Are mesoscale eddies enclosed deserts or productive oases? Satellite-derived evidence of chlorophyll depletion in eddy cores of the southeastern Mediterranean Sea

Gilboa, Merav; Lazar, Ayah; Berman-Frank, Ilana; Guy-Haim, Tamar

doi:10.5194/egusphere-2026-4744

Preprints

https://doi.org/10.5194/egusphere-2026-4744

Preprints

21 Aug 2026

| 21 Aug 2026

Are mesoscale eddies enclosed deserts or productive oases? Satellite-derived evidence of chlorophyll depletion in eddy cores of the southeastern Mediterranean Sea

Merav Gilboa, Ayah Lazar, Ilana Berman-Frank, and Tamar Guy-Haim

Abstract. Mesoscale eddies are fundamental drivers of physical and biological variability in the ultra-oligotrophic Southeastern Mediterranean Sea (SEMS). By integrating two long-term satellite altimetry datasets (DYNED and PET, 1993–2025), this study characterizes the seasonal spatiotemporal dynamics of cyclonic (CE) and anticyclonic (ACE) eddies and assesses their ecological impacts on phytoplankton biomass. Our analysis reveals polarity-based dynamics, though several physical metrics are sensitive to eddy-detection method: ACEs showed a tendency toward longer lifetimes than CEs, rotational velocity rankings reversed between datasets, and CEs consistently dominated high-density spatial hotspots. Seasonal patterns indicate peak eddy activity between February and April, followed by a basin-wide decline during the summer months. Our investigation of surface chlorophyll-a (Chl a) distributions challenges the classical "productive oasis" paradigms; most CEs demonstrate negative Chl a anomalies and depleted cores, potentially driven by intense top-down grazing pressure. Conversely, ACEs encompass larger areas of influence and may therefore facilitate greater horizontal retention, producing a larger area-integrated surface Chl a footprint and potentially supporting greater phytoplankton biomass through bottom-up nutrient accumulation. Our findings suggest that mesoscale features in the SEMS function as active ecological mechanisms, restructuring the basal food web via differential transport, retention, and community succession. This work emphasizes the need for coupled physical-biological frameworks to understand ecosystem resilience in nutrient-impoverished basins facing global climate change.

Received: 06 Aug 2026 – Discussion started: 21 Aug 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Merav Gilboa, Ayah Lazar, Ilana Berman-Frank, and Tamar Guy-Haim

Status: final response (author comments only)

RC1:
'Comment on egusphere-2026-4744', Cem Serimozu, 06 Oct 2026
Review: “Are mesoscale eddies enclosed deserts or productive oases? Satellite-derived evidence of chlorophyll depletion in eddy cores of the southeastern Mediterranean Sea”
This review was prepared by Cem Serimozu in collaboration with Frederike Benz in the context of the EGU peer review training. The manuscript investigates the physical characteristics and surface chlorophyll-a signatures of cyclonic and anticyclonic eddies in the southeastern Mediterranean Sea using two long-term eddy products, DYNED and PET, together with a Copernicus biogeochemical chlorophyll product. The manuscript is nicely structured and well written. The topic is scientifically relevant, particularly in the ultra-oligotrophic southeastern Mediterranean, where mesoscale variability may influence upper-ocean structure and biological distributions. The study's long temporal coverage and comparison between two eddy products are strengths. In particular, the manuscript illustrates that estimates of eddy radius, lifetime, velocity, and displacement can depend appreciably on the product and detection/tracking framework. The attempt to distinguish local Chl-a concentrations from broader spatial footprints is also potentially useful, and the public availability of the analysis code and datasets is a positive aspect of the study. However, several methodological definitions remain insufficiently clear for reproducibility, while some physical and ecological interpretations extend beyond what the diagnostics directly establish. In particular, the manuscript sometimes moves from geometrical or statistical relationships to statements about material retention, fluid isolation, transport, and causal ecological mechanisms that are not directly quantified.
We therefore recommend major revision. Our comments focus primarily on the manuscript's physical-oceanographic and methodological aspects.
1) Definition and reproducibility of the analyzed eddy and Chl-a populations
The manuscript needs a clearer, internally consistent description of the eddy and chlorophyll populations used in each analysis.
The direct comparison between DYNED and PET combines different temporal sampling periods. DYNED covers approximately 2000-2017, whereas PET Historical covers 1993-2021. Yet lifetime, radius, displacement, and rotational velocity are compared directly, despite appreciable interannual variability in eddy occurrence. Temporal sampling could therefore contribute to the reported differences. We recommend repeating the principal DYNED-PET physical-property comparison over their common temporal interval using the same study-domain selection and post-processing inclusion/exclusion criteria. If the differences persist, this would strengthen the conclusion that they are robust to temporal sampling.
DYNED and PET should also be treated as distinct eddy products rather than as a controlled comparison of two detection/tracking algorithms applied to identical inputs. Although both are ultimately based on gridded satellite-altimetry products, their input fields and processing chains differ. The Mediterranean DYNED Atlas applies AMEDA to daily surface geostrophic velocity fields derived from regional 1/8° ADT maps produced by SSALTO/DUACS and distributed through CMEMS (Le Vu et al., 2018; Barboni et al., 2021). A cyclostrophic correction is applied to these geostrophic velocities when quantifying the eddy dynamical properties (Ioannou et al., 2019; Barboni et al., 2021), which can appreciably modify the estimated intensity of strong anticyclones. In contrast, the META3/PET framework underlying META3.2 is based on global daily 1/4° ADT fields, applies spatial high-pass filtering before contour-based PET detection, and defines its speed contour from the contour of maximum circum-averaged speed (Pegliasco et al., 2022). Differences in eddy lifetime, radius, displacement, and particularly rotational velocity therefore cannot be attributed uniquely to the detection/tracking algorithm. We suggest describing these results as sensitivity to the respective eddy products and their associated processing and detection/tracking frameworks.
The statement in Sect. 2.1 that both datasets provide trajectories at a spatial resolution of 0.25° should also be checked or clarified. Published DYNED methodology describes the Mediterranean Atlas as being derived from regional 1/8° ADT-based velocity fields (Barboni et al., 2021), whereas META3/PET is based on global 1/4° fields (Pegliasco et al., 2022). If 0.25° refers to subsequent resampling or output representation rather than the native field used for DYNED detection, this distinction should be stated explicitly.
The spatial selection also requires clarification. The Methods define the study region as 31-34.5°N and 32-36°E, whereas the trajectory and AOI figures extend considerably west of 32°E. Please specify whether eddies were selected when entering or intersecting the nominal region and their complete trajectories retained, or whether a larger analysis region was actually used. It should also be clear whether reported lifetime and displacement represent complete atlas trajectories or only their in-domain segments. This affects interpretation of both the physical-property statistics and statements concerning eddy source regions.
The Chl-a sampling procedure also appears internally inconsistent. Section 2.3 states that mean and integrated surface Chl-a were calculated for each DYNED eddy at the time of maximum azimuthal velocity. Section 2.5, however, describes cross-correlation and frequency-domain coherence between rotational velocity and Chl-a along continuous individual eddy tracks. These descriptions imply different sampling designs. Similarly, 'life stage' enters the statistical analysis without a sufficiently clear definition. Please state explicitly whether Chl-a is sampled once per eddy, repeatedly throughout each trajectory, or differently for separate analyses, and associate each procedure with the relevant figure and statistical test.
An additional clarification concerns the provenance of the Chl-a field. Section 2.3 identifies MEDSEA_MULTIYEAR_BGC_006_008 as the chlorophyll source. This is a Mediterranean Sea biogeochemical reanalysis produced with the MedBFM3 model system, in which satellite surface chlorophyll observations are assimilated, rather than a direct satellite ocean-color retrieval product. According to the Copernicus Marine product documentation, the reanalysis has a horizontal resolution of 1/24° and assimilates ESA-CCI surface chlorophyll on a weekly cycle. Please distinguish clearly between reanalysis/model-derived Chl-a and direct satellite observations throughout the manuscript, and reconsider or qualify the term "satellite-derived" where appropriate. The product resolution and data-assimilation procedure should also be stated explicitly. This distinction is particularly relevant to Sect. 4.4, where limitations of direct satellite optical retrievals are discussed in relation to the analyzed Chl-a field.
The spatial Chl-a zones need clearer, internally consistent definitions. Section 2.4 defines three spatial categories: an internal core, a combined "effective or speed contour" margin, and an external 1.5R-2.5R background annulus. In contrast, Fig. 5 treats the effective margin and speed margin as separate zones, resulting in four spatial categories. Please reconcile these definitions and specify exactly how each zone is constructed. For the 1.5R-2.5R background annulus, please also state how land or invalid Chl-a pixels and overlap with other contemporaneous eddies are treated.
2) Physical interpretation of the AOI and area-integrated metrics
Several central metrics are mathematically well defined, but the manuscript assigns them physical interpretations that are not directly supported by how the metrics are calculated. The Area of Influence (AOI) is generated by buffering eddy positions by the eddy radius and subsequently calculating a convex hull around the lifetime geometry. A convex hull can include regions that were never contained within an instantaneous eddy footprint, particularly for curved or looping trajectories. The resulting quantity is therefore better interpreted as a lifetime trajectory envelope than as the area actually occupied or influenced by the eddy. A useful and relatively simple sensitivity check would be to compare the convex-hull area with the union of the instantaneous radius-buffered footprints before applying the hull. This would show whether the convex hull is a reasonable approximation or whether the AOI terminology and interpretation should be revised.
The lifetime hull also contains no information about how long an eddy occupies each location, so it should not be interpreted as a measure of temporal occupancy. If persistent occupancy is an intended result, a time-resolved measure such as the fraction of observation days for which each cell lies within an eddy footprint would be more directly interpretable.
A similar issue applies to the integrated Chl-a metric. The manuscript defines this as mean near-surface Chl-a multiplied by eddy area, with units of mg m^-1. This represents a horizontal area integral of a volumetric concentration field. It is not equivalent to total chlorophyll mass, water-column-integrated biomass, or primary production. Because the integrated metric explicitly includes eddy area, and the area of a radius-defined circular footprint scales as R², part of the relationship between integrated Chl-a and radius is built into the metric by construction. This geometric dependence is quadratic, a power law with exponent 2, rather than exponential. A strong relationship with radius therefore cannot by itself demonstrate enhanced accumulation, carrying capacity, or retention. We suggest clearly distinguishing between mean Chl-a concentration, eddy area, and the combined area-integrated surface signal. If correlations between the integrated metric and polarity or rotational velocity are interpreted mechanistically, the authors should establish that the relationship is not simply a consequence of eddy size, or else moderate the interpretation.
3) Statistical independence, effect magnitude and temporal inference
The manuscript applies a substantial statistical framework, but the underlying sampling structure is not sufficiently clear. The most important issue is how you define the independent observational unit. Measurements from different spatial zones of the same eddy are paired, while track-based analyses may contain repeated observations over the lifetime of the same eddy. Such observations are not necessarily statistically independent. Please state the sample size and the independent unit of replication for each principal statistical analysis, and explain how repeated measurements from the same eddy are accounted for.
This is particularly important for the reported three-way ANOVA. Polarity, season, and life stage are described as 'highly significant drivers,' with p<0.001, yet the reported effect sizes are extremely small: η²=0.00041, 0, and 0.0001, respectively. Statistical significance should therefore be distinguished from effect magnitude and does not by itself imply scientific importance. Please specify the definition of η², report sample sizes and degrees of freedom, and discuss whether effects of this magnitude are scientifically meaningful. The manuscript also refers to interactions between polarity, season, and life stage, but it should report the corresponding interaction statistics explicitly if these interactions are important to the interpretation.
The authors also conduct multiple month-specific CE–ACE comparisons under a stated α=0.05 threshold. Please clarify whether any correction for multiple testing was applied. Given the number of comparisons, we suggest reporting the exact p-values without describing values such as p=0.055 and p=0.064 as “approaching statistical significance,” as this wording may overstate the evidence.
The spectral analysis also requires sufficient methodological detail for its robustness to be evaluated. The manuscript identifies characteristic periods of approximately 3.6 years for ACEs and 9 years for CEs, but does not sufficiently describe preprocessing, detrending, windowing, spectral resolution, or significance assessment. Only a small number of complete 9-year cycles are present within the available record. The physical interpretation of this spectral peak should therefore remain cautious unless its robustness can be demonstrated.
Similarly, for the cross-correlation and coherence analysis, please clarify how you treat tracks of different lengths and whether the study accounts for temporal autocorrelation and common seasonality. A zero-lag relationship shows contemporaneous covariance but does not, by itself, establish a rapid causal Chl-a response.
4) Distinguishing tracked eddy features from material retention, transport and causal mechanisms
This is the main physical-interpretation concern. The analyses characterize detectable eddy features through trajectory, lifetime, radius, and rotational velocity. Several sections, however, interpret these feature-tracking diagnostics as direct evidence of material fluid retention. For example, Sect. 4.3 describes the relationship between rotational velocity and integrated footprint as 'direct physical evidence of fluid isolation,' and states that high peripheral velocities create a closed boundary that traps internal water and prevents cross-frontal dispersion. A persistent eddy identified from sea-surface height or velocity structure is not necessarily a materially coherent body of water, because exchange can occur across an evolving boundary while the feature remains identifiable. This distinction is consistent with Condie and Condie (2016), already cited by the manuscript, who estimate plankton retention within eddies using Lagrangian particle tracking and report retention times of 5-67 days (median 19 days), which they note are short relative to an estimated average eddy lifetime of approximately eight months. Chelton et al. (2011), also cited by the manuscript, characterize eddy nonlinearity using the ratio of rotational velocity to translation speed (U/c), rather than rotational velocity alone. More generally, Eulerian feature lifetime and Lagrangian/material coherence can occur on substantially different timescales, with leakage and entrainment occurring while an eddy remains identifiable (Denes et al., 2022). Consequently, tracked eddy lifetime and trajectory cannot automatically be equated with water-parcel residence time or coherent water-mass transport, and maximum rotational velocity alone does not quantify boundary permeability, entrainment, or retention.
The manuscript later states that future Lagrangian analyses are required to examine transport, entrainment, and internal retention/trapping, which is inconsistent with the earlier claim that the present diagnostics already provide direct evidence of fluid isolation. We do not consider a full Lagrangian experiment necessary for revision if it lies outside the intended scope of the study. The authors could instead moderate the interpretation, describing large, persistent and rapidly rotating eddies as having characteristics potentially favorable to enhanced retention rather than as direct evidence of material trapping. The same distinction applies to terms such as 'transport pathways' or 'high-volume eddy corridors.' Eddy tracks can demonstrate recurrent pathways of eddy-feature propagation or regions of high track density, but they do not by themselves quantify water-volume transport; this would require an appropriate volumetric or material-transport diagnostic.
The ecological mechanisms should also be distinguished from the observed Chl-a patterns. The proposed grazing and nutrient-related explanations may be plausible and supported by previous studies, but grazing pressure and nutrient supply are not directly quantified here. These mechanisms should therefore be presented as interpretations consistent with the observations rather than processes demonstrated by the present analysis.
5) Explain the novelty compared to previous studies
The novelty of the study could be articulated more explicitly, particularly in relation to previous work, to help clarify the specific contribution of the present manuscript.
Minor and specific comments
Seasonal intervals in Fig. 7. The periods February-April, May-July, August-November and December-January are labelled MAM, JJA, SON, and DJF, although they do not correspond to those meteorological seasons. They also have unequal durations. If trajectory or hotspot counts are cumulative, normalize the intervals by duration or label them explicitly as the custom periods used.

'Offshore migration.' Seasonal changes in the spatial distribution of tracks do not necessarily demonstrate that the same eddies migrate offshore. Unless this is shown from individual trajectories, 'the spatial distribution shifts offshore' would be more precise.

Mean Dynamic Topography. MDT minima and maxima should not be described simply as 'depressions' and 'ridges' if this terminology is intended to imply seafloor topography. If dynamic-topography structure is intended, the terminology should be made explicit. If bathymetric control is invoked, bathymetry should be considered separately. The source and resolution of the MDT field should also be specified.

Persistence terminology. DYNED ACEs have a longer mean lifetime than CEs, but the reported difference is p=0.053, while the PET lifetime difference is also non-significant. Statements that ACEs are categorically 'more persistent' should therefore be qualified, for example as a tendency toward longer mean lifetimes.

Month-specific CE-ACE differences. The manuscript is internally inconsistent regarding the months in which significant differences occur: one section reports June, July, October and December, whereas another reports only July and December. Please reconcile these statements.

Fig. 7 methodology. The figure introduces clustering, significant hotspots, trajectory weighting, and origin/terminal markers that need sufficient methodological definition to be reproducible. In particular, clarify whether origin/terminal points represent true eddy genesis/dissipation or simply the first and last positions within a seasonal subset.

Reference-list consistency. Please cross-check the in-text citations against the bibliography. Le Vu et al. (2018), cited in Table 1 as a methodological reference for AMEDA, and Mason et al. (2014), cited in Sect. 2.1 for py-eddy-tracker, appear to be missing from the reference list. We also noted two bibliographic errors: the DOI for Belkin et al. (2022) is listed as 10.5194/os-18-693-20 but should be 10.5194/os-18-693-2022; and the Zhao et al. entry for ‘Global chlorophyll distribution induced by mesoscale eddies’ is incomplete/incorrect. The paper was published in 2021 in Remote Sensing of Environment, 254, 112245 (https://doi.org/10.1016/j.rse.2020.112245), rather than the currently listed ‘Not Available, 2022.’ Please verify and correct these entries.

In Fig. 2, the legend includes 'DYNED Atlas', 'PET NRT', and 'PET Historical', but these distinct lines do not appear clearly distinguishable within the figure itself, making the graphic representation confusing for the reader.

We highly recommend adding a general geographical overview map at the very beginning of the manuscript to help orient the reader to the study domain within the southeastern Mediterranean Sea.

References supporting the methodological comments
Le Vu, B., Stegner, A., and Arsouze, T. (2018). Angular Momentum Eddy Detection and Tracking Algorithm (AMEDA) and Its Application to Coastal Eddy Formation. Journal of Atmospheric and Oceanic Technology, 35(4), 739-762. https://doi.org/10.1175/JTECH-D-17-0010.1

Barboni, A., Lazar, A., Stegner, A., and Moschos, E. (2021). Lagrangian eddy tracking reveals the Eratosthenes anticyclonic attractor in the eastern Levantine Basin. Ocean Science, 17, 1231-1250. https://doi.org/10.5194/os-17-1231-2021

Ioannou, A., Stegner, A., Tuel, A., Le Vu, B., Dumas, F., and Speich, S. (2019). Cyclostrophic Corrections of AVISO/DUACS Surface Velocities and Its Application to Mesoscale Eddies in the Mediterranean Sea. Journal of Geophysical Research: Oceans, 124(12), 8913-8932. https://doi.org/10.1029/2019JC015031

Pegliasco, C., Delepoulle, A., Mason, E., Morrow, R., Faugère, Y., and Dibarboure, G. (2022). META3.1exp: a new global mesoscale eddy trajectory atlas derived from altimetry. Earth System Science Data, 14, 1087-1107. https://doi.org/10.5194/essd-14-1087-2022

Condie, S., and Condie, R. (2016). Retention of plankton within ocean eddies. Global Ecology and Biogeography, 25, 1264-1277. https://doi.org/10.1111/geb.12485

Chelton, D. B., Gaube, P., Schlax, M. G., Early, J. J., and Samelson, R. M. (2011). The influence of nonlinear mesoscale eddies on near-surface oceanic chlorophyll. Science, 334(6054), 328-332. https://doi.org/10.1126/science.1208897

Denes, M. C., Froyland, G., and Keating, S. R. (2022). Persistence and material coherence of a mesoscale ocean eddy. Physical Review Fluids, 7, 034501. https://doi.org/10.1103/PhysRevFluids.7.034501

Mason, E., Pascual, A., & McWilliams, J. C. (2014). A new sea surface height–based code for oceanic mesoscale eddy tracking. Journal of atmospheric and oceanic technology, 31(5), 1181-1188. https://doi.org/10.1175/JTECH-D-14-00019.1
Citation: https://doi.org/10.5194/egusphere-2026-4744-RC1
RC2:
'Comment on egusphere-2026-4744', Anonymous Referee #2, 12 Oct 2026
Review: Are mesoscale eddies enclosed deserts or productive oases? Satellite-derived evidence of chlorophyll depletion in eddy cores of the southeastern Mediterranean Sea
Major revision is recommended. The title and abstract suggest that the authors will show if eddies are “deserts or productive oases” driven by ecological processes such as grazing and retention. Biological mechanisms are not directly addressed. Satellite-derived datasets are used to characterise the physical structure of eddies at the surface. The vertical eddy structure is not explored. Without including analysis of the biological and vertical processes, the conclusions of this manuscript cannot be reached.
This study characterises cyclonic (CEs) versus anticyclonic eddies (ACEs) in the Southeastern Mediterranean Sea from eddy atlases calculated from satellite-based altimetry (DYNED and PET). By applying eddy metrics to a daily surface chlorophyll dataset (Teruzzi et al., 2021), the study addresses the possible role of CEs versus ACEs in primary productivity in this ultra-oligotrophic region. While they show that CEs are associated with “negative chlorophyll anomalies” and ACEs are larger (but only for one of the datasets), the conclusion that these correspond to depletion in CEs and greater “carrying capacity” in ACEs, respectively, is not supported by the analysis.
While characterising the role of eddies in the region and their effect on productivity in such an oligotrophic region is a valuable contribution, the connection between the chosen surface eddy metrics and the influence on surface chlorophyll concentration is not clear. The chlorophyll dataset reference is titled “Mediterranean Sea Biogeochemical Reanalysis (CMEMS MED-Biogeochemistry, MedBFM3 system) (Version 1)” suggesting that the missing vertical analysis could be done.
General Comments
The study does not show if the eddies from the chosen atlas are consistent with the chlorophyll dataset. Chlorophyll concentrations within and surrounding an eddy are analysed with boundaries defined by the eddy dataset e.g. the core uses the maximum speed contour. This has consequences to the spatially-averaged chlorophyll values defined by boundaries from a separate dataset. Without addressing this, eddy chlorophyll values cannot be attributed to eddy trapping or due to eddy evolution.

There is no discussion on how robust the methodology is to the high chlorophyll concentration at the coast that exists throughout the year (Fig. 3). For example, if the eddy is located close to this region, spatial averaging for the eddy external domain could incorporate the high values; the eddy now appears depleted relative to the higher background. How are eddies close to the coast handled?

The use and limitations of the chosen datasets need addressing, the choices could bias the analysis. The larger footprint of ACEs compared to CEs leads to much discussion and conclusions. However, this stems from the use of the DYNED eddy atlas together with surface chlorophyll. When the DYNED and PET datasets were compared, the size comparison is reversed. DYNED eddy mean radii for ACE and CE are 27 km and 23 km, while for PET Historical are 38 km and 41 km.
The value of using multiple datasets is not clear, especially when they contradict each other, “ACEs showed a tendency toward longer lifetimes than CEs, rotational velocity rankings reversed between datasets” (lines 18-19).

The authors state that two datasets are used but describe and show the results of three datasets (DYNED, PET Historical, PET NRT). They acknowledge that the eddy detection algorithm affects the eddy metrics and directly compare the resultant eddy characteristics but they do not state which dataset is more applicable to their case.

the time period over which the means are calculated and compared are not stated, this could also result in differences between the products in addition to the mentioned detection method.

The Rossby Radius for this region is not defined nor the limitations of the resolution of the datasets addressed.

The approach of using surface data while delineating vertical versus horizontal effects of the eddy is not addressed, e.g. the role of the deep chlorophyll maximum.

Chlorophyll concentration is often used as a proxy for phytoplankton biomass, results should be limited to this and not extended to incorporate "planktonic communities” and ecological processes (e.g. grazing, retention).

I found Section 3.4 difficult to follow. The reasoning that starts with where eddies occur and results in “transport pathways”, combined with confusing figures (Fig. 6 and 7) needs re-evaluation.

Many statistical tests were performed but the motivation and results for these tests are hidden in the text and should be presented in figures or in a table.

The chosen figures/presentation do not adequately address the problem statement.
The reader could benefit from an orientation to the region and its dynamics, a snapshot of chlorophyll with eddy contours (speed, R) overlaid.

Figures 6 and 7 are difficult to interpret. Besides the hotspot for ACE occurrence in Figure 6, a clear pattern is not readily distinguishable. Perhaps Figure 7 can be summarised into a clearly annotated schematic.

Where there are a lot of values being compared in the text, consider tabulating the results.

Where possible, a clearly annotated schematic would help.

Line plots need gridlines

X-labels for Months should be named or initialed to be consistent with the text

Some figure captions use new terminology not already introduced in the main text

Some inconsistencies need to be fixed in the text to prevent reader confusion:
“divergence“ is a physical oceanography metric, so use difference/contrast instead

“occurrence” seems to refer to different quantities

chlorophyll and “pigment levels” are used interchangeably which makes one question if the units have changed

“integrated chlorophyll” in this paper refers to an area integral rather than the more common depth-integrated chlorophyll

ring definitions should be clearly indicated in Figure A1 with annotations and R shown on the eddies

chlorophyll anomaly when only chlorophyll means are shown

“topography” or ”topographic features” is more commonly used for bathymetry such as the Eratosthenes Seamount, avoid when referring to dynamic topography. MDT reflects the ocean circulation.

Using month names in the text but the x-labels are for numerical months (Month 1, 2, etc as in Fig 2b)

Using seasons as categories, e.g. “winter” in Fig. 3 but months in Fig. 7 “a: Dec-Jan (DJF)”

jargon e.g. line 321 “Chl anomaly variability is modulated by synergistic multi-variable interactions”

Specific Comments

22: The "productive oasis" paradigm should be explicitly stated so the reader can confirm that your results are contrary to this.
26-28: Data insufficient to state this
51: introducing bathymetric features while discussing eddy detection is out of place
62-66: What data did they use? Can their results be applied to all eddies in the region?
96: the chlorophyll analysis used only the DYNED dataset
127: Table 1- define LNAM
139-140: “density”, do you mean number? What is the effect of counting over different periods?
172: This is 2 separate boundaries, this sentence is ambiguous
202: Putting the comparison in a table would help the reader
212: “The distinct structural… products” what does this refer to? The sentences before or after?
237-238, Figure 2a: the intensified period is difficult to see in the plot. Noticeable is after 2015 in the PET Historical CEs the amplitude of the occurrences decreases.
277-281, 330-332: Similar paragraphs with different months
Fig. 3: Use a logarithmic colormap, ”mean monthly surface Chl a” should be mean seasonal
345-348: the area covered by type of eddies, if needed, should be visualised in a bar chart or table so that their connection to the argument is more obvious
354: “persistent topographic features”, clarify if this refers to bathymetry then the snapshot of the region with bathymetry included would help.
362: “maximum track line thickness” and how the zones connect is not apparent
371: eddy formation has not been addressed previously, the propagation/transport pathways connection is not obvious
411: The generation of new eddies “source” has not been presented
Citation: https://doi.org/10.5194/egusphere-2026-4744-RC2

Merav Gilboa, Ayah Lazar, Ilana Berman-Frank, and Tamar Guy-Haim

Viewed

Total article views: 273 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
125	74	74	273	81	64

HTML: 125
PDF: 74
XML: 74
Total: 273
BibTeX: 81
EndNote: 64

Views and downloads (calculated since 21 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	26	19	13	58
Sep 2026	49	35	49	133
Oct 2026	50	20	12	82

Cumulative views and downloads (calculated since 21 Aug 2026)

Month	HTML	PDF	XML	Total
Aug 2026	26	19	13	58
Sep 2026	49	35	49	133
Oct 2026	50	20	12	82

Viewed (geographical distribution)

Total article views: 267 (including HTML, PDF, and XML) Thereof 267 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 22 Oct 2026

Short summary

This study examined how eddies or rotating currents, shape surface chlorophyll in the nutrient-poor southeastern Mediterranean Sea. Satellite records from 1993–2025 showed that most counterclockwise eddies had chlorophyll-depleted centers, contrary to the idea that they usually boost phytoplankton growth. Clockwise eddies covered larger areas and may retain more organisms and material. Eddy effects varied with direction, season, and age rather than following a simple desert-or-oasis pattern.


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