Pith. sign in

REVIEW 4 major objections 7 minor 53 references

LASSE: Learning Active Sampling for Storm Tide Extremes in Non-Stationary Climate Regimes

T0 review · 4 major / 7 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read By iteratively retraining a surrogate on the most destructive-looking unsimulated cyclones, the paper claims, a search can pull every retrieved storm from the rare set of storm-tide-producing cyclones using fewer than a fifth of the full…

desk verdict A useful active-learning extension for storm tide extremes, but the headline 100% precision is measured on the self-selected pool and needs a held-out evaluation before it carries weight. read the letter →

arxiv 2501.00149 v2 pith:SHAR3O2Z submitted 2024-12-30 physics.ao-ph cs.LGphysics.geo-ph

classification physics.ao-phcs.LGphysics.geo-ph
keywords stormtideextremestropicalcyclonesactivelearningsurrogatemodelingensemble-approximatedconditionalGaussianprocessclimatechangeBangladeshextremevaluesampling
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper claims that a surrogate model can replace most of the expensive hydrodynamic simulations needed to find the rare tropical cyclones that cause destructive storm tides. It shows that an active, iterative learning loop—rank unsimulated storms by predicted destructiveness, simulate the top few, retrain the surrogate—can retrieve these rare storms with 100% precision while evaluating under 20% of a large catalog of downscaled cyclones. The paper further claims that the surrogate generalizes from storms downscaled from present-climate reanalysis to storms from a high-emissions future climate scenario, retaining roughly 80% precision. If these claims hold, screening very large synthetic storm catalogs for climate risk assessment becomes computationally tractable.

What carries the argument

The central object is the ensemble-approximated conditional Gaussian process (Ens-CGP), a surrogate that maps a 165-dimensional vector of cyclone parameters (latitude, longitude, maximum wind speed, minimum sea-level pressure, and radius of maximum wind at 33 hourly steps around landfall) to a 54-element vector of peak storm tides at stations along the Bangladesh coast. The model is the linear estimator $M = C_{yx}(C_{xx} + \sigma^2 I)^{-1}$, computed cheaply with a reduced-rank singular-value decomposition of the training ensemble. In the informative sampling loop, the surrogate predicts storm tides for all unsimulated cyclones, ranks them by destructiveness—the number of stations where the predicted tide exceeds the 3 meter threshold—and the top 1% are simulated and added to the training set before the surrogate is refit. Adjusting the threshold biases the loop toward precision or recall.

What would settle it

Run the same precision-tuned Ens-CGP active sampler on a fully simulated, independently downscaled storm catalog that was not used in any way to train or select the data, and check whether 100% precision is reached before 20% of the new catalog has been simulated; if precision drops substantially on the unseen storms, the central claim is refuted.

Watch

Extended reading notes

Core claim

The central claim is that an ensemble-approximated conditional Gaussian process (Ens-CGP), retrained online on the most destructive unsimulated cyclones, identifies tropical cyclones that generate storm tides above 3 meters at any of 54 Bangladesh coastal stations with 100% precision after evaluating about 20% of the catalog. In the precision-tuned configuration, perfect precision is reached early in the iterative process while recall continues to climb toward total recall, and a recall-tuned configuration achieves total recall. The paper also reports that a batch-trained surrogate using only 20% of present-climate storm data reaches about 80% precision when tested on future-climate storms, supporting the generalization claim. All results treat water levels relative to mean sea level and do not include sea-level rise.

Load-bearing premise

The paper assumes that precision computed on the shrinking pool of unsimulated cyclones, after the surrogate has been retrained on the very storms it selected as most destructive, is a faithful measure of how the search would perform on a fresh, independent catalog of storms.

Editorial extensions

If this is right

  • Large synthetic storm catalogs can be screened for extreme storm tides with only about a fifth of the full hydrodynamic simulations, making tail-risk estimates feasible for regions with sparse historical storm records.
  • A surrogate trained on present-climate storms carries its skill to future-climate scenarios, so the active search can be run on downscaled storm sets from different climate models and emission scenarios.
  • The threshold knob lets users trade precision for recall, so the same machinery can serve both conservative risk screening and exhaustive hazard discovery.
  • Because the surrogate replaces storm-tide simulations, the active loop can be redeployed to other hazard variables, such as inundation depth from rainfall, without changing the learning framework.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A stronger test than the paper's residual-pool evaluation would hold out an entire climate-model ensemble, run the active loop only on the rest, and score on the held-out storms; this would separate genuine generalization from self-confirmation inside the selection loop.
  • The destructiveness score could be replaced by expected economic damage or by exceedance of a high quantile of the tide distribution, which would directly optimize what risk managers care about rather than an arbitrary 3-meter cut.
  • Adding sea-level rise as an input parameter or post-processed offset would test whether the perfect-precision claim survives under the higher baseline water levels expected in the future climate.
  • The fixed 54-station output vector could be extended to continuous coastline maps, letting the sampler choose which locations are informative rather than assuming all stations are equally important.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 7 minor

Summary. This paper addresses the computational bottleneck of identifying tropical cyclones (TCs) whose storm tides exceed a 3 m damage threshold at any of 54 Bangladesh coastal stations, where gold-standard labels come from expensive ADCIRC hydrodynamic simulations. The authors evaluate surrogate strategies in two modes. In batch mode, XGBoost trained on 75% of an ERA5-downscaled catalog (4100 TCs) achieves 81% precision and 84% recall on the held-out quarter; trained on all ERA5 and tested on 2000 EC-EARTH-3 SSP5-8.5 TCs it achieves 82% precision and 87% recall; and trained on only 20% of ERA5 it retains 80% precision and 84% recall when tested on SSP5-8.5. The headline contribution is an active, 'informative' learning loop (LASSE): starting from a 1% seed, a surrogate iteratively selects the most destructive predicted storms, which are then hydrodynamically simulated and added to the training set. With an ensemble-approximate conditional Gaussian process (Ens-CGP) surrogate, the paper claims 100% precision in retrieving destructive storms after evaluating roughly 20% of the catalog, with an adjustable operating point favoring precision or recall; an XGBoost-based active variant is presented as comparatively ineffective.

Significance. The batch results are internally consistent and practically useful: the ERA5-to-SSP5-8.5 transfer, retained with only 20% of the training fraction, is a genuine cross-climate generalization test and is presented with a clean 50-trial protocol. The paper also deserves credit for flagging its own main caveats (constant sea level in Section VI; ad-hoc threshold mechanism in Section VII), and the Ens-CGP operator in Eq. (8) is consistent with reduced-rank ridge regression. The active-learning contribution, if substantiated on independent evaluation, would be significant: locating rare destructive storms while simulating only about 20% of a catalog would directly reduce the cost of climate hazard screening. However, the headline 100%-precision claim currently rests on an in-pool evaluation without reported trial variance, without a held-out or future-climate test of the active sampler, and at an optimized operating point; its significance is therefore conditional on the additional evaluations requested below. If those evaluations support the claim, this would be a strong contribution to applied climate risk assessment.

major comments (4)
  1. [§V.A–V.B, Figs. 7–8; Abstract] The active-learning evaluation has no held-out test set. In Experiments IVa and IVb the surrogate labels the remaining unsimulated pool, the most destructive predictions are simulated and appended to the training set, and the reported precision and recall are computed on this self-selected, shrinking pool; no random subset of ERA5 is reserved, and the SSP5-8.5 catalog is never presented to the active sampler. Because each selected storm's label comes from the hydrodynamic simulation, the metric is not circular by construction, but the evaluation set is the algorithm's own evolving selection, so the reported precision has an unknown relationship to performance on an untouched catalog. Consequently, the abstract's claim that the informative sampling approach is 'generalizable to climate scenarios' is unsupported: generalization is demonstrated only for the batch-trained XGBoost of Experiments II and III, and even that claim is subject to the paper's own Section VI caveats (constant sea level, embedded SSP5-8.5 characteristics). The authors should hold out a stratified sample of ERA5 or run the trained sampler on the SSP5-8.5 catalog and report precision and recall on that untouched set, and should temper the generalization statement in the abstract until such evidence exists.
  2. [§V.B, Fig. 8] No repetition count or spread is reported for the Ens-CGP active-learning experiment that produces the headline result. Experiments I and III state 50 trials and Experiment IVa states 10 trials, but Experiment IVb reports none, and Fig. 8 shows no error bars or seed-to-seed variability. Since the 1% seed set is drawn randomly and the early iterations of the loop depend on it, a single favorable run could produce the reported 100% precision. The authors should run the Ens-CGP sampler over multiple seeds (comparable to the 50-trial batch protocol), report the distribution of precision, recall, and accuracy versus evaluation fraction, and state how many of the trials reached 100% precision at the 20% point.
  3. [§V.B.2, Fig. 8 caption, §VII] The abstract and Section I state, without qualification, that the active online learning system 'achieves 100% precision after evaluating only 20% of the dataset,' but Fig. 8 attributes 100% precision to the precision-tuned case, which is reached by varying the damage threshold, a mechanism Section VII itself describes as ad hoc. The threshold values used for the untuned, precision-tuned, and recall-tuned cases are not stated, and the recall achieved at the 20% evaluation point in the precision-tuned case is not quantified; the caption's 'maintaining progress toward total recall' is not backed by a number. The paper should state the exact operating points, report precision and recall jointly at 20% for each case, and qualify the abstract claim so that the tuned and untuned results are not conflated.
  4. [§IV–V, Figs. 7–8] The rarity of the target class is never quantified: the paper nowhere states the fraction of the 4100 ERA5 TCs whose maximum simulated storm tide exceeds 3 m, despite calling them 'rare destructive storms,' and neither figure includes a random-query baseline at matched simulation budget. The 'naive' curve in Fig. 7 is not such a baseline: it trains batch models on increasing random fractions and tests on the shrinking complement, so training-set size and test-set composition change together. Without the catalog base rate and a random-sampling active baseline, the improvement attributable to informative selection cannot be separated from the base rate, and the central efficiency claim ('100% precision ... using less than 20% of the simulations') lacks a quantitative benchmark.
minor comments (7)
  1. [§V.A vs. §VI–VII] Experiment IVa is described in §V.A as selecting the top 5% of TCs by predicted destructiveness, but §VI and the Conclusions describe the XGBoost selection as randomly sampling TCs predicted to exceed the threshold; these two descriptions of the same protocol should be reconciled.
  2. [§V.B.1, Eq. (8)] The ridge parameter σ² is said only to be 'chosen empirically'; unlike the exhaustive XGBoost hyperparameter list in §IV, no value or sensitivity analysis is provided for this parameter, which directly controls the Ens-CGP operator in Eq. (8).
  3. [§III.C] The reported catalog sizes (4100 and 2000 tracks) are counts before applying the exclusion of TCs that never reach land or that make two or more landfalls; the effective numbers of tracks actually used in each experiment should be stated.
  4. [§II] The paper surveys adaptive Kriging surrogate models (refs. [16], [18], [47]) but gives no comparison against them on the same catalog; a standard adaptive Kriging or other published active-surrogate baseline would contextualize the claimed Ens-CGP gains.
  5. [§II, §V.B.2] Minor typos: 'adaptive obswervations' in §II should be 'adaptive observations,' and the sentence beginning 'the training iteration is updated (j ← j + 1) after simulating the top 1% ...' in §V.B.2 is grammatically incomplete and should be rewritten.
  6. [§IV.B, Fig. 4] Experiment II (ERA5 to SSP5-8.5) is reported as a single run with no trial count while Experiments I and III use 50 trials; a multi-seed estimate for Experiment II would make the cross-scenario comparison more robust.
  7. [Reproducibility] The paper provides no data or code availability statement; given that the ADCIRC outputs and downscaled track catalogs are the basis of all experiments, a reproducibility statement would be expected.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the active-learning precision claim is an empirical result validated by independent hydrodynamic simulations.

full rationale

The paper's load-bearing claims are not circular. The surrogate models (XGBoost and Ens-CGP) are trained on hydrodynamic simulation outputs from ADCIRC and tested on held-out or cross-scenario storm catalogs; the Ens-CGP equations are stated in the paper, so the result does not reduce to a self-citation. The active-learning loop selects TCs predicted destructive, simulates them, and then evaluates precision against the true ADCIRC labels of the unselected pool; true labels are not defined by the surrogate, so the 100% precision claim is an empirical retrieval outcome rather than a construction. The same damage threshold defines both the target class and the informativeness score, but that is a consistent objective rather than a circular reduction. The paper explicitly acknowledges limitations: the future-scenario test tracks embed SSP5-8.5 characteristics, sea-level rise is neglected, and the threshold tuning is ad hoc (Section VI and Conclusions). These are robustness concerns, not evidence that any equation reduces to its own input. No Eq. X = Eq. Y by construction was found; therefore the circularity score is 0.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central empirical claims rest on the correctness of the ADCIRC storm tide labels, the representativeness of the downscaled TC catalogs, and the sufficiency of the selected input features. The method itself introduces several hand-chosen settings: the Ens-CGP ridge parameter, singular-value rank truncation, the 3 m damage threshold, seed size and selection fraction, and XGBoost hyperparameters. No new physical entities are postulated.

free parameters (5)
  • Ens-CGP ridge parameter sigma^2 = not reported
    Chosen empirically as regularization in Eq. (1) and Eq. (8); no value or selection procedure is given (Section V.B).
  • Rank truncation k (99% cumulative singular energy) = k less than Rank(Sigma), unspecified
    Heuristic based on training data; affects the Ens-CGP approximation (Section V.B.1, step 4).
  • Damage threshold 3 meters = 3 m
    Defines positive class and query selection; adjustable to trade precision and recall (Sections IV.A and V.B.2).
  • Seed size and per-iteration selection fraction = 1% seed and 1% per iteration for Ens-CGP; 1% seed and 5% per iteration for XGBoost
    Algorithm hyperparameters chosen without stated optimization (Sections V.A and V.B.2).
  • XGBoost hyperparameters = eta=0.3, max_depth=6, subsample=1, etc.
    Fixed after several attempts at other models; no tuning or search protocol is reported (Section IV).
assumptions (5)
  • domain assumption Downscaled synthetic TC catalogs from ERA5 and EC-EARTH-3 represent the relevant storm climatology for current and future Bangladesh risk.
    Used as training and testing populations; generated by statistical-deterministic downscaling cited to Emanuel et al. (Section III.A).
  • domain assumption ADCIRC hydrodynamic simulations accurately compute storm tides for each TC.
    Labels come from ADCIRC; validation against Sidr and tide gauges is summarized but not shown (Section III.B).
  • domain assumption The 33 hourly time steps and five storm parameters contain enough information to predict peak storm tide at 54 stations.
    Surrogate inputs are extracted this way (Section III.C).
  • domain assumption Constant sea level, with zero sea-level rise, is an acceptable simplification for evaluating the surrogate generalization claim.
    Stated caveat in Section VI; affects the future climate generalization result.
  • domain assumption Excluding TCs with no landfall or two or more landfalls does not bias the extreme storm tide results.
    Data filtering rule in Section III.C; such storms are outside the evaluated catalog.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LASSE: Learning Active Sampling for Storm Tide Extremes in Non-Stationary Climate Regimes." pith.science (2026). https://pith.science/paper/SHAR3O2Z

@misc{pith2026250100149,
  author       = {Pith},
  title        = {Pith review of: LASSE: Learning Active Sampling for Storm Tide Extremes in Non-Stationary Climate Regimes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SHAR3O2Z}},
  note         = {Machine review of arXiv:2501.00149}
}
read the original abstract

Identifying tropical cyclones that generate destructive storm tides for risk assessment, such as from large downscaled storm catalogs for climate studies, is often intractable because it entails many expensive Monte Carlo hydrodynamic simulations. Here, we show that surrogate models are promising from accuracy, recall, and precision perspectives, and they "generalize" to novel climate scenarios. We then present an informative online learning approach to rapidly search for extreme storm tide-producing cyclones using only a few hydrodynamic simulations. Starting from a minimal subset of TCs with detailed storm tide hydrodynamic simulations, a surrogate model selects informative data to retrain online and iteratively improves its predictions of damaging TCs. Results on an extensive catalog of downscaled TCs indicate 100% precision in retrieving rare destructive storms using less than 20% of the simulations as training. The informative sampling approach is efficient, scalable to large storm catalogs, and generalizable to climate scenarios.

Figures

Figures reproduced from arXiv: 2501.00149 by the authors.

Figure 1
Figure 1. The standard workflow to quantify TC-induced [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. ML-based surrogate allows for faster risk [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Average confusion matrix from repeated trials of [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 5
Figure 5. Figure 5: Averaged confusion matrix from repeated trials of [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: The flowchart for active and informative sampling [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]
Figure 7
Figure 7. Figure 7: Comparison of naive and iterative approaches for storm tide prediction using the XGBoost surrogate model. The [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]
Figure 8
Figure 8. Figure 8: Comparison of naive batch training (top), [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

53 extracted references · 51 canonical work pages

  1. [1]

    Given a training dataset △j = {(x1, y1),

    Training Procedure Instead of explicitly calculating the covariances, a reduced-rank square-root ensemble representation pro- vides computational efficiency. Given a training dataset △j = {(x1, y1), . . . ,(xEj , yEj )} at the jth retraining iter- ation, the following steps are performed:

  2. [2]

    Calculate T raining Ensemble Means: ¯xj = 1 Ej EjX i=1 xi, (3) ¯yj = 1 Ej EjX i=1 yi. (4)

  3. [3]

    , xEj − ¯xj i , (5) ˜Yj = h y1 − ¯yj,

    Compute the Deviation Ensemble Matrices: ˜Xj = h x1 − ¯xj, . . . , xEj − ¯xj i , (5) ˜Yj = h y1 − ¯yj, . . . , yEj − ¯yj i . (6)

  4. [4]

    Perform Singular V alue Decomposition (SVD): ˜Xj = U ΣV T . (7)

  5. [5]

    The selected value of k cov- ers 99% of the cumulative energy of the singular values, ensuring k <Rank(Σ)

    Estimate the Reduced Ens-CGP Model: Mj = ˜YjVβΣβ Σ2 β + σ2I −1 U T β , (8) where Uβ = U (:, 1 : k), Vβ = V (:, 1 : k), and Σβ = Σ(1 : k, 1 : k). The selected value of k cov- ers 99% of the cumulative energy of the singular values, ensuring k <Rank(Σ)

  6. [6]

    From these predictions, TCs exceeding the dam- age threshold (e.g., > 3 meters) are ranked based on their destructiveness

    Iterative Refinement Process At iteration j, the model Mj predicts storm tide vec- tors for storm parameter vectors not in the training dataset. From these predictions, TCs exceeding the dam- age threshold (e.g., > 3 meters) are ranked based on their destructiveness. the training iteration is updated (j ← j + 1) after simulating the top 1% of the most dam...

  7. [7]

    Corwin L. Atwood. Sequences Converging to D-Optimal Designs of Experiments. The Annals of Statistics , 1(2):342 – 352, 1973

  8. [8]

    Xgboost: A scalable tree boosting system

    Tianqi Chen and Carlos Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Dis- covery and Data Mining , KDD ’16, page 785–794. ACM, August 2016

Show all 53 references
  1. [9]

    Sequential design of experiments

    Herman Chernoff. Sequential design of experiments. Ann. Math. Statist. , 30(3):755–770, September 1959

  2. [10]

    D. A. Cohn, Z. Ghahramani, and M. I. Jordan. Active learning with statistical models, 1996

  3. [11]

    David A. Cohn. Neural network exploration using op- timal experiment design. Neural Networks , 9(6):1071– 1083, 1996

  4. [12]

    D¨ oscher, M

    R. D¨ oscher, M. Acosta, A. Alessandri, P. Anthoni, T. Ar- souze, T. Bergman, R. Bernardello, S. Boussetta, L.-P. Caron, G. Carver, M. Castrillo, F. Catalano, I. Cvi- janovic, P. Davini, E. Dekker, F. J. Doblas-Reyes, D. Docquier, P. Echevarria, U. Fladrich, R. Fuentes- Franco...

  5. [13]

    Tropical cyclone risk in bangladesh

    Kerry Emanuel. Tropical cyclone risk in bangladesh. MAUSAM, 72(1):27–34, 2021

  6. [14]

    Environmental control of tropical cyclone intensity

    Kerry Emanuel, Christopher DesAutels, Christopher Holloway, and Robert Korty. Environmental control of tropical cyclone intensity. Journal of the atmospheric sciences, 61(7):843–858, 2004

  7. [15]

    A statistical deterministic approach to hur- ricane risk assessment

    Kerry Emanuel, Sai Ravela, Emmanuel Vivant, and Camille Risi. A statistical deterministic approach to hur- ricane risk assessment. Bulletin of the American Meteo- rological Society, 87(3):299–314, 2006

  8. [16]

    Hurricanes and global warming: Results from downscaling ipcc ar4 simulations

    Kerry Emanuel, Ragoth Sundararajan, and John Williams. Hurricanes and global warming: Results from downscaling ipcc ar4 simulations. Bulletin of the Amer- ican Meteorological Society, 89(3):347–368, 2008

  9. [17]

    The ensemble kalman filter: theoretical formulation and practical implementation

    Geir Evensen. The ensemble kalman filter: theoretical formulation and practical implementation. Ocean Dy- namics, 53(4):343–367, Nov 2003

  10. [18]

    V.V. Fedorov. Theory of Optimal Experiments. Academic Press, 1972

  11. [19]

    The era5 global reanalysis

    Hans Hersbach, Bill Bell, Paul Berrisford, Shoji Hira- hara, Andr´ as Hor´ anyi, Joaqu ´ ın Mu˜ noz-Sabater, Julien Nicolas, Carole Peubey, Raluca Radu, Dinand Schepers, et al. The era5 global reanalysis. Quarterly Journal of the Royal Meteorological Society, 146(730):1999–2049, 2020

  12. [20]

    Greg J. Holland. An analytic model of the wind and pressure profiles in hurricanes. Monthly Weather Review, 108(8):1212–1218, 1980

  13. [21]

    Holland et al

    Greg J. Holland et al. Tropical cyclone motion: Envi- ronmental interaction plus a beta effect. Colorado State University Atmospheric Science Paper Series , 355, 1982. Available through the CSU Libraries repository

  14. [22]

    Taflanidis, Norberto C

    Gaofeng Jia, Alexandros A. Taflanidis, Norberto C. Nadal-Caraballo, Jeffrey A. Melby, Andrew B. Kennedy, and Jane M. Smith. Surrogate modeling for peak or time-dependent storm surge prediction over an extended coastal region using an existing database of synthetic storms. Natu...

  15. [23]

    Storm surge hazard over bengal delta: a probabilistic– deterministic modelling approach

    Md Jamal Uddin Khan, Fabien Durand, Kerry Emanuel, Yann Krien, Laurent Testut, and AKM Islam. Storm surge hazard over bengal delta: a probabilistic– deterministic modelling approach. Natural Hazards and Earth System Sciences , 22(7):2359–2379, 2022

  16. [24]

    Kyprioti, Christopher Irwin, Alexandros A

    Aikaterini P. Kyprioti, Christopher Irwin, Alexandros A. Taflanidis, Norberto C. Nadal-Caraballo, Madison C. Yawn, and Luke A. Aucoin. Spatio-temporal storm surge emulation using gaussian process techniques. Coastal En- gineering, 180:104231, 2023

  17. [25]

    Physically based assessment of hurri- cane surge threat under climate change

    Ning Lin, Kerry Emanuel, Michael Oppenheimer, and Erik Vanmarcke. Physically based assessment of hurri- cane surge threat under climate change. Nature Climate Change, 2(6):462–467, 2012

  18. [26]

    A machine learning-based parameterized tropical cyclone precipitation model

    Yi Lu, Jie Yin, Peiyan Chen, Hui Yu, and Sirong Huang. A machine learning-based parameterized tropical cyclone precipitation model. International Journal of Disaster Risk Science, 15(6):972–985, Dec 2024

  19. [27]

    Adcirc: an advanced three- dimensional circulation model for shelves, coasts, and estuaries

    Richard Albert Luettich, Joannes J Westerink, Nor- man W Scheffner, et al. Adcirc: an advanced three- dimensional circulation model for shelves, coasts, and estuaries. report 1, theory and methodology of adcirc- 2dd1 and adcirc-3dl. Technical report, US Army Corps of Engineers, 1992

  20. [28]

    David J. C. MacKay. Information-based objective func- tions for active data selection. Neural Computation , 4(4):590–604, 1992

  21. [29]

    Learning surrogate rainfall-driven inundation models with few data

    Marzieh Alireza Mirhoseini. Learning surrogate rainfall-driven inundation models with few data. arXiv:2411.19323, 2024

  22. [30]

    Morss, Kerry A

    Rebecca E. Morss, Kerry A. Emanuel, and Chris Sny- der. Idealized adaptive observation strategies for improv- ing numerical weather prediction. Journal of the Atmo- spheric Sciences, 58(2):210 – 232, 2001

  23. [31]

    A review of tropical cyclone-generated storm surges: Global data sources, observations, and impacts

    Hal F Needham, Barry D Keim, and David Sathiaraj. A review of tropical cyclone-generated storm surges: Global data sources, observations, and impacts. Reviews of Geo- physics, 53(2):545–591, 2015

  24. [32]

    Risks of coastal storm surge and the effect of sea level rise in the red river delta, vietnam

    James E Neumann, Kerry A Emanuel, Sai Ravela, Lind- say C Ludwig, and Caroleen Verly. Risks of coastal storm surge and the effect of sea level rise in the red river delta, vietnam. Sustainability, 7(6):6553–6572, 2015

  25. [33]

    The scenario model intercomparison project (scenariomip) for cmip6

    Brian C O’Neill, Claudia Tebaldi, Detlef P Van Vuuren, Veronika Eyring, Pierre Friedlingstein, George Hurtt, Reto Knutti, Elmar Kriegler, Jean-Francois Lamarque, Jason Lowe, et al. The scenario model intercomparison project (scenariomip) for cmip6. Geoscientific Model De- velo...

  26. [34]

    A framework for flexible peak storm surge prediction

    Benjamin Pachev, Prateek Arora, Carlos del Castillo- Negrete, Eirik Valseth, and Clint Dawson. A framework for flexible peak storm surge prediction. arXiv preprint arXiv:2204.13168, 2022

  27. [35]

    Global storm tide modeling with adcirc v55: unstructured mesh design and performance

    William J Pringle, Damrongsak Wirasaet, Keith J Roberts, and Joannes J Westerink. Global storm tide modeling with adcirc v55: unstructured mesh design and performance. Geoscientific Model Development , 14(2):1125–1145, 2021

  28. [36]

    Bangladesh’s accelerating coastal flood hazard

    Jiangchao Qiu, Sai Ravela, and Kerry Emanuel. Bangladesh’s accelerating coastal flood hazard. arXiv:2312.06051, 2023

  29. [37]

    Rasmussen and C.K.I

    C.E. Rasmussen and C.K.I. Williams. Gaussian Pro- cesses for Machine Learning . Adaptive Computation And Machine Learning. MIT Press, 2005

  30. [38]

    Ravela and Dennis B

    S. Ravela and Dennis B. McLaughlin. Fast ensemble smoothing. Ocean Dynamics, 57:123–134, 2007

  31. [39]

    Tractable Non-Gaussian Representations in Dynamic Data Driven Coherent Fluid Mapping , pages 29–46

    Sai Ravela. Tractable Non-Gaussian Representations in Dynamic Data Driven Coherent Fluid Mapping , pages 29–46. Springer International Publishing, Cham, 2018

  32. [40]

    Statistical- deterministic approach to natural disaster prediction, June 8 2010

    Sai Ravela and Kerry A Emanuel. Statistical- deterministic approach to natural disaster prediction, June 8 2010. US Patent 7,734,245

  33. [41]

    A realtime observatory for labora- tory simulation of planetary flows.Experiments in Fluids, 48(5):915–925, May 2010

    Sai Ravela, John Marshall, Chris Hill, Andrew Wong, and Scott Stransky. A realtime observatory for labora- tory simulation of planetary flows.Experiments in Fluids, 48(5):915–925, May 2010

  34. [42]

    Oceanmesh2d 1.0: Matlab-based software for two-dimensional unstructured mesh generation in coastal ocean modeling

    Keith J Roberts, William J Pringle, and Joannes J Westerink. Oceanmesh2d 1.0: Matlab-based software for two-dimensional unstructured mesh generation in coastal ocean modeling. Geoscientific Model Develop- ment, 12(5):1847–1868, 2019

  35. [43]

    On the automatic and a priori design of unstructured mesh resolution for coastal ocean circulation models

    Keith J Roberts, William J Pringle, Joannes J Westerink, Maria Teresa Contreras, and Damrongsak Wirasaet. On the automatic and a priori design of unstructured mesh resolution for coastal ocean circulation models. Ocean Modelling, 144:101509, 2019. 11

  36. [44]

    Tidal varia- tions associated with sea level changes in the northern bay of bengal

    Linta Rose and Prasad K Bhaskaran. Tidal varia- tions associated with sea level changes in the northern bay of bengal. Estuarine, Coastal and Shelf Science , 272:107881, 2022

  37. [45]

    Statistical-physical ad- versarial learning from data and models for downscaling rainfall extremes

    Anamitra Saha and Sai Ravela. Statistical-physical ad- versarial learning from data and models for downscaling rainfall extremes. Journal of Advances in Modeling Earth Systems, 16(6):e2023MS003860, 2024. e2023MS003860 2023MS003860

  38. [46]

    A tutorial on gaussian process regression: Modelling, ex- ploring, and exploiting functions

    Eric Schulz, Maarten Speekenbrink, and Andreas Krause. A tutorial on gaussian process regression: Modelling, ex- ploring, and exploiting functions. Journal of Mathemat- ical Psychology, 85:1–16, 2018

  39. [47]

    Active learning literature survey

    Burr Settles. Active learning literature survey. Technical Report 1648, Computer Sciences, University of Wiscon- sin–Madison, January 2010

  40. [48]

    Characteristics of tides in the bay of bengal

    B Sindhu and AS Unnikrishnan. Characteristics of tides in the bay of bengal. Marine Geodesy , 36(4):377–407, 2013

  41. [49]

    Informative neural ensemble kalman learning

    Margaret Trautner, Gabriel Margolis, and Sai Rav- ela. Informative neural ensemble kalman learning. arXiv:2008.09915, 2020

  42. [50]

    Westerink, R

    Joannes J. Westerink, R. A. Luettich Jr, C. A. Blain, and Norman W. Scheffner. Adcirc: an advanced three- dimensional circulation model for shelves, coasts, and es- tuaries. report 2. user’s manual for adcirc-2ddi. Technical report, US Army Corps of Engineers, 1994

  43. [51]

    Multirotor ensem- ble model predictive control i: Simulation experiments

    Erina Yamaguchi and Sai Ravela. Multirotor ensem- ble model predictive control i: Simulation experiments. arXiv, 2305.12625, 2023

  44. [52]

    Zahura, Jonathan L

    Fatema T. Zahura, Jonathan L. Goodall, Jeffrey M. Sadler, Yu Shen, Mohamed M. Morsy, and Monica Behl. Training machine learning surrogate models from a high- fidelity physics-based model: Application for real-time street-scale flood prediction in an urban coastal commu- nity. ...

  45. [53]

    Taflanidis, Norberto C

    Jize Zhang, Alexandros A. Taflanidis, Norberto C. Nadal- Caraballo, Jeffrey A. Melby, and Fatimata Diop. Ad- vances in surrogate modeling for storm surge prediction: storm selection and addressing characteristics related to climate change. Natural Hazards, 94(3):1225–1253, Dec 2018

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.