the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Jan 2026
Variability in the Cretan Sea (Eastern Mediterranean) from six years’ glider observations (2017–2023)
Abstract. The Cretan Sea is an intermediate and occasionally deep water formation area within the Eastern Mediterranean that accumulates and transforms water masses from the adjacent Aegean, Levantine and Ionian Seas. Six years of glider observations (2017–2023) in the Cretan Sea were analysed to study the properties, variability and dynamics of the water masses during the study period. The analysis revealed progressive warming and salinification of the intermediate and deep layers. The mean temperature increased by around 0.05–0.07 °C per year, and the salinity by approximately 0.02 per year. Furthermore, comparisons with climatological data from 2000 to 2015 show temperature departures of + 0.4 to 0.6 °C in the upper 400 m and salinity increases of up to + 0.3 at the surface. Both of these values decline with depth, highlighting the intensified warming and increased salinity near the surface and in the upper intermediate layers. Additionally, the analysis of salinity and temperature datasets revealed the formation of intermediate water annually, except in winter 2022 when an intense mixing event occurred in the Cretan Sea triggered by exceptionally cold atmospheric conditions. The mixed layer, as captured by the glider, extended below 600 m inside the Cretan basin. These newly formed waters almost reached the deep layers, significantly modifying the properties of the intermediate and deep waters, although full deep convection was not reached. The observed downward displacement of the TMW core below 1000 m is associated with the strong convective event as well as with the redistribution of the heat and salt in the intermediate and deep layers. These findings emphasize the importance of sustained, high-resolution observations in capturing both gradual trends and extreme events, and in improving our understanding of the evolving thermohaline circulation of the Eastern Mediterranean.
- Preprint
(4602 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-6542', Anonymous Referee #1, 28 Jan 2026
-
AC1: 'Reply on RC1', Evi Bourma, 01 Apr 2026
We thank the reviewer for the thorough evaluation of our manuscript and for the constructive comments and suggestions, which have significantly contributed to improving the quality of our work. All points raised have been carefully addressed in the attached document.
-
AC1: 'Reply on RC1', Evi Bourma, 01 Apr 2026
-
RC2: 'Comment on egusphere-2025-6542', Anonymous Referee #2, 24 Mar 2026
General Comments:
This paper provides the community with a highly significant data set for the Cretan Sea, and useful context for understanding how the values/vertical structures vary over the period of study, and in relation to previous studies, both in the Cretan Sea and in the surrounding seas. The data seem of high quality and are provided for the reader. Overall conclusions about the heating anomaly in one year, the trends in temperature and salinity over the 6 years, and the changes in deeper layers below the intermediate water are supported by data and well-described.
While all of this is excellent, the 3 overall conclusions do not include a convincing discussion of the statistical significance or representativeness of the measurements. As in nearly all observational studies, there was undersampling in both space and time which makes discussion of the implications of the gaps critical. Neither is there a discussion of processes or mechanisms that may provide insight into the details of how the events/trends happened, and how it may help future studies better design sampling strategies to capture relevant mechanisms.
I accept that all of this is challenging, if not impossible, to perform and to include in a single paper. However these points need to be raised and discussed at the very least, and the authors should be careful not to claim quantitative conclusions without proper caveats and error analysis.
Specific Comments: (individual scientific questions/issues)
The main conclusion seems to be that the anomalous heating event was detected and in line with other studies. The magnitude of the event was described well and compared to other studies, but no further insights were provided (such as representativeness spatially, role of and impact on mesoscale dynamics and structure). A second conclusion about the rates of change of temperature and salinity is qualitatively convincing, but not quantitatively robust (statistically, representativeness spatially) nor did it provide any insight as to details of how/where this came about exactly (spatial extent, exact timing, or episodes not captured by measurements, and no discussion about those). The third conclusion about the role of intermediate water formation on deeper layers was not well-explained in terms of possible processes or convincing in terms of the measurements that might support.
No other tools or methods (like model simulations) were used to support the arguments for any of the above.
Technical Corrections and instances of specific comments:
TMW used in abstract without definition
Line 33: typo ‘Theoharis’
Line 48: it is not clear if EMT events occur in the Cretan Sea or outside, or both.
Line 60: thermohaline pump” mechanism ….are you implying this is in contrast to the BiOS? please explain.
Line 65: use of ‘where’ is confusing since just after that you say exported from. Clarify that exported from Cretan Sea.”…Aegean Sea, where the export of …”
Line 83: dynamics are not really discussed in this paper. The current fields in 3D are not presented or discussed. One 2-D plot mentioned, with illegible vectors, and inadequate discussion.
Line 101: Define “mission”; is it a single transect?
Line 105 and after: Details on glider specs, calibration of which sensors when, summary of QC methods and quantitative assessment of data quality (e.g. how much each flag was used, typical problems found and how dealt with)…lots missing here. Doesn’t TEOS-10 depend on regional composition, requiring local salinity samples to be used?
Line 119: what type of interpolation was used? Vertical bin averages used or other filters? Smoothing used for plots?
Line 130: A more detailed chart of time coverage would be interesting (time series of number of profiles per day for the duration, and again with all months collapsed to see coverage per month). Or a table of dates and length of each mission. (instead of Fig. 3)
Line 163: what signifies these ‘continuous exchanges’ exactly, from looking at the data?
Line 178: The T-S diagrams are not convincing with regard to spatial gradients described.
Line 199: “deep layers of salinity transects..” which depths, salinity values, and which transects specifically? Quantitatively.
Line 200-205: this section gives a 1-D description of a 2-d image...why? No thoughts on spatial structures for temp/density? No thoughts on time variability within the mission (10 days) or any biasing by internal waves or other non 1-D process? The dynamic topography maps could be used to explain the horizontal structure seen (location of eddies in both plots pointed out). As it is the figure is not legible, and the text does not help the reader make this connection, if there even is one.
Table 1: does this table show anything not also shown in time series in Figure 7 below? No need for table in my opinion.
Line 244: please explain how 'spreading from east to west' is to be inferred from the figure. What do you mean?
Line 244 (fig 8). at 26 deg from 29 to 30 March, huge changes seen in upper 600 m, and only a few more days for observed changes at 25.6 deg. Surely advection played a role.
Line 245: How can you be sure that the difference in density (fig. 9a) is vertical convection and not horizontal advection of the denser water while the glider was absent?
Line 259: Again, how can you be sure the difference in density, (fig 9b) is not horizontal advection of denser water into that area since the previous year? Any evidence from other sources?
Line 261 (and fig. 9): convective events not well-justified. Difficult to understand which contour is 'zero' change, and if there is really positive change of density at any particular depth (all some shade of yellow). Spatial structure and advection could also be major reason for difference between years, since cannot be sure of rest of the Cretan Sea, outside this transect.
Line 273: unclear how this calculation is done. Rewrite.
Line 276-286: this text fits better in the introduction.
Line 288: Below not ‘Bellow’
Line 294 and Fig 10: why standard deviation and not standard error to take into account of number of observations? (of course assuming a normal distribution is not ideal in either case but the reader at least should have an idea of the number of observations used to make the average).
Line 302: how well does the glider CTD sample the upper 2 m given the geometry of the sensor, the time response and reliability of the pump? Has there been any comparison with the CTD casts? With remote sensing? How many points given 30 sec sampling period could be expected in each? This kind of information about the sensor and general calibration/comparisons should be in the methods section and any comparisons relevant to trends made here.
Line 306: “delayed restratification” How is this shown? In figure? Or is it from a reference?
Line 308: “potential to disrupt patterns” can you be more specific? What did you have in mind exactly, and is it supported by something we can see in the figure? Or supports other work you can reference? Or just general comment?
Line 313 (fig 10): so is it true that there are 14 transects (each time point in the figure is one transect with its averages by layer) or 14 missions? 1 transect = 1 mission?
Line 314: you mean plots a, b, and c (not a, b, d)?
Line 324: why not examine by density and not depth ( evaluating the T and S changes over a range of densities representing the water mass)
Line 340 (fig 11): I don’t believe this figure adds much information. Replot with different axis of some of fig. 10 lines. Slopes of lines in table 2.
Line 388: what do you mean ‘selected to be correlated’? Why those years?
Line 400: Please explain. How can a change in annual mean, relative to climatology be 'consistent' with winter convection? You mean consistent with a particular change in winter convection? What is the process?
Line 413: how can the salinity and depth of a water mass change by 'exchanges'? Does this mean lateral mixing with Med Sea masses? And the 2nd mechanism is vertical mixing ('in the layers of Cretan Sea..')? This needs more explanation.
Line 415: Salt fingering: Has any evidence been found of this in the CTD profiles? Shouldn't it be visible during this period if indeed the water mass was changing for this reason?
Line 428: Mean core depth: How were the limits of the 'core' defined? Max/min depth? Max/min density? Values?
Fig 16 and 17: Why not show the 2 glider profiles nearest to these to avoid the smoothing by the averaging process?
Line 493: I am not clear on the mechanism for 'accelerating the downward shift' of TMW.
Citation: https://doi.org/10.5194/egusphere-2025-6542-RC2
Data sets
Dataset: Variability in the Cretan Sea (Eastern Mediterranean) from 6 years' glider observations (2017–2023) Evi Bourma https://doi.org/10.5281/zenodo.18076232
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 257 | 150 | 22 | 429 | 16 | 27 |
- HTML: 257
- PDF: 150
- XML: 22
- Total: 429
- BibTeX: 16
- EndNote: 27
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors present an analysis of a 6-year glider observational dataset in the Cretan Sea. The region is highly relevant to Eastern Mediterranean variability, and several changes described over the study period appear consistent with recent findings and with what is being observed more broadly in the area. However, the manuscript in its current form lacks a clear narrative and prioritization of objectives, and many conclusions are insufficiently supported by quantitative results and methodological detail. For these reasons, I recommend rejection in its current form, but encourage re-submission after substantial restructuring and additional analysis, which in my view go beyond a single major-revision cycle. I hope the authors will view these comments as guidance to strengthen the scientific robustness and clarity of the work.
Major comments
1. Clarify the main scope and re-balance the manuscript accordingly
At present, it is difficult to identify a single central research question (or a small set of clearly prioritized objectives). The abstract places strong emphasis on the winter 2022 convection/mixing event, but the manuscript does not consistently follow through with event-focused analysis beyond MLD and selected transects. If winter 2022 is intended as a main focus, it should be supported with additional diagnostics (see comment 5). If not, the abstract and discussion should be revised to better reflect the broader goals (seasonal/interannual variability, anomalies vs climatology, water-mass evolution, etc.).
2. Too many figures relative to the amount of quantified results
There are many figures (some could be merged), yet relatively few results are explicitly quantified in the text. In several places the figure captions/results text remain descriptive, while key numbers (magnitudes of changes, anomalies by layer, interannual differences, uncertainties) are not summarized clearly. I recommend reducing redundancy, merging related panels, and adding concise quantitative summaries (tables or compact metrics) to strengthen the Results section.
The inclusion of satellite ADT/geostrophic velocity is potentially valuable to interpret mesoscale control on the transects, but in the current manuscript it remains largely qualitative. Please either (i) reduce/relocate ADT panels (in a potential Appendix) if only illustrative, or (ii) leverage ADT quantitatively (eddy polarity/strength indices per mission; distance to eddy center; relation between ADT anomalies and isopycnal heave / T–S anomalies; simple eddy tracking) to demonstrate how mesoscale circulation explains the observed hydrographic variability.
3. Spatial definitions and data-selection methodology need to be explicit
Several analyses refer to “east/central/west” regions, specific profiles, or climatological comparisons, but the manuscript does not clearly define:
These definitions are critical for reproducibility and interpretation, and should be stated explicitly (ideally with a map and/or a table of coordinates/criteria).
4. Structure: Results vs Discussion are currently mixed
The Results section contains substantial interpretation and comparison that would fit better in the Discussion, while the Results themselves often lack quantitative reporting. I suggest a clearer separation:
5. Attribution and context for the 2022 strong convection/mixing event
If winter 2022 is a major highlight, it would benefit from stronger supporting evidence on the atmospheric forcing and buoyancy loss. For example, ERA5-based diagnostics over the event window (turbulent heat fluxes, wind stress, buoyancy flux components, evaporation–precipitation, air–sea temperature differences) could substantiate the proposed driver and help distinguish local vs basin-scale forcing.
6. Trend analysis: statistical methodology and significance need more detail
Given the short record (6 years), limited number of missions, and uneven seasonal sampling, I am cautious about strong statements regarding trends without detailed statistical justification. The manuscript should specify:
If possible, additional confirmation should be attempted via independent sources (e.g., regional in-situ data, Argo where available, and/or CMEMS reanalysis) over the same region and time span.
Finally, the E1M3A buoy time series is not described in enough methodological detail (sampling frequency, temporal coverage, QC/processing), and the manuscript does not explicitly discuss any discrepancies between buoy- and glider-derived trend estimates. Please add these details and a quantitative buoy–glider trend comparison over the overlapping period.
7. External references / climatology description should be more concrete
The manuscript refers to an external climatology and uses it as a reference. Please describe more clearly:
If additional historical observations are available (SeaDataNet/CMEMS/ship CTDs/Argo), they could be used to provide context beyond the 6-year record and to corroborate the reported changes.
8. Presentation and clarity
Several figures are difficult to read (small labels, low contrast, dark text on dark backgrounds). Increasing font sizes and improving contrast would substantially improve accessibility.
Consider merging figures and reducing the total number by combining related information into fewer, more information-dense plots.
Summary recommendation
The dataset and topic are valuable, but the manuscript requires substantial restructuring, clearer definitions and methods, stronger quantitative reporting in the Results, and more robust support for trend statements (including statistical significance and/or external validation). If the authors choose to resubmit a substantially reworked manuscript, the work could become a solid and publishable contribution to Eastern Mediterranean variability studies.