arxiv: 2604.22524 · v1 · submitted 2026-04-24 · ⚛️ physics.ao-ph

Recognition: unknown

Demographics of Mesoscale Eddies in an Eddy-Permitting Ocean Model and Reanalysis

Benjamin Lombardi , Ian Grooms , William Kleiber

Authors on Pith no claims yet

Pith reviewed 2026-05-08 09:06 UTC · model grok-4.3

classification ⚛️ physics.ao-ph

keywords mesoscale eddiesocean modelingsatellite altimetryreanalysiseddy trackingsea surface heightclimate projections

0 comments

The pith

Eddy-permitting ocean models and reanalysis data miss nearly 30 percent of the mesoscale eddies detected in satellite altimetry and produce eddies that live longer, grow larger, and remain weaker.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper applies the same detection and tracking method to sea-surface height fields from 1/4-degree satellite altimetry, reanalysis products, and ocean model output to count and characterize mesoscale eddies that persist longer than six weeks. It reports that the model and reanalysis fields contain only about 70 percent as many such eddy trajectories as the observations, and that the captured eddies differ systematically in duration, size, and intensity. Because these eddies drive much of the ocean's mixing, heat transport, and biogeochemical cycling, any systematic shortfall or distortion in their representation directly affects the reliability of climate projections and marine forecasts built on eddy-permitting simulations.

Core claim

When compared to eddies observed in satellite altimetry data, eddies in reanalysis data and ocean model output are missing almost 30% of the number of eddy trajectories. In addition to missing eddy trajectories, the characteristics of eddies in reanalysis data and ocean model output differ from eddies observed in satellite altimetry data. At a high level, eddies in reanalysis data and ocean model output tend to live longer, are larger, and are weaker than eddies in observed altimetry data. This paper presents a variety of statistics describing these differences both spatially and in global aggregate.

What carries the argument

Consistent application of an eddy detection and tracking algorithm to sea-surface height fields drawn from satellite altimetry, reanalysis, and eddy-permitting ocean model output at identical 1/4-degree resolution.

If this is right

Climate projections that rely on eddy-permitting ocean components will undercount the total number of mesoscale eddies active in the real ocean.
Longer-lived eddies in the simulations may cause models to overestimate the cumulative distance and time over which individual eddies influence tracer transport.
Larger yet weaker eddies alter estimates of horizontal mixing and vertical velocities that control heat uptake and nutrient distribution.
Spatial variations in the reported biases imply that regional ocean forecasts and marine heat-wave predictions carry geographically uneven errors.
Any hypothesis test involving eddy-driven biogeochemistry or air-sea exchange in these models rests on the unverified assumption that the resolved eddies are statistically realistic.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The shortfall could shrink in next-generation models that incorporate improved data assimilation or explicit sub-grid eddy parameterizations tuned against the same altimetry record.
Weaker modeled eddies may lead to underprediction of eddy-induced upwelling and its effects on surface productivity and carbon export in ecosystem models.
A direct test would compare eddy statistics from the same model run at both 1/4-degree and 1/12-degree resolution to isolate whether the bias is primarily a resolution or a representation issue.

Load-bearing premise

The detection and tracking algorithm yields unbiased, directly comparable eddy populations and properties when applied across observational, reanalysis, and model-generated fields despite differences in noise, resolution artifacts, and how the underlying height fields are produced.

What would settle it

Repeating the identical tracking procedure on an independent eddy-detection algorithm or on higher-resolution altimetry data that still shows the same 30 percent shortfall and the same shifts toward longer, larger, weaker eddies would confirm the result; if those differences vanish under the new method or data, the reported demographic mismatch would be falsified.

Figures

Figures reproduced from arXiv: 2604.22524 by Benjamin Lombardi, Ian Grooms, William Kleiber.

**Figure 1.** Figure 1: A cyclonic eddy trajectory with associated DEDs. Each DED is denoted with an ‘x’. The speed view at source ↗

**Figure 2.** Figure 2: Cyclonic Eddy Census. (D) is the zonal average cyclonic eddy census per year. (A-C) depict for view at source ↗

**Figure 3.** Figure 3: Cyclonic Eddy Amplitude. (A-C) display the spatial maps of mean amplitude for ADT, ORAS5, view at source ↗

**Figure 4.** Figure 4: Cyclonic Equivalent Radius. (A-C) display the spatial maps of mean equivalent radius for ADT, view at source ↗

**Figure 5.** Figure 5: Cyclonic Eddy Acircularity. (A-C) display the local mean maps of acircularity for ADT, ORAS5, view at source ↗

**Figure 6.** Figure 6: Cyclonic Eddy speed average. (A-C) display the spatial maps of mean speed average for ADT, view at source ↗

**Figure 7.** Figure 7: Cyclonic Eddy Births. (D) is the zonal average cyclonic eddy births per year. (A-C) depict the view at source ↗

**Figure 8.** Figure 8: Cyclonic Eddy Azimuth Mean. (A-C) display the circular mean for eddy cyclonic eddy azimuth view at source ↗

**Figure 9.** Figure 9: Anticyclonic Eddy Azimuth Mean. (A-C) display the circular mean for eddy anticyclonic eddy view at source ↗

**Figure 10.** Figure 10: Cyclonic Eddy Azimuth Variance. (A-C) display the circular variance for eddy anticyclonic eddy view at source ↗

**Figure 11.** Figure 11: Cyclonic Eddy Lifetime. (A-C) display the local mean maps of lifetime for ADT, ORAS5, and view at source ↗

**Figure 12.** Figure 12: Cyclonic Eddy Displacement. (A-C) display the spatial maps of mean displacement for ADT, view at source ↗

**Figure 13.** Figure 13: Cyclonic Ratio of Displacement to Lifetime. (A-C) display the spatial maps of the mean ratio of view at source ↗

**Figure 14.** Figure 14: Cyclonic Eddy Distance Traveled. (A-C) display the spatial maps of mean distance traveled for view at source ↗

**Figure 15.** Figure 15: Ratio of Distance to Lifetime. (A-C) display the spatial maps of the mean ratio of distance view at source ↗

**Figure 16.** Figure 16: Ratio of Displacement to Distance. (A-C) display the spatial maps of the mean ratio of dis view at source ↗

read the original abstract

Ocean mesoscale eddies can be thought of as the "weather" of the ocean and strongly influence the ocean's physics, chemistry, and biology; they influence other components of the Earth system via air-sea and sea-ice interactions, and are crucial drivers of marine heat waves. Thus, proper modeling of eddies in both historical and future climates is crucial to accurately capturing the Earth system. Climate projections using global coupled models with eddying ocean components are only recently starting to be more widely used. Despite their critical role in understanding and forecasting climate characteristics, these so-called eddy-permitting models have not been explored to verify that resolved eddies are realistic, and thus any downstream scientific testing of hypotheses in biogeochemistry, ocean physics or other associated Earth systems impacted by eddies hinge on this critical assumption. This paper compares observed eddies with lifetimes longer than 6 weeks present in $1/4^\circ$ satellite altimetry data with observed eddies in $1/4^\circ$ reanalysis data and ocean model output. When compared to eddies observed in satellite altimetry data, eddies in reanalysis data and ocean model output are missing almost 30% of the number of eddy trajectories. In addition to missing eddy trajectories, the characteristics of eddies in reanalysis data and ocean model output differ from eddies observed in satellite altimetry data. At a high level, eddies in reanalysis data and ocean model output tend to live longer, are larger, and are weaker than eddies in observed altimetry data. This paper presents a variety of statistics describing these differences both spatially and in global aggregate.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper reports that 1/4° models and reanalysis miss ~30% of altimetry eddy trajectories and produce longer-lived, larger, weaker eddies, but the fixed detection algorithm may not treat noisy observations and smooth fields equivalently.

read the letter

The headline result is that eddy-permitting ocean model output and reanalysis at 1/4° resolution recover only about 70% of the eddy trajectories found in satellite altimetry for features lasting longer than six weeks, with the captured eddies tending to live longer, reach larger radii, and show weaker amplitudes. The work supplies global totals plus spatial maps of these differences, which is the concrete new piece relative to earlier eddy-detection studies. Applying one consistent tracking pipeline across the three datasets is a straightforward way to expose the mismatch without introducing extra variables. The numbers are presented clearly enough that a reader can see both the aggregate shortfall and where the discrepancies concentrate geographically. That kind of descriptive benchmark is useful for groups running climate projections that rely on resolved eddies for heat transport or marine heat-wave statistics. The main limitation is that the abstract gives no evidence of tests that would confirm the algorithm returns unbiased counts and properties when the input fields differ in noise level and dynamical consistency. Altimetry contains measurement noise and interpolation effects that the model and reanalysis lack, so a single set of detection thresholds and lifetime cutoffs could systematically suppress short or weak tracks in the observations while retaining them in the smoother fields. Without synthetic-eddy injections or cross-filtering experiments, it is hard to separate physical model error from methodological artifact. The paper is aimed at ocean modelers and reanalysis developers who need to know how well their eddy populations match observations. It is not a methods paper and does not claim new theory, but the empirical comparison is direct enough to merit referee time. I would send it for review; the discrepancy is large enough that referees can decide whether the detection pipeline needs additional validation before the numbers are used to judge model fidelity.

Referee Report

2 major / 2 minor

Summary. The manuscript compares mesoscale eddies (lifetimes >6 weeks) detected via a single algorithm in 1/4° satellite altimetry against those in reanalysis and an eddy-permitting ocean model. It reports that reanalysis and model output contain ~30% fewer eddy trajectories than altimetry and that the detected eddies in the latter are longer-lived, larger, and weaker, with supporting spatial and global aggregate statistics.

Significance. If the detection pipeline is shown to be unbiased across data regimes, the quantitative demographics would provide valuable benchmarks for validating eddy-permitting models used in climate projections, directly informing studies of ocean transport, biogeochemistry, air-sea interactions, and marine heat waves.

major comments (2)

[Methods] Methods (eddy detection and tracking section): The paper applies one fixed algorithm and lifetime cutoff (>6 weeks) to noisy altimetry, smoother reanalysis, and model fields but provides no synthetic-eddy tests, noise-injection experiments, or cross-filtering controls to demonstrate that detection thresholds and tracking criteria return comparable statistics across these regimes. This directly affects the central ~30% trajectory deficit and the reported shifts in lifetime, size, and amplitude.
[Results] Results (global and spatial comparisons): The headline differences in eddy counts and properties are presented as physical but rest on the untested assumption that the algorithm's sensitivity to small-scale fluctuations (present in altimetry but absent in model/reanalysis) does not systematically suppress short-lived or weak features in the noisier dataset.

minor comments (2)

[Methods] Clarify the precise definition of 'eddy amplitude' and 'radius' used in the tracking algorithm and whether any post-processing filters differ across the three datasets.
[Figures] Figure captions should explicitly state the number of trajectories and the exact time period analyzed for each data source to allow direct reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their insightful comments on our manuscript comparing mesoscale eddy demographics across satellite altimetry, reanalysis, and ocean model data. We address each of the major comments below and indicate the revisions we plan to implement in the updated version of the paper.

read point-by-point responses

Referee: [Methods] Methods (eddy detection and tracking section): The paper applies one fixed algorithm and lifetime cutoff (>6 weeks) to noisy altimetry, smoother reanalysis, and model fields but provides no synthetic-eddy tests, noise-injection experiments, or cross-filtering controls to demonstrate that detection thresholds and tracking criteria return comparable statistics across these regimes. This directly affects the central ~30% trajectory deficit and the reported shifts in lifetime, size, and amplitude.

Authors: We acknowledge the validity of this concern. Our study utilizes a consistent eddy detection and tracking algorithm with identical parameters across all three datasets to facilitate a fair comparison. While we did not include synthetic tests in the original submission, we note that the algorithm has been validated in previous literature for both observational and modeled sea surface height fields. To directly address the referee's point, we will add a new subsection in the Methods discussing the potential impacts of data noise characteristics on detection sensitivity and will include results from a sensitivity test varying the amplitude and lifetime thresholds to demonstrate that the reported ~30% deficit and property shifts are robust to reasonable parameter choices. revision: yes
Referee: [Results] Results (global and spatial comparisons): The headline differences in eddy counts and properties are presented as physical but rest on the untested assumption that the algorithm's sensitivity to small-scale fluctuations (present in altimetry but absent in model/reanalysis) does not systematically suppress short-lived or weak features in the noisier dataset.

Authors: We agree that this assumption underlies our interpretation and that it should be more explicitly discussed. In the revised manuscript, we will modify the Results section to present the differences as those in the detected eddy populations rather than purely physical, and we will add text in the Discussion section acknowledging that some portion of the observed differences may arise from detection biases due to noise levels. This will provide a more balanced view while preserving the value of the quantitative comparison as a benchmark for model evaluation. revision: yes

Circularity Check

0 steps flagged

No circularity: direct empirical comparison of detected eddies

full rationale

The paper applies one fixed eddy detection and tracking algorithm to three independent data sources (1/4° altimetry, reanalysis, and model output) and reports counts and statistics of the resulting trajectories. No derivation chain, fitted parameters, or equations are present; the ~30% deficit and shifts in lifetime/size/amplitude are raw outputs of the same pipeline on different inputs. The algorithm itself is treated as an external tool rather than derived within the paper, and no self-citation supplies a uniqueness theorem or ansatz that the results reduce to. The comparison is therefore self-contained and externally falsifiable by re-running the detector on the source fields.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an empirical comparison study with no free parameters, mathematical axioms, or invented entities; it relies on standard oceanographic data analysis techniques applied to existing datasets.

pith-pipeline@v0.9.0 · 5600 in / 1294 out tokens · 58324 ms · 2026-05-08T09:06:10.596233+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

5 extracted references · 5 canonical work pages

[1]

doi:https://doi.org/10.1038/s41586-022-05162-6. C. Pegliasco, A. Delepoulle, E. Mason, R. Morrow, Y. Faug´ ere, and G. Dibarboure. Meta3.1exp: a new global mesoscale eddy trajectory atlas derived from altimetry.Earth System Science Data, 14(3):1087– 1107, 2022a. doi:https://doi.org/10.5194/essd-14-1087-2022. Qinbiao Ni, Xiaoming Zhai, Zhibin Yang, and Dak...

work page doi:10.1038/s41586-022-05162-6 2022
[2]

30 Reindert J Haarsma, Malcolm J Roberts, Pier Luigi Vidale, Catherine A Senior, Alessio Bellucci, Qing Bao, Ping Chang, Susanna Corti, Neven S Fuˇ ckar, Virginie Guemas, et al

doi:https://doi.org/10.1016/j.pocean.2008.10.013. 30 Reindert J Haarsma, Malcolm J Roberts, Pier Luigi Vidale, Catherine A Senior, Alessio Bellucci, Qing Bao, Ping Chang, Susanna Corti, Neven S Fuˇ ckar, Virginie Guemas, et al. High resolution model intercom- parison project (highresmip v1. 0) for cmip6.Geoscientific Model Development, 9(11):4185–4208, 20...

work page doi:10.1016/j.pocean.2008.10.013 2008
[3]

Issued by CLS under ser- vice contract 2022/C3S2 312b MOi/SC1

URLhttps://dast.copernicus-climate.eu/documents/satellite-sea-level/vDT2024/ C3S2-D312b-WP2-FDDP-SL-v3.0-202411-PQAR-of-vDT2024-v1.1_Final.pdf. Issued by CLS under ser- vice contract 2022/C3S2 312b MOi/SC1. Maxime Ballarotta, Cl´ ement Ubelmann, Marie-Isabelle Pujol, Guillaume Taburet, Florent Fournier, Jean- Fran¸ cois Legeais, Yannice Faug` ere, Antoine...

work page doi:10.5194/os- 2022
[4]

Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing,

doi:https://doi.org/10.1111/j.2517-6161.1995.tb02031.x. Daniel S. Wilks. “the stippling shows statistically significant grid points”: How research results are routinely overstated and overinterpreted, and what to do about it.Bulletin of the American Meteorological Society, 97(12):2263–2273, 2016. doi:https://doi.org/10.1175/BAMS-D-15-00267.1. N. I. Fisher...

work page doi:10.1111/j.2517-6161.1995.tb02031.x 1995
[5]

32 Elizabeth Yankovsky, Scott Bachman, K Shafer Smith, and Laure Zanna

doi:https://doi.org/10.1016/j.ocemod.2023.102265. 32 Elizabeth Yankovsky, Scott Bachman, K Shafer Smith, and Laure Zanna. Vertical structure and energetic constraints for a backscatter parameterization of ocean mesoscale eddies.J. Adv. Model. Earth Syst., 16 (7):e2023MS004093, 2024. doi:https://doi.org/10.1029/2023MS004093. Laure Zanna and Thomas Bolton. ...

work page doi:10.1016/j.ocemod.2023.102265 2023