
I cannot see the shaded area in the downloaded pdf file. Perhaps the rendering of the figure is not totally correct
6) Section 5 consists of just one big paragraph (!). I would recommend splitting the paragraph, perhaps by the variables being described, to help the reader.
7) 'STR (q–t) presents positive Present–Past differences, implying reduced net longwave loss in the present climate, consistent with a warmer, moister atmosphere returning more longwave radiation to the surface (Trenberth et al., 2015; Vautardet al., 2023)'
I think this paragraph and the corresponding figure warrant a more detailed discussion or description. In principle, attributing it to climate change would entail stronger long-wave forcing from greenhouse gases. Feedbacks increase specific humidity, thereby amplifying the greenhouse effect. But in theory, another scenario is possible: solar radiation increases, warming the ocean surface and increasing specific humidity. In this scenario, anthropogenic climate change would not be ultimately responsible, or only indirectly through reduced cloudiness. So there are subtle questions to discuss based on this figure.
A clean attribution would find that the same MHW is linked to a similar or smaller solar radiation and to increased longwave radiation. I am aware that this reduces the number of possible analogs, but , again, this seems to me to be a complication of this novel attribution approach.