Pith. sign in

REVIEW 3 major objections 5 minor 70 references

Breaking the Statistical Similarity Trap in Extreme Convection Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper provides quantitative evidence that correlation-based evaluation can hide total failure on extreme convection, and presents a dual-decoder model that recovers operational skill by separating background from extreme residual.

desk verdict Genuine task formulation and a credible metric-trap demo, but the ground-truth channel and an unexplained ablation inconsistency make the headline numbers unreliable. read the letter →

arxiv 2509.09195 v2 pith:UJ4SNR67 submitted 2025-09-11 cs.LG cs.CV

Md Tanveer Hossain Munim This is my paper
classification cs.LGcs.CV
keywords extremeconvectiondetectionbrightnesstemperaturedownscalingCriticalSuccessIndeximbalancedregressiondual-decoderarchitectureIVTparadoxatmosphericriverforecastingdeeplearningweathermodels
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

This paper argues that standard deep-learning weather evaluations reward similarity to the average scene and therefore miss rare, dangerous convection. It demonstrates the trap empirically: a classical statistical model reaches 0.979 correlation and 0.968 SSIM while scoring exactly zero on the Critical Success Index for pixels colder than 220 K, and modern neural baselines plateau below CSI 0.20 under conventional training. The proposed fix, DART, splits the target brightness-temperature field into a smooth background and a sparse extreme residual learned by separate decoders, oversamples event-rich samples, and tunes a bias-skill trade-off with a beta parameter. The paper also reports an 'IVT paradox': removing integrated water-vapor transport, normally considered essential for atmospheric-river work, improves dangerous-convection CSI by 270.66%, and removing 500 hPa temperature by 248.65%. On the test set DART reaches CSI 0.273 at bias 2.52 versus 0.270 at bias 6.72 for the strongest single-decoder baseline, and the authors verify the system on the August 2023 Chittagong flood cases.

What carries the argument

The load-bearing mechanism is a mathematical decomposition of the target field: ground-truth brightness temperature is split at 225 K into a background target (all convection capped at 225 K) and an extreme residual target (the zero-everywhere-except-cold-cores difference). A shared ResNet-style encoder feeds two decoders—a continuity head trained on the background and an extreme head trained on the residual—and their outputs are added to form the final field. A weighted random sampler up-weights samples by their fraction of pixels below 220 K, and a composite loss with weight beta on the extreme head acts as an 'aggressiveness knob' for deployment. This design forces the network to learn ra

What would settle it

Replace the ground-truth labels with an independently derived cloud-top temperature product from an infrared window channel (about 10-11 micrometers) for the same scenes and recompute DART's CSI at the 220 K threshold; if the spatial pattern of detected 'dangerous convection' does not match that independent product, or if the improvement over baselines disappears, then the paper's ground truth—and therefore the Statistical Similarity Trap numbers—is an artifact of channel choice.

Watch

Extended reading notes

Core claim

The paper's central claim is that the 'Statistical Similarity Trap' is real and measurable: models that optimize bulk agreement on coarse atmospheric-to-satellite downscaling can look near-perfect while missing the extreme convection that matters. The paper asserts four findings: (1) classical and simple ML baselines, as well as standard U-Net-style deep models, fail operationally (CSI = 0.00 or below 0.20) even when correlation and SSIM are high; (2) an ablation over all 21 input combinations shows that removing IVT and 500 hPa temperature improves detection, while removing 700 hPa relative humidity degrades it; (3) DART's architectural separation of background and extreme residual, combine

Load-bearing premise

The load-bearing premise is that the 6.2 micrometer water-vapor brightness temperature from the geostationary satellite, thresholded at 220 K, is a faithful proxy for cloud-top temperature and thus dangerous convection; if the water-vapor channel responds to upper-tropospheric moisture rather than cloud tops, the ground-truth labels, CSI thresholds, and all reported conclusions measure a different quantity.

Editorial extensions

If this is right

  • If the Statistical Similarity Trap diagnosis is right, models that look excellent on correlation or SSIM—and are often published as state-of-the-art—can be operationally useless; evaluation must include event-based metrics such as CSI before deployment decisions.
  • The IVT Paradox, if it generalizes, means predictor selection for convective-scale downscaling cannot be inherited from synoptic-scale preferences: removing IVT and 500 hPa temperature improves CSI@220K by 270.66% and 248.65% respectively, while removing 700 hPa relative humidity hurts.
  • The dual-decoder decomposition with event-weighted sampling reaches CSI 0.273 at bias 2.52, against 0.270 at bias 6.72 for the strongest single-decoder baseline, meaning architectural separation, not just training tricks, is what keeps false alarms manageable at equal skill.
  • The beta-tuning knob gives operators a range from conservative (bias near 1.5) to aggressive (bias near 3) settings from one trained system, so the same model can serve routine forecasting and emergency-response modes.
  • With training under 10 minutes on a single GPU and standard forecast inputs, the framework is deployable in regions that lack supercomputing resources.

Reading between the lines

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

  • Inference: The IVT paradox may be partly an artifact of bilinear interpolation: IVT at roughly 25 km spacing is smoothed before being regridded to a 2 km grid, so the model sees contaminated fine-scale signal; ingesting IVT at native resolution or through a separate multi-scale stream could reverse the sign of the effect. The paper raises interpolation artifacts as a possible cause but does not te
  • Inference: The reported 270.66% improvement is relative to a near-zero baseline (CSI 0.043 for the full five-variable model); the practical lesson is about input selection and metric choices, not a claim that dropping IVT produces high absolute skill on its own.
  • Inference: The background/extreme decomposition is a generic template for any heavy-tailed target that is a smooth field plus sparse anomalies—precipitation, fire radiative power, power-grid failures, disease-surveillance maps—and the paper's core architectural idea should transfer, though the 225 K split threshold is domain-specific.
  • Inference: A direct sensitivity test would recompute all CSI numbers using an infrared window channel (around 10-11 micrometers) cloud-top product instead of the 6.2 micrometer water-vapor band; if the spatial agreement with deep convective cores changes materially, the ground-truth definition—not the model—is the limiting factor.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes DART, a dual-decoder U-Net that downscales ERA5 atmospheric predictors to Himawari brightness-temperature fields, with the stated goal of detecting extreme convection. It reports four headline findings: (1) a 'Statistical Similarity Trap' in which classical baselines such as MOS achieve correlation 0.979 and SSIM 0.968 yet CSI=0; (2) an 'IVT Paradox' in which removing Integrated Water Vapor Transport improves CSI@220K by 270.66%; (3) DART achieves CSI=0.273 with bias 2.52 versus 6.72 for an enhanced Attention U-Net at similar CSI; and (4) a Chittagong flood case study demonstrating operational skill. The evaluation uses 150 atmospheric-river test samples over the Bay of Bengal (2015-2023) and WMO-style CSI/HSS/POD/FAR/bias verification. The target is stated in §3.1.1 as Himawari-8/9 Band 8 (6.2 µm water vapor) brightness temperature, but the paper repeatedly labels this 'cloud-top brightness temperature' and applies IR-window convective thresholds.

Significance. If the findings hold, the paper provides a useful warning about the inadequacy of bulk similarity metrics for rare-event verification and offers a practical architecture for imbalanced regression in a meteorological downscaling task. The paper's strengths include full code/data availability, explicit reproducibility statements, a physically motivated oversampling strategy, and an unusually transparent discussion of the 'IVT Paradox' as an empirical result rather than an expected one. However, the central quantitative claims currently rest on two load-bearing assumptions: the identification of a 6.2 µm water-vapor channel with cloud-top IR-window temperature, and the internal consistency of the ablation baseline. Both need to be resolved before the conclusions can be accepted; the ablation inconsistency in particular undermines the headline '270.66%' improvement as currently stated.

major comments (3)
  1. [§3.1.1 and §3.4] The ground truth is defined as 'Himawari-8/9 brightness temperature observations, derived from Band 8 (Water Vapor)' (§3.1.1), yet the paper repeatedly calls this 'cloud-top brightness temperature' (Figs. 5–7) and applies the 210/220/230 K thresholds from Johnson et al. [55] and Martin et al. [56]. Those thresholds are established for IR-window (10–12 µm) brightness temperature, not for the 6.2 µm water-vapor band. In 6.2 µm imagery, clear-sky upper-tropospheric water vapor can produce brightness temperatures below 220 K in the absence of deep convection, and cloudy weighting functions peak at different heights than IR-window bands. Consequently, every CSI/HSS/POD/bias result, the MOS trap demonstration, the IVT ablation, and the Chittagong pixel verification are computed against a target whose meteorological meaning is unverified. Please either switch to an IR-window channel (e.g., Hima
  2. [§4.2, Table 3, Appendix C] The ablation study reports CSI@220K = 0.043 for the full five-variable Attention U-Net (Table C2, 'all variables'; Table C3 baseline), but the architecture study, which uses 'an identical training protocol' (Appendix C), reports CSI@220K = 0.198±0.201 for the Attention U-Net trained on all five variables (Table 3; Table B1). If 0.198 is the correct all-variable baseline, removing IVT to obtain 0.161 is a 18.7% decrease, not the +270.66% improvement claimed in §4.2 and Table C3. The same inconsistency affects the claim that T500+W500 (0.176) outperforms the full variable set. Please rerun the ablation with the exact model and protocol used in the architecture study, or reconcile the two baselines. This is required because the IVT Paradox is one of the four headline findings.
  3. [§4.3, Appendix G.4, Abstract] The abstract and conclusion report 'CSI = 0.273 with bias = 2.52' for DART without qualifying that this is computed on 111 of 150 test samples selected by requiring ≥1% dangerous-convection pixel coverage. The selection criterion is disclosed in §4.3 and Appendix G.4, but the headline is not. Since filtering out null-event samples can inflate CSI, please report CSI/HSS/POD/FAR/bias on the full 150-sample test set and compare DART with the Attention U-Net on that full set, or explicitly frame the headline as 'on events with ≥1% convection coverage.'
minor comments (5)
  1. [Appendix B vs Appendix F] Parameter counts for the architecture study are inconsistent: Table B1 reports 34.5M parameters for Attention U-Net and 8.1M for Lightweight CNN, while Table F6 and Table 3 report 21.4M and 1.1M respectively. These should be reconciled.
  2. [Table D4, §4.1] The MOS baseline is credited with 'bulk similarity' success, but Table D4 shows RMSE = 94.04 K and R² = -16.38. These values are not characteristic of a good bulk forecast; please clarify what correlation/SSIM are being computed over and report additional diagnostics such as conditional bias or quantile performance.
  3. [§4.3.3, Fig. 3] The beta-sensitivity discussion states numerical values in the text, but Figure 3 would benefit from explicit axis labels and confidence intervals, especially since the reported CSI differences across beta values are within one standard deviation of each other.
  4. [§4.4.1] The third case is selected as the 'best-performing' small-convection case by algorithmic search. This is not a blind test. Including an average- or poor-performing case would give a more representative picture of operational skill, or the selection should be clearly labeled as a best-case illustration in the main text.
  5. [§5.1] The comparison of DART's CSI=0.273 with 'the 0.19-0.22 range achieved by leading convection-allowing models' in NOAA evaluations [57] is apples-to-oranges: the NOAA numbers are for severe-weather object verification, not thresholded satellite brightness temperature. Please either remove this comparison or provide a direct peer it is matched to in event definition and verification protocol.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: headline results are empirical train/test measurements; flagged issues (Band 8 ground truth, beta selection) are correctness/statistical concerns, not by-construction reductions.

full rationale

The paper does not attempt an analytic derivation; its four headline findings are empirical evaluations on a chronological 80/10/10 split. The 'Statistical Similarity Trap' result (MOS correlation 0.979 / CSI 0.000, §4.1) is exactly the intended contrast between bulk-similarity and event-based metrics; CSI is computed from thresholded predictions, not defined from the correlation, so there is no by-construction equivalence. The DART decomposition T_GT = T_BG + T_RES (§3.3.2) is a target-identity used for supervision; the model must still learn T_RES from ERA5 inputs, and Table 4 shows it does so imperfectly (POD 0.573, FAR 0.593). The IVT ablation (§4.2, Table C3) is a measured leave-one-out comparison; the 270.66% figure is a relative change in test CSI and is not forced by the loss or the sampling weights. Aligning the tiered-loss thresholds (220K/210K), oversampling criterion (220K), and evaluation thresholds is standard supervised calibration, not circularity. Two concerns raised by the text — using Himawari Band 8 (6.2 µm water vapor) as 'cloud-top brightness temperature' (§3.1.1) and reporting test-set CSI for beta values tuned post hoc (§4.3.3) — are physical-correctness and statistical-inference weaknesses, respectively; neither reduces a claimed prediction to its inputs by construction. There are no load-bearing self-citations or imported uniqueness theorems. Therefore the paper is not circular; score 1 reflects the mild post-hoc test-set beta reporting, which is a methodological caveat rather than a circular step.

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

The central claims are empirical ML results, so the ledger is dominated by hand-set loss weights, thresholds, and assumptions about data fidelity. The most consequential assumption is the use of Himawari Band 8 water vapor BT as a cloud-top temperature proxy; if wrong, all CSI thresholds lose their stated physical meaning.

free parameters (5)
  • Extreme-head loss weight beta = headline beta=1.2; range 1.0-2.0
    Controls aggressiveness and the bias-skill trade-off; the paper does not state whether the headline value was selected on validation or test data.
  • Tiered loss weights for extreme head = 10.0 for T<=220K, 25.0 for T<=210K
    Hand-set weights directly supervise rare-event detection and shape the reported CSI values.
  • Background/extreme decomposition threshold = 225 K
    Chosen as a buffer between the 230K and 220K convection levels; determines the residual target used for training.
  • Oversampling base weight = 0.1 added to eventfulness percentage
    Hand-set to prevent event-free samples from being excluded; affects the training curriculum and final performance.
  • Composite loss weights alpha and delta = alpha=0.5, delta=0.4
    Weights for continuity and final fused losses; no sensitivity analysis is provided.
assumptions (5)
  • domain assumption ERA5 reanalysis accurately represents the atmospheric state at AR timestamps
    All training predictors come from ERA5; errors in reanalysis propagate into the learned transfer function (Section 3.1.1).
  • domain assumption The UCLA AR Catalog correctly identifies AR events in the Bay of Bengal region
    Dataset curation filters on this catalog; incorrect detections would mislabel non-AR samples as AR samples (Section 3.1.1).
  • domain assumption Himawari-8/9 Band 8 water vapor brightness temperature can be treated as cloud-top brightness temperature for convection thresholds
    Load-bearing ground-truth assumption; no spectral reference or justification is given (Section 3.1.1).
  • domain assumption Bilinear interpolation of 0.25-degree ERA5 fields to 0.02-degree grid introduces no fatal information loss for convective detection
    All predictors are interpolated before training; the model must recover fine-scale extremes from smoothed inputs (Section 3.1.2).
  • domain assumption Filtering the test cohort to 111 events with at least 1% dangerous-convection coverage does not bias headline performance comparisons
    Evaluation is reported only on the favorable filtered subset; no full-test-set metrics are given (Appendix G.4).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Breaking the Statistical Similarity Trap in Extreme Convection Detection." pith.science (2026). https://pith.science/paper/UJ4SNR67

@misc{pith2026250909195,
  author       = {Pith},
  title        = {Pith review of: Breaking the Statistical Similarity Trap in Extreme Convection Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UJ4SNR67}},
  note         = {Machine review of arXiv:2509.09195}
}
read the original abstract

Current evaluation metrics for deep learning weather models create a "Statistical Similarity Trap", rewarding blurry predictions while missing rare, high-impact events. We provide quantitative evidence of this trap, showing sophisticated baselines achieve 97.9% correlation yet 0.00 CSI for dangerous convection detection. We introduce DART (Dual Architecture for Regression Tasks), a framework addressing the challenge of transforming coarse atmospheric forecasts into high-resolution satellite brightness temperature fields optimized for extreme convection detection (below 220 K). DART employs dual-decoder architecture with explicit background/extreme decomposition, physically motivated oversampling, and task-specific loss functions. We present four key findings: (1) empirical validation of the Statistical Similarity Trap across multiple sophisticated baselines; (2) the "IVT Paradox", removing Integrated Water Vapor Transport, widely regarded as essential for atmospheric river analysis, improves extreme convection detection by 270%; (3) architectural necessity demonstrated through operational flexibility (DART achieves CSI = 0.273 with bias = 2.52 vs. 6.72 for baselines at equivalent CSI), and (4) real-world validation with the August 2023 Chittagong flooding disaster as a case study. To our knowledge, this is the first work to systematically address this hybrid conversion-segmentation-downscaling task, with no direct prior benchmarks identified in existing literature. Our validation against diverse statistical and deep learning baselines sufficiently demonstrates DART's specialized design. The framework enables precise operational calibration through beta-tuning, trains in under 10 minutes on standard hardware, and integrates seamlessly with existing meteorological workflows, demonstrating a pathway toward trustworthy AI for extreme weather preparedness.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

70 extracted references · 2 canonical work pages

  1. [55]

    Earth and Space Science 10(5), 2022–002651 (2023)

    Johnson, A., Wang, X., Blake, B.T., Rogers, E., Wang, Y., Carley, J.R., Clark, A., Beck, J., Alexander, C.: Verification and model configuration sensitivity of simulated abi radiance forecasts with the fv3-lam model. Earth and Space Science 10(5), 2022–002651 (2023)

    2022

  2. [56]

    Journal of Applied Meteo- rology and Climatology47(2), 525–543 (2008)

    Martin, D.W., Kohrs, R.A., Mosher, F.R., Medaglia, C.M., Adamo, C.: Over- ocean validation of the global convective diagnostic. Journal of Applied Meteo- rology and Climatology47(2), 525–543 (2008)

    2008

  3. [1]

    Nature619(7970), 533–538 (2023) https://doi.org/10.1038/s41586-023-06185-3

    Bi, K., Xie, L., Zhang, H., Chen, X., Gu, X., Tian, Q.: Accurate medium-range global weather forecasting with 3d neural networks. Nature619(7970), 533–538 (2023) https://doi.org/10.1038/s41586-023-06185-3

    doi:10.1038/s41586-023-06185-3 2023

  4. [2]

    Science382(6677), 1416–1421 (2023) 37

    Lam, R., Sanchez-Gonzalez, A., Willson, M., Wirnsberger, P., Fortunato, M., Alet, F., Ravuri, S., Ewalds, T., Eaton-Rosen, Z., Hu, W.,et al.: Learning skillful medium-range global weather forecasting. Science382(6677), 1416–1421 (2023) 37

    2023

  5. [3]

    arXiv preprint arXiv:2202.11214 (2022)

    Pathak, J., Subramanian, S., Harrington, P., Raja, S., Chattopadhyay, A., Mardani, M., Kurth, T., Hall, D., Li, Z., Azizzadenesheli, K., Hassanzadeh, P., Kashinath, K., Anandkumar, A.: Fourcastnet: A global data-driven high- resolution weather model using adaptive fourier neural operators. arXiv preprint arXiv:2202.11214 (2022)

    arXiv 2022

  6. [4]

    Quarterly Journal of the Royal Meteorological Society146(730), 1999– 2049 (2020) https://doi.org/10.1002/qj.3803

    Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Hor´ anyi, A., Mu˜ noz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D.,et al.: The ERA5 global reanalysis. Quarterly Journal of the Royal Meteorological Society146(730), 1999– 2049 (2020) https://doi.org/10.1002/qj.3803

    doi:10.1002/qj.3803 1999

  7. [5]

    Technical Report ERL ESG-15, Environmental Research Laboratories, Boulder, CO (1985)

    Doswell, C.A.: The operational meteorology of convective weather, volume 2: Storm scale analysis. Technical Report ERL ESG-15, Environmental Research Laboratories, Boulder, CO (1985)

    1985

  8. [6]

    Geophysical Research Letters48(6), 2020–092261 (2021)

    Proud, S.R., Bachmeier, S.: Record-low cloud temperatures associated with a tropical deep convective event. Geophysical Research Letters48(6), 2020–092261 (2021)

    2020

Show all 70 references
  1. [7]

    The Penn- sylvania State University

    The Pennsylvania State University: Observing Weather from Space. The Penn- sylvania State University. https://learningweather.psu.edu/node/46 Accessed 2025-09-06

    2025

  2. [8]

    World Meteorological Organization

    World Meteorological Organization: Guidelines for Satellite-based Nowcasting in Africa. World Meteorological Organization. WMO-No. 1309. https://cdn.knmi. nl/system/data center publications/files/000/071/978/original/1309 en.pdf Accessed 2025-09-06

    2025

  3. [9]

    World Meteorological Organization

    WMO-CGMS VLab: Guidelines on Satellite Skills and Knowledge for Opera- tional Meteorologists. World Meteorological Organization. https://rammb2.cira. colostate.edu/wp-content/uploads/2025/06/Guidelines SatSkills 2018 en.pdf Accessed 2025-09-06

    2025

  4. [10]

    European Centre for Medium-Range Weather Forecasts

    ECMWF: Integrated Forecasting System. European Centre for Medium-Range Weather Forecasts. https://www.ecmwf.int/en/forecasts/ documentation-and-support/changes-ecmwf-model Accessed 2025-09-06

    2025

  5. [11]

    Journal of Hydrometeorology19, 409–423 (2018) https://doi.org/10.1175/JHM-D-17-0135.1

    DeFlorio, M.J., Shields, M.D., Ralph, F.M., Patricola, C.M.: Global assessment of atmospheric river prediction skill. Journal of Hydrometeorology19, 409–423 (2018) https://doi.org/10.1175/JHM-D-17-0135.1

    2018 doi

  6. [12]

    Natural Hazards and Earth System Sciences21(6), 1807–1823 (2021)

    Lee, D., Ahmadul, H., Patz, J., Block, P.: Predicting social and health vulnera- bility to floods in bangladesh. Natural Hazards and Earth System Sciences21(6), 1807–1823 (2021)

    2021

  7. [13]

    LSE Grantham Research Institute on Climate Change and 38 the Environment (2023)

    Letsch, L., Dasgupta, S., Robinson, E.: Tackling flooding in Bangladesh in a changing climate. LSE Grantham Research Institute on Climate Change and 38 the Environment (2023). https://www.lse.ac.uk/granthaminstitute/publication/ tackling-flooding-in-bangladesh-in-a-changing-climate/

    2023

  8. [14]

    Natural Hazards and Earth System Sciences 22(3), 729–751 (2022)

    Bernard, A., Long, N., Becker, M., Khan, J., Fanchette, S.: Bangladesh’s vulner- ability to cyclonic coastal flooding. Natural Hazards and Earth System Sciences 22(3), 729–751 (2022)

    2022

  9. [15]

    Artificial Intelligence for the Earth Systems4(1), 240033 (2025)

    Pasche, O.C., Wider, J., Zhang, Z., Zscheischler, J., Engelke, S.: Validating deep learning weather forecast models on recent high-impact extreme events. Artificial Intelligence for the Earth Systems4(1), 240033 (2025)

    2025

  10. [16]

    arXiv preprint arXiv:2406.04678 (2024)

    Kim, D., Seo, M., Choi, Y.: Ace metric: Advection and convection evaluation for accurate weather forecasting. arXiv preprint arXiv:2406.04678 (2024)

    2024 arXiv

  11. [17]

    Technical Report WMO/TD-No

    Stanski, H.R., Wilson, L.J., Burrows, W.R.: Survey of common verification meth- ods in meteorology. Technical Report WMO/TD-No. 358, World Meteorological Organisation (1989)

    1989

  12. [18]

    Earth Interactions15(28), 1–25 (2011)

    Pijanowski, B., Moore, N., Mauree, D., Niyogi, D.: Evaluating error propagation in coupled land–atmosphere models. Earth Interactions15(28), 1–25 (2011)

    2011

  13. [19]

    Meteorologische Zeitschrift21(1), 1–7 (1904)

    Bjerknes, V.: Das Problem der Wettervorhersage, betrachtet von Standpunkt der Mechanik und Physik. Meteorologische Zeitschrift21(1), 1–7 (1904)

    1904

  14. [20]

    Selz, T., Zanna, L., Bueso-Borr´ as, P., Al-Hassani, O.: Can artificial intelligence- based weather prediction models simulate the butterfly effect? Geophys- ical Research Letters50(23), 2023–105747 (2023) https://doi.org/10.1029/ 2023GL105747

    2023

  15. [21]

    part ii: Daily pre- cipitation

    Sha, Y., Gagne II, D.J., West, G., Stull, R.: Deep-learning-based gridded down- scaling of surface meteorological variables in complex terrain. part ii: Daily pre- cipitation. Journal of Applied Meteorology and Climatology59(12), 2075–2092 (2020)

  16. [22]

    Scientific Reports14(1), 17748 (2024)

    Guo, Q., He, Z., Wang, Z.: Monthly climate prediction using deep convolutional neural network and long short-term memory. Scientific Reports14(1), 17748 (2024)

    2024

  17. [23]

    Atmosphere 14(4), 708 (2023)

    Fouotsa Manfouo, N.C., Potgieter, L., Watson, A., Nel, J.H.: A comparison of the statistical downscaling and long-short-term-memory artificial neural network models for long-term temperature and precipitations forecasting. Atmosphere 14(4), 708 (2023)

    2023

  18. [24]

    CoRR (2024) 39

    Saleem, H., Salim, F.D., Purcell, C.: Conformer: Embedding continuous attention in vision transformer for weather forecasting. CoRR (2024) 39

    2024

  19. [25]

    Advances in Neural Information Processing Systems37, 68740–68771 (2024)

    Nguyen, T., Shah, R., Bansal, H., Arcomano, T., Maulik, R., Kotamarthi, R., Fos- ter, I., Madireddy, S., Grover, A.: Scaling transformer neural networks for skillful and reliable medium-range weather forecasting. Advances in Neural Information Processing Systems37, 68740–68771 (2024)

    2024

  20. [26]

    Communications Earth & Envi- ronment1(1), 39 (2020)

    Lavers, D.A., Ralph, F.M., Richardson, D.S., Pappenberger, F.: Improved fore- casts of atmospheric rivers through systematic reconnaissance, better modelling, and insights on conversion of rain to flooding. Communications Earth & Envi- ronment1(1), 39 (2020)

    2020

  21. [27]

    Geophysical Research Letters51(2), 2023–106602 (2024)

    Zhang, Y., Chen, X., Stensrud, D.J., Clothiaux, E.E.: Enhancing severe weather prediction with microwave all-sky radiance assimilation: The 10 august 2020 midwest derecho. Geophysical Research Letters51(2), 2023–106602 (2024)

    2020

  22. [28]

    ACM computing surveys (CSUR)49(2), 1–50 (2016)

    Branco, P., Torgo, L., Ribeiro, R.P.: A survey of predictive modeling on imbalanced domains. ACM computing surveys (CSUR)49(2), 1–50 (2016)

    2016

  23. [29]

    Artificial Intelligence Review 57(6), 137 (2024)

    Chen, W., Yang, K., Yu, Z., Shi, Y., Chen, C.P.: A survey on imbalanced learning: latest research, applications and future directions. Artificial Intelligence Review 57(6), 137 (2024)

    2024

  24. [30]

    In: First International Workshop on Learning with Imbalanced Domains: Theory and Applications, pp

    Branco, P., Torgo, L., Ribeiro, R.P.: Smogn: a pre-processing approach for imbal- anced regression. In: First International Workshop on Learning with Imbalanced Domains: Theory and Applications, pp. 36–50 (2017). PMLR

    2017

  25. [31]

    In: ECAI, vol

    Kukar, M., Kononenko, I.,et al.: Cost-sensitive learning with neural networks. In: ECAI, vol. 15, pp. 88–94 (1998). Citeseer

    1998

  26. [32]

    Advances in Neural Information Processing Systems31(2018)

    Zhang, L., Zhou, Z.-H.: L1-regression with heavy-tailed distributions. Advances in Neural Information Processing Systems31(2018)

    2018

  27. [33]

    Srinivasan, V., Prasad, A., Balakrishnan, S., Ravikumar, P.K.: Efficient estima- tors for heavy-tailed machine learning (2021)

    2021

  28. [34]

    Journal of Advances in Modeling Earth Systems16(6), 2023–004019 (2024)

    Rasp, S., Hoyer, S., Merose, A., Langmore, I., Battaglia, P., Russell, T., Sanchez- Gonzalez, A., Yang, V., Carver, R., Agrawal, S.,et al.: Weatherbench 2: A benchmark for the next generation of data-driven global weather models. Journal of Advances in Modeling Earth Systems16...

    2023

  29. [35]

    Quarterly Journal of the Royal Meteorological Society 150(760), 1267–1287 (2024)

    Fahad, A.A., Hasan, M., Sharmili, N., Islam, S., Swenson, E.T., Roxy, M.: Climate change quadruples flood-causing extreme monsoon rainfall events in bangladesh and northeast india. Quarterly Journal of the Royal Meteorological Society 150(760), 1267–1287 (2024)

    2024

  30. [36]

    International Journal of Climatology38(2), 1010–1021 (2018) 40

    Yan, Y., Zhao, T., Ni, G., Sun, T.: Atmospheric rivers over the bay of bengal lead to northern indian extreme rainfall. International Journal of Climatology38(2), 1010–1021 (2018) 40

    2018

  31. [37]

    Natural Hazards28, 35–64 (2003) https://doi.org/10.1023/A: 1021169731325

    Mirza, M.: Three recent extreme floods in bangladesh: A hydro-meteorological analysis. Natural Hazards28, 35–64 (2003) https://doi.org/10.1023/A: 1021169731325

    2003 doi

  32. [38]

    Accessed: 2025-09-07

    Hussain, M.S.: Alluvial Soil. Accessed: 2025-09-07. https://en.banglapedia.org/ index.php?title=Alluvial Soil

    2025

  33. [39]

    EBSCO Research Starters

    Ferreira, F.: Bangladesh and sea-level rise. EBSCO Research Starters. Accessed: September 7, 2025 (2024). https://www.ebsco.com/research-starters/ environmental-sciences/bangladesh-and-sea-level-rise

    2025

  34. [40]

    Monthly Weather Review149(12), 3897–3921 (2021)

    Lagerquist, R., Stewart, J.Q., Ebert-Uphoff, I., Kumler, C.: Using deep learning to nowcast the spatial coverage of convection from himawari-8 satellite data. Monthly Weather Review149(12), 3897–3921 (2021)

    2021

  35. [41]

    Atmosphere15(3), 243 (2024)

    Li, Y., Shi, X., Deng, G., Li, X., Sun, F., Zhang, Y., Qin, D.: Dugs-unet: A novel deep semantic segmentation approach to convection detection based on fy-4a geostationary meteorological satellite. Atmosphere15(3), 243 (2024)

    2024

  36. [42]

    Environmental Data Science2, 31 (2023)

    Bansal, A.S., Lee, Y., Hilburn, K., Ebert-Uphoff, I.: Leveraging spatiotemporal information in meteorological image sequences: From feature engineering to neural networks. Environmental Data Science2, 31 (2023)

    2023

  37. [43]

    Journal of Geophysical Research: Atmospheres120(24), 12514–12535 (2015)

    Guan, B., Waliser, D.E.: Detection of atmospheric rivers: Evaluation and appli- cation of an algorithm for global studies. Journal of Geophysical Research: Atmospheres120(24), 12514–12535 (2015)

    2015

  38. [44]

    Weather and Forecasting28(6), 1337–1352 (2013)

    Wick, G.A., Neiman, P.J., Ralph, F.M., Hamill, T.M.: Evaluation of forecasts of the water vapor signature of atmospheric rivers in operational numerical weather prediction models. Weather and Forecasting28(6), 1337–1352 (2013)

    2013

  39. [45]

    Journal of Geophysical Research: Atmospheres130(17), 2025–043896 (2025)

    Zavadoff, B.L., Cheng, K.-Y., Lopez, H., Kim, D., West, R., Gao, K., Chen, J.-H., Harris, L.M.: Western us subseasonal atmospheric river prediction in the global- nested gfdl shield model. Journal of Geophysical Research: Atmospheres130(17), 2025–043896 (2025)

    2025

  40. [46]

    https://doi

    GUAN, B.: [Data] Global Atmospheric Rivers Database, Version 3. https://doi. org/10.25346/S6/YO15ON . https://doi.org/10.25346/S6/YO15ON

    doi

  41. [47]

    Bulletin of the American Meteorological Society100(2), 269–289 (2019)

    Ralph, F.M., Rutz, J.J., Cordeira, J.M., Dettinger, M., Anderson, M., Reynolds, D., Schick, L.J., Smallcomb, C.: A scale to characterize the strength and impacts of atmospheric rivers. Bulletin of the American Meteorological Society100(2), 269–289 (2019)

    2019

  42. [48]

    Weather and forecasting11(4), 560–581 (1996) 41

    Doswell III, C.A., Brooks, H.E., Maddox, R.A.: Flash flood forecasting: An ingredients-based methodology. Weather and forecasting11(4), 560–581 (1996) 41

    1996

  43. [49]

    Monthly Weather Review127(12), 2977–2991 (1999)

    Sherwood, S.C.: Convective precursors and predictability in the tropical western pacific. Monthly Weather Review127(12), 2977–2991 (1999)

    1999

  44. [50]

    International Geophysics, vol

    Holton, J.R., Hakim, G.J.: An Introduction to Dynamic Meteorology, 5th edn. International Geophysics, vol. 88. Academic Press, Waltham, MA (2012)

    2012

  45. [51]

    arXiv preprint arXiv:2110.03613 (2021)

    Motamedi, M., Sakharnykh, N., Kaldewey, T.: A data-centric approach for training deep neural networks with less data. arXiv preprint arXiv:2110.03613 (2021)

    2021 arXiv

  46. [52]

    Neural networks106, 249–259 (2018)

    Buda, M., Maki, A., Mazurowski, M.A.: A systematic study of the class imbalance problem in convolutional neural networks. Neural networks106, 249–259 (2018)

    2018

  47. [53]

    Journal of artificial intelligence research16, 321–357 (2002)

    Chawla, N.V., Bowyer, K.W., Hall, L.O., Kegelmeyer, W.P.: Smote: synthetic minority over-sampling technique. Journal of artificial intelligence research16, 321–357 (2002)

    2002

  48. [54]

    Proceedings of the national academy of sciences115(39), 9684–9689 (2018)

    Rasp, S., Pritchard, M.S., Gentine, P.: Deep learning to represent subgrid pro- cesses in climate models. Proceedings of the national academy of sciences115(39), 9684–9689 (2018)

    2018

  49. [57]

    Weather and Forecasting34(5), 1395–1416 (2019)

    Potvin, C.K., Carley, J.R., Clark, A.J., Wicker, L.J., Skinner, P.S., Reinhart, A.E., Gallo, B.T., Kain, J.S., Romine, G.S., Aligo, E.A.,et al.: Systematic com- parison of convection-allowing models during the 2017 noaa hwt spring forecasting experiment. Weather and Forecastin...

    2017

  50. [58]

    arXiv preprint arXiv:2404.17752 (2024)

    Watt, R.A., Mansfield, L.A.: Generative diffusion-based downscaling for climate. arXiv preprint arXiv:2404.17752 (2024)

    2024 arXiv

  51. [59]

    Geoscientific Model Development17(21), 7915–7962 (2024)

    Olivetti, L., Messori, G.: Do data-driven models beat numerical models in forecasting weather extremes? a comparison of ifs hres, pangu-weather, and graphcast. Geoscientific Model Development17(21), 7915–7962 (2024)

    2024

  52. [60]

    Applied Sciences13(21), 12019 (2023)

    Chen, L., Han, B., Wang, X., Zhao, J., Yang, W., Yang, Z.: Machine learning methods in weather and climate applications: A survey. Applied Sciences13(21), 12019 (2023)

    2023

  53. [61]

    Atmospheric Environment338, 120797 (2024)

    Yang, R., Hu, J., Li, Z., Mu, J., Yu, T., Xia, J., Li, X., Dasgupta, A., Xiong, 42 H.: Interpretable machine learning for weather and climate prediction: A review. Atmospheric Environment338, 120797 (2024)

    2024

  54. [62]

    arXiv preprint arXiv:2501.06907 (2025)

    Shi, J., Shirali, A., Jin, B., Zhou, S., Hu, W., Rangaraj, R., Wang, S., Han, J., Wang, Z., Lall, U., et al.: Deep learning and foundation models for weather prediction: A survey. arXiv preprint arXiv:2501.06907 (2025)

    2025 arXiv

  55. [63]

    https://www.ecmwf.int/en/forecasts/documentation-and-support/ changes-ecmwf-model (2025)

    European Centre for Medium-Range Weather Forecasts: Changes to the ECMWF Model. https://www.ecmwf.int/en/forecasts/documentation-and-support/ changes-ecmwf-model (2025)

    2025

  56. [64]

    Atmosphere33(4), 399–411 (2023)

    Kim, G., Back, S.-Y., Kwon, Y., Son, S.-W.: Comparison of atmospheric river detection algorithms in east asia. Atmosphere33(4), 399–411 (2023)

    2023

  57. [65]

    Nature637(8044), 84–90 (2025)

    Price, I., Sanchez-Gonzalez, A., Alet, F., Andersson, T.R., El-Kadi, A., Masters, D., Ewalds, T., Stott, J., Mohamed, S., Battaglia, P.,et al.: Probabilistic weather forecasting with machine learning. Nature637(8044), 84–90 (2025)

    2025

  58. [66]

    Artificial Intelligence for the Earth Systems (2025)

    B¨ ulte, C., Horat, N., Quinting, J., Lerch, S.: Uncertainty quantification for data- driven weather models. Artificial Intelligence for the Earth Systems (2025)

    2025

  59. [67]

    arXiv preprint arXiv:2412.14048 (2024)

    Khot, A., Luo, X., Kagawa, A., Yoo, S.: Evidential deep learning for probabilistic modelling of extreme storm events. arXiv preprint arXiv:2412.14048 (2024)

    2024 arXiv

  60. [68]

    European Centre for Medium- Range Weather Forecasts

    ECMWF: Medium Range Ensemble Forecasts. European Centre for Medium- Range Weather Forecasts. https://confluence.ecmwf.int/display/FUG/Section+ 2.1.2.1+Medium+Range+Ensemble+forecasts Accessed 2025-09-09

    2025

  61. [69]

    Sensors (Basel, Switzerland)24(23), 7837 (2024)

    Wu, R., Liang, Y., Lin, L., Zhang, Z.: Spatiotemporal multivariate weather pre- diction network based on cnn-transformer. Sensors (Basel, Switzerland)24(23), 7837 (2024)

    2024

  62. [70]

    IEEE Access11, 135979–135993 (2023) 43

    Shi, L., Zhao, Y.: Edge detection of high-resolution remote sensing image based on multi-directional improved sobel operator. IEEE Access11, 135979–135993 (2023) 43

    2023

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.