- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper introduces a structured predictor for global surface geostrophic currents. It separates velocity into solenoidal, irrotational, and uniform components, predicts vorticity corrections followed by spectral Biot-Savart inversion, applies damped persistence plus a learned correction to the irrotational component, and persists the uniform flow. Models trained for one-day prediction are rolled out for 20 days on AVISO/DUACS data, with evaluation of velocity and eddy kinetic energy (EKE). The central contribution is a task-specific combination of physical representation, residual learning, and component-dependent forecasting priors.
Critical strengths
-
The application contribution is meaningful.** Predicting global geostrophic currents on a 680 x 1440 grid, while explicitly examining EKE over 20-day rollouts, presents a materially different setting from small synthetic fluid benchmarks or regional scalar-field prediction. HelmFluid already uses Helmholtz-inspired dynamics, but its reported sea-temperature experiment uses four 64 x 64 regions. HelmGeoNet's global current-vector task and energy-focused formulation deserve independent credit; they are not rendered unoriginal merely because the decomposition is classical. See HelmFluid.
-
The methodological contribution is a coherent domain-specific adaptation.** Vorticity-residual prediction, an explicit inversion, unequal correction capacity across components, and a persistence-based initialization form a sensible architecture around the properties of this dataset. These are established ingredients, so I view algorithmic originality as moderate rather than foundational. Nevertheless, their arrangement addresses a concrete forecasting difficulty and is more substantive than applying an unchanged generic backbone.
-
The reported gains are measurable and extend beyond comparison with generic models.** Day-20 EKE RMSE decreases from 0.0411 for TAU to 0.0382, a 7.1% reduction, and from 0.0437 for persistence, a 12.6% reduction. EKE ACC improves from 0.3975 to 0.4194. Importantly, the physics-only variant at rho = 0.97 has RMSE 0.0435, so the full model improves over its own non-learned scaffold by 12.2%. This supports a useful contribution from learned corrections, even though it does not identify the contribution of every architectural choice.
-
The statistical motivation is supported, with an appropriate practical interpretation.** Recomputing Table 1 gives solenoidal persistence-error shares of approximately 90.6%, 93.6%, 94.8%, and 95.3% at 1, 5, 10, and 20 days. Ignoring the reported near-zero uniform term, the remaining accounting discrepancy is approximately 0.4-0.9% of total error. These values support approximate separation and allocating most capacity to the solenoidal branch. They do not prove independence or exact orthogonality, but they provide useful empirical motivation.
Critical weaknesses
-
The physical interpretation is stronger than the numerical formulation establishes.** Equations (2)-(4) describe planar spectral operators, whereas the application uses a masked latitude-longitude grid and a cos(latitude)-weighted loss. The treatment of spherical metric factors, latitude boundaries, and land filling is not sufficiently explicit. Orthogonality of a planar FFT projection does not automatically survive the change of inner product or masking. I verified this distinction with a periodic analytic example: an unweighted cross inner product numerically equal to zero becomes 1/3 under cosine weighting. This is a counterexample to an automatic guarantee, not evidence that the submitted implementation produces that error. The small discrepancy in Table 1 suggests the approximation may work well in practice. The paper should distinguish that empirical success from exact physical claims.
-
The sources of improvement are only partly identified.** The physics-only comparison establishes that learning helps, but does not show how much the decomposition itself improves over a similarly initialized TAU-style velocity-residual predictor. Likewise, the comparison with rho = 0.8 and 1.1 omits rho = 1, the most natural no-damping control. At day 20 these settings have RMSEs 13.3 and 1.8 times the default, respectively, showing substantial sensitivity. A small set of targeted controls would clarify whether the benefit comes primarily from representation, initialization, damping, or their interaction. A direct HelmFluid comparison would be informative, but is not the sole basis for judging originality because its original task and implementation differ.
-
The energy-stability diagnostics need a narrower interpretation.** For a hypothetical constant normalized energy error e(t) = 0.5, both reported variation measures are zero despite a persistent 50% error. Thus, these metrics measure variation across lead time, not energy fidelity by themselves. The EKE RMSE results still provide meaningful evidence of accuracy. However, the use of velocity clipping during rollout makes the clipping rate and its influence relevant to the stability interpretation; clipping is applied across models, so its presence alone does not establish an unfair comparison.
-
Some technical statements and selection details require correction or clarification.** Small cross terms do not imply statistical independence; Eq. (8) is optimal within scalar-multiple predictors, not arbitrary linear predictors. Appendix A's Eq. (14) also appears inconsistent with the earlier definition of E_irr as persistence error. Finally, the text should reconcile estimating rho from historical statistics with the test-set comparison described in Appendix C. I would not infer test leakage from the wording alone. These issues limit the theoretical explanation and reproducibility, but do not by themselves invalidate the reported forecasting improvements.
Questions for authors
- Which spherical metric factors, spatial units, boundary conditions, and land-filling procedure are used? Can you report divergence/curl and weighted orthogonality residuals for the actual masked implementation?
- How does the model compare with a matched velocity-residual predictor using the same backbone, initialization, seasonal inputs, and training budget? What happens for rho = 1 or a learned rho?
- What fraction of predicted ocean points is clipped at each lead, and how do unclipped EKE errors and failure rates compare? Can mean energy bias accompany the variation diagnostics?
- Were rho, climatology, and all other design choices fixed without consulting test outcomes? Can repeated runs or a temporally appropriate uncertainty estimate establish the reliability of the improvement over TAU?
6 comments
Response to Reviewer hPzN [2/4]
[Response to Q2]: We answer the two parts in turn.
Matched velocity-residual predictor. In the revised manuscript we submitted, on Page 17, Line 913, we state that "In Table S4, we isolate the role of the Helmholtz structure with a residual control. The control keeps the backbone and the training configuration of HelmGeoNet, removes the Helmholtz decomposition and the relaxation mechanism, and predicts the velocity residual around persistence with the TAU modules alone." The control shares the three-level encoder-translator-decoder with TAU translator blocks, the zero-initialized final projection, the seasonal channels, and the parameter budget. Table 1 lists the day-20 results.
Table 1. Day-20 EKE anomaly correlation (ACC) and latitude-weighted RMSE on the test set spanning 2023 to 2024. The best value per column is bolded. (Table S4 in the revised manuscript.)
| Model | EKE ACC | EKE RMSE (m$^2$/s$^2$) |
|---|---|---|
| TAU (direct prediction) | 0.3975 | 0.0411 |
| Residual control | 0.3810 | 0.0390 |
| HelmGeoNet | 0.4194 | 0.0382 |
The control improves the day-20 EKE RMSE over the direct TAU baseline, from to m$^2$/s$^2$, and its EKE ACC of stays below the TAU value of . On Page 17, Line 917, we state that "Residual learning with zero initialization thus accounts for much of the gain, mainly in EKE RMSE. The three-branch structure and the relaxation mechanism of HelmGeoNet provide the remaining gain in EKE ACC."
and a learned . Both are added in the revised manuscript. On Page 8, Line 385, we state that "As the most natural no-damping control, also gives a stable rollout, with a day-20 EKE RMSE of m$^2$/s$^2$ and an EKE ACC of ." On Page 8, Line 387, we state that "A control that learns as a free parameter converges to , and every metric stays within of the fixed setting, with the better day-20 EKE forecasts at ." Table 2 lists the full sensitivity study.
Table 2. Sensitivity of HelmGeoNet to the divergence damping coefficient . EKE ACC and latitude-weighted RMSE at leads 5, 10, and 20 days on the test set spanning 2023 to 2024. The best value per column is bolded and the second-best is italicized, with tied best values both bolded. (Table 2 in the revised manuscript.)
| lead 5d ACC | lead 5d RMSE | lead 10d ACC | lead 10d RMSE | lead 20d ACC | lead 20d RMSE | |
|---|---|---|---|---|---|---|
| 0.8664 | 0.0188 | 0.4445 | 0.0710 | 0.0375 | 0.5077 | |
| 0.8700 | 0.0185 | 0.6872 | 0.0283 | 0.4194 | 0.0382 | |
| 0.8732 | 0.0184 | 0.6947 | 0.0284 | 0.4280 | 0.0389 | |
| 0.8705 | 0.0187 | 0.6775 | 0.0297 | 0.2527 | 0.0683 | |
| learned () | 0.8720 | 0.0184 | 0.6897 | 0.0282 | 0.4166 | 0.0383 |
On Page 8, Line 390, we state that "within this range the corrector compensates for the imposed damping, so the differences in skill remain small at all leads. We keep because it is fixed from training data alone and keeps the rollout away from the unstable settings."
Comparison with HelmFluid. HelmFluid is an impressive work and the manuscript cites it. Our work applies the Helmholtz decomposition to the evolution of geostrophic currents specially, so that each component learns its own evolution and the kinetic-energy forecast becomes more accurate and stable. The two frameworks differ in the input structure. Our framework consumes one velocity field per step and forecasts the next field, while HelmFluid consumes a sequence of historical frames to capture the evolution, so a direct comparison requires adapting one of the two input protocols, and we leave it to future work. This could also answer the concerns in W2.
Response to Reviewer hPzN [3/4]
[Response to Q3]: We answer the five asked items in turn.
Clipping fraction. For HelmGeoNet, the count of 18,361 clamping events is accumulated over the 711 daily initializations of the two test years, each rolled out to 20 leads, counting every grid point at every lead once. The count rises from 3 events at lead 1 to 2,235 events at lead 20, about 3 grid points per rollout at the worst lead. All events lie inside the excluded equatorial band, at latitudes between N and N.
Unclipped comparison. Table 1 lists the day-20 EKE skill with and without the clamp for HelmGeoNet and for SimVP, the baseline with the largest change.
Table 1. Day-20 EKE skill on the 2023 to 2024 test set with and without the velocity clamp.
| Model | EKE ACC (with clamp) | EKE RMSE (with clamp) | EKE ACC (without clamp) | EKE RMSE (without clamp) |
|---|---|---|---|---|
| HelmGeoNet | 0.4194 | 0.0382 | 0.4195 | 0.0381 |
| SimVP | 0.3422 | 0.0520 | 0.3393 | 0.1384 |
HelmGeoNet and the other baselines stay essentially unchanged with or without the clamp, with SimVP the exception shown in the table. On Page 9, Line 460, we state that "Notably, HelmGeoNet stays ahead of every baseline with or without the clamp. Removing the clamp gives a day-20 EKE ACC of and an EKE RMSE of m$^2$/s$^2$."
Mean energy bias. Yes. On Page 7, Line 345, we state that "measure the variation of across lead time, and we also report the mean normalized error ... to assess the energy fidelity." Table 2 lists the mean and the maximum absolute normalized energy error of every model over the rollout.
Table 2. Mean and maximum absolute normalized energy error over the 20-lead rollout, in percent. The maximum absolute error is reported with the sign of the error at the lead where the absolute error is largest. (Table S6 in the revised manuscript.)
| Model | Mean error (%) | Max abs error (signed, %) |
|---|---|---|
| HelmGeoNet | ||
| ConvLSTM | ||
| SwinLSTM | ||
| TAU | ||
| SimVP | ||
| PredFormer | ||
| VideoMamba | ||
| FourCastNet | ||
| Earthformer |
On Page 18, Line 967, we state that "SimVP is the only baseline that attains both a smaller mean error and a smaller maximum absolute error, yet its day-20 EKE RMSE of m$^2$/s$^2$ far exceeds the m$^2$/s$^2$ of HelmGeoNet." and "HelmGeoNet keeps both errors steady across leads and stays below in absolute error at every lead, so it avoids the large energy drifts and swings of the baselines."
[Response to Q4]: We answer the two asked items in turn.
Fixed design choices. Yes for the two named choices. On Page 5, Line 263, we state that "The damping coefficient is set to according to the historically measured one-day autocorrelation of Eq. (8), computed on 2000 to 2009, a period entirely inside the training range." On Page 7, Line 331, we state that "where is computed from the training period spanning 2000 to 2020." The 2023-2024 test set are used exclusively for evaluation and are not consulted at any stage of model development.
Reliability over TAU. We have not added independent training runs in this revision, since they exceed our current compute budget. We quantify the variability over the 711 daily initializations with a paired block bootstrap. On Page 18, Line 931, we state that "A paired block bootstrap over the test initializations uses a block length of 30 days and 2000 resamples. The confidence interval for the day-20 EKE RMSE difference to TAU runs from to m$^2$/s$^2$." and "Both comparisons give ." and "The ranking of the three forecasts is the same in each of the two test years, with no reversal at any lead." We will actively pursue multi-seed training in an updated version. This could also answer the concerns in W3.
Response to Reviewer hPzN [1/4]
Thank the reviewer for kind review and the valuable feedback!
[Response to Q1]: We answer the four asked items and then report the residuals for the masked implementation.
Metric factors and spatial units. In the revised manuscript, on Page 15, Line 792, we state that "The spectral operators use a uniform reference zonal spacing dx given by the mean of the local spacing RcosϕΔλ over the latitude grid, and the meridional spacing dy=RΔϕ." On Page 15, Line 794, we state that "Velocities are in m s$^{-1}$ and vorticities and divergences in s$^{-1}$."
Boundary conditions and land filling. On Page 15, Line 787, we state that "Longitude is periodic by construction. The latitude boundaries bound land and sea ice, where the flow is already near zero, and are zero-padded. Before the FFT, land points are initialized to zero and updated by Laplace relaxation (five iterations, with ocean values fixed), producing a smooth extrapolation across the land-sea boundary. The ocean mask is re-applied after the inversion, so the filled values never enter the forecast or the evaluation."
Divergence and curl residuals. Let usol and uirr denote the divergence-free and the curl-free components. In our own verification, the spectral divergence of usol and the spectral curl of uirr on the filled periodic domain vanish with relative residuals below 10−10. Table 1 reports the finite-difference residuals on the ocean points of the masked grid with ∥ϕ∥≥5∘, over the daily fields spanning 2000 to 2009. The raw field is the geostrophic velocity before the decomposition and serves as the reference level.
Table 1. Finite-difference readings on the ocean points with ∥ϕ∥≥5∘, averaged over the daily fields spanning 2000 to 2009.
The readings are dominated by the discretization error of the finite-difference operator itself, so the divergence reading of usol sits at the level of the raw field. When the equatorial band is included, the weighted RMS is 7.51×10−7 s$^{-1}$ for usol against 8.74×10−7 s$^{-1}$ for the raw field. In short, the divergence residual of the decomposition therefore stays at the level already present in the raw data, and the curl reading of uirr stays at the same numerical level.
Weighted orthogonality residuals. Table 2 reports the normalized cross inner products ⟨a,b⟩w/(∣a∣w∣b∣w) under the masked, latitude-weighted inner product, with w=cosϕ and the sum over the valid ocean grid points, on the same daily fields.
Table 2. Normalized cross inner products of the three components under the masked, latitude-weighted inner product. The value is zero under orthogonality. The energy share of a pair, a and b, is 2⟨a,b⟩w/⟨u,u⟩w in percent, with the sign retained. (Table S1 in the revised manuscript.)
On Page 15, Line 780, we state that "Table S1 reports its statistics for the three component pairs on the daily fields spanning 2000 to 2009, and every daily value stays below 0.10 in absolute value." and "The three components are therefore approximately orthogonal on the masked and latitude-weighted grid."
Your periodic analytic example confirms that an unweighted orthogonality does not transfer under cosine weighting. In the revised manuscript, on Page 3, Line 124, we state that "Though the inner product may change on the masked and latitude-weighted grid used for training and evaluation, the components of the AVISO/DUACS geostrophic currents remain approximately orthogonal for component-wise forecasting." In our case, the weighted orthogonality is reported as a measured approximation. On Page 15, Line 767, we update that "Under the masked, latitude-weighted inner product of Eq. (11) the cross inner products are small but not exactly zero." This could also answer the concerns in W1.