Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

WxC-Bench: A Novel Dataset for Weather and Climate Downstream Tasks

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read WxC-Bench packages six ML-ready weather and climate datasets, from turbulence labels to forecast text, and validates each with a baseline model.

desk verdict Useful public resource of six weather/climate ML tasks with real artifacts; the GW flux labels rest on an unvalidated spectral filter and the validation baselines are uneven, so value the resource but discount the claims. read the letter →

arxiv 2412.02780 v1 pith:LYUQCDGN submitted 2024-12-03 cs.LG cs.AI

Rajat Shinde , Christopher E. Phillips , Kumar Ankur , Aman Gupta , Simon Pfreundschuh , Sujit Roy , Sheyenne Kirkland , Vishal Gaur
show 5 more authors
Amy Lin Aditi Sheshadri Udaysankar Nair Manil Maskey Rahul Ramachandran
This is my paper · ORCID
classification cs.LGcs.AI
keywords WxC-BenchweatherandclimatefoundationmodelsML-readydatasetgravitywaveparameterizationaviationturbulencepredictionsubseasonalprecipitationforecastinghurricanetrackintensitynatural-languageforecastgeneration
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

WxC-Bench is a collection of six curated, machine-learning-ready datasets for weather and climate research, each formulated as an ML task: aviation turbulence classification, gravity wave momentum flux regression, weather analog search, subseasonal precipitation forecasting, hurricane track and intensity prediction, and natural-language forecast report generation. The paper's claim is that, together, these datasets span atmospheric scales from meso-$\beta$ (about $20$--$200$ km) to synoptic scales (about $2500$ km) and multiple data modalities, so they can support training and benchmarking of generalizable weather and climate AI models rather than a single forecasting model. The datasets are released openly with the code used to prepare them, and each is checked with a baseline model in the technical validation section. If the dataset labels are faithful, researchers can use WxC-Bench as a common testbed for foundation-model transfer across tasks that previously had no shared benchmark.

What carries the argument

The load-bearing mechanism is the conversion of heterogeneous raw sources into standardized input–label pairs for six ML formulations. For each task, the paper fixes a data source, a preprocessing recipe, and a label construction rule: turbulence labels come from pilot reports binned to MERRA-2 grid cells with more than 25% moderate-or-greater reports; gravity wave fluxes come from a Helmholtz decomposition of ERA5 winds into rotational and divergent parts, removal of the first 21 harmonics, and formation of the products of divergent winds with eddy vertical velocity; analog search uses overlapping 1800 km subgrids of MERRA-2 fields; precipitation uses geostationary, polar-orbiting, and microwave satellite observations regridded to 0.625° by 0.5° with PERSIANN-CDR and IMERG estimates as targets; hurricanes use HURDAT best tracks cubically interpolated to 3-hourly intervals; and forecast reports pair analysis stacks with processed forecast discussions. The pipeline also applies per-task normalization, such as scaling winds by three standard deviations and applying a cube root to gravity wave fluxes, so that each dataset is ML-ready as published.

What would settle it

Compare the WxC-Bench gravity wave momentum flux labels to fluxes diagnosed from a storm-resolving global simulation with grid spacing near 2–4 km on the same dates and grid; if the field correlation is near zero or the magnitudes are systematically off, the label construction does not support learning the intended physics.

Watch

Extended reading notes

Core claim

The central discovery is that a single publicly released suite of six preprocessed datasets can represent weather and climate phenomena at very different scales and in very different ML formats, and that baseline models can learn from each of them. The six tasks are intended to sample the space of downstream problems: classification of aviation turbulence from MERRA-2 atmospheric profiles, regression of subgrid gravity wave momentum fluxes from ERA5 background states, similarity search over weather analogs encoded as subgrids, autoregressive forecasting of daily precipitation four weeks ahead from four decades of satellite observations, hurricane track and intensity forecasting from HURDAT and MERRA-2, and generation of textual weather discussions conditioned on analysis maps. The validation shows each dataset is learnable: an ANN detects turbulence with 80% overall accuracy, an attention-based convolutional network reproduces global gravity wave flux patterns with $R^2$ up to 0.6 in the midlatitudes, a convolutional encoder-decoder retrieves analogous past weather states, an autoregressive CNN competes with operational subseasonal-to-seasonal models beyond ten-day lead, a Fourier-neural-operator forecast model tracks hurricanes with small track errors, and a vision-language model generates reports that match key forecast wording. The intended consequence is that weather and climate foundation models can be evaluated on one multi-modal, multi-scale benchmark instead of being trained and scored separately for each application.

Load-bearing premise

The load-bearing premise is that the residual of ERA5's divergent wind after removing the first 21 harmonics really represents gravity wave momentum fluxes; if that residual is dominated by numerical noise or balanced flow, the gravity wave regression targets are biased.

Editorial extensions

If this is right

  • A model trained or fine-tuned on WxC-Bench can be scored on six tasks at once, giving a direct measure of transfer across spatial scales and data modalities.
  • The gravity wave dataset provides a global, multi-year regression target that can be used to train ML parameterizations, which could later be coupled into coarse climate models.
  • The precipitation benchmark's four-decade record of satellite inputs and precipitation references allows training and evaluation of subseasonal forecasts against operational NWP baselines at lead times beyond ten days.
  • The hurricane dataset merges Atlantic and Pacific best-track records from 1980 to 2022, so models can be tested for cross-basin generalization rather than only in a single basin.
  • The natural-language task offers text labels that make it possible to train or fine-tune vision-language models on weather report generation, a step toward automatically communicating forecasts.

Reading between the lines

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

  • Editorial inference: because the gravity wave labels are computed from ERA5's resolved divergent flow, any systematic deficiency in ERA5's representation of mesoscale gravity waves will be baked into the training targets; comparing these labels against flux estimates from storm-resolving simulations on overlapping dates would quantify that bias.
  • Editorial inference: the analog-search subgrid encoding, which stores each 1800 km tile with its location, could be reused as a general retrieval interface over other reanalysis products or extended to multi-variable queries, but the paper only demonstrates single-variable lookups.
  • Editorial inference: the natural-language weather report task is framed as caption generation conditioned on an analysis map, so the benchmark does not yet test whether a model could generate a forecast discussion from a predicted future state; connecting WxC-Bench's forecast tasks to the text labels would close that loop.
  • Editorial inference: because all tasks are aligned to common reanalysis-era grids and formats, the same pretrained embedding or foundation model could be probed for zero-shot performance across the suite, which would give a cheap signal about where transfer learning fails.
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 / 6 minor

Summary. The paper introduces WxC-Bench, a collection of six ML-ready datasets for weather and climate downstream tasks: aviation turbulence detection from MERRA-2 and PIREPs, gravity wave momentum flux regression from ERA5, weather analog search over MERRA-2 subgrids, long-range precipitation forecasting from satellite observations, hurricane track and intensity data from HURDAT with MERRA-2 atmospheric states, and natural-language forecast report generation from HRRR and SPC discussion text. For each task, the authors describe the data sources, preprocessing steps, and a baseline model with quantitative or qualitative validation. The datasets and generation code are released publicly on Hugging Face and GitHub.

Significance. If the data curation is sound and the label definitions are revised, WxC-Bench would be a useful multi-modal benchmark resource for weather and climate foundation models. Its strengths include public release of data and code, six tasks spanning meso-β to synoptic scales, and a diversity of ML modalities (classification, regression, retrieval, and captioning). The paper is appropriately cautious in presenting baselines as technical validation rather than state-of-the-art results. The principal value lies in the assembled data and open release, which lower the barrier for future evaluation of generalist weather and climate AI models.

major comments (3)
  1. [Gravity Wave (GW) Parameterization, Data Description] The regression labels are defined as (Fu,Fv) = (udiv * omega', vdiv * omega') from ERA5 divergent winds with the first 21 spherical harmonics removed, then conservatively coarse-grained to 2.8 degrees. The manuscript calls these both 'resolved GW fluxes' and 'subgrid-scale momentum fluxes' in the same paragraph. Because the T21 residual contains divergent scales from roughly 1800 km down to the ERA5 effective resolution (~150-200 km), and the 2.8-degree grid resolves scales above roughly 300 km, the coarse-grained flux is largely a resolved flux rather than a subgrid-scale parameterization target. No independent validation against observations, storm-resolving simulations, or other reanalyses is provided, and the 21-harmonic cutoff is not justified. This is load-bearing for the GW task, which is presented as a central contribution; please add sensitivity analyses to the spectral cutoff, compare against independent GW flux estimates, or explicitly relabel the targets as resolved GW fluxes and remove the subgrid-scale claim.
  2. [Long-range Precipitation Forecasting, Figure 14] The ML baseline is trained on precipitation estimates from PERSIANN-CDR and IMERG and then evaluated against the IMERG Final product, as the text acknowledges: 'expected given that the precipitation estimates it was trained on were derived from the same precipitation product that is used for the evaluation.' This makes the comparison with the ECMWF/UKMO S2S baselines in Figure 14 circular for the ML model; the claimed lower bias and higher correlation at later lead times are in part an artifact of training on the reference. The dataset can still be useful, but the technical validation should either use a genuinely independent reference (e.g., gauge-based products or a withheld IMERG period) or explicitly frame the ML curve as an in-distribution sanity check rather than a skill comparison.
  3. [Technical Validation, Generation of Natural Language-based Weather Forecast Reports, Table 7] Table 7 reports ROUGE-L scores for only three dates (2017-03-30, 2018-08-03, 2019-08-30), with no information on the test-set size, no variance estimates, and no comparison to a trivial baseline. This is insufficient to support the claim that the dataset is ML-ready for end-to-end weather report generation. At minimum, report scores over the full held-out set and include variance or confidence intervals, or restrict the claim to a proof-of-concept.
minor comments (6)
  1. [Gravity Wave (GW) Parameterization, Data Description] The text states that ERA5 is 'publicly available at a horizontal resolution of 0.3 by 0.3 degrees,' but ERA5 is on a 0.25-degree grid (about 31 km), as the later '30 km' phrase implies; please correct this.
  2. [Aviation Turbulence Prediction, Dataset Description] The label definition 'any cell with MODG report frequency more than 25%' is an arbitrary threshold; please report sensitivity to this threshold or provide a citation to prior usage.
  3. [Weather Analog Search, Figure 11] The text says that 'the first four images retrieved by the similarity search have overall SSIM scores greater than 0.5, while the fourth and fifth images have scores less than 0.2'; the fourth image cannot satisfy both statements, so please correct the figure or the sentence.
  4. [Gravity Wave (GW) Parameterization, Data Description] The phrase 'the dataset comprises of a total of 64x128x24x1461 (~287 million) columns' would be clearer as 'samples' or 'grid-point-time columns'.
  5. [Natural Language-based Weather Forecast Reports, Table 5] Since Table 5 appears in the natural-language forecasting section, the first column should specify that the 'Number of samples' refers to HRRR-analysis/report pairs, and the caption should give the full date range and dataset name.
  6. [Hurricane Forecasting based on FourCastNet] The validation is a single-event case study (Hurricane Michael); please clarify in the text that this is illustrative and not a dataset-level benchmark.

Circularity Check

1 steps flagged · score 3.0 of 10

One disclosed self-referential precipitation baseline; dataset construction and other validations are otherwise self-contained.

  1. fitted input called prediction [Technical Validation > Long-range Precipitation Forecasting (discussion of Fig. 14)]
    "The ML baseline forecasts is closest to the reference data, which is expected given that the precipitation estimates it was trained on were derived from the same precipitation product that is used for the evaluation."

    The ML baseline for long-range precipitation is trained with IMERG/PERSIANN-CDR precipitation estimates as regression targets, and then its skill is evaluated against the same IMERG Final product used for the reference. The paper itself concedes that being 'closest to the reference' is expected because the training and evaluation products are the same. The comparison against ECMWF and UKMO NWP forecasts is therefore not an independent test of forecast quality: part of the baseline's apparent advantage is built into its training target. This is a disclosed limitation and affects only the precipitation validation, not the construction of the other five datasets.

full rationale

WxC-Bench is a dataset paper rather than a derivation paper: its central contribution is the curation and public release of six ML-ready datasets, and its technical validations demonstrate that each dataset can drive a baseline model. No fitted constants, uniqueness theorems, or ansatz-carrying self-citations are used to force the claimed results. The gravity-wave labels are constructed from ERA5 via Helmholtz decomposition and a 21-harmonic high-pass filter; this is a physically debatable label definition, and the paper explicitly cautions that ERA5 does not fully resolve GW scales, but it is not a circular reduction because the labels are not fitted from the input features or defined as the model output. The precipitation evaluation is the one genuine circular element: the baseline is trained on the same precipitation product family used as the reference, making the reported 'closest to the reference' result partly self-referential. The authors flag this explicitly, and it concerns a supporting validation rather than the core dataset claim, so it raises the circularity score only moderately.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central claim is a practical contribution (a public dataset collection), not a physical law, so the axiom ledger records the domain assumptions embedded in label construction rather than free physical parameters.

free parameters (4)
  • Turbulence positive label threshold = 25% MODG PIREPs per cell-day
    Chosen to convert sparse point reports into binary labels; arbitrary threshold affects class balance and baseline scores.
  • GW high-pass filter cutoff = remove first 21 harmonics of divergent flow
    Chosen so that scales >150-200 km remain; no sensitivity analysis.
  • GW dataset selected years = 2010, 2012, 2014, 2015
    Chosen for rich wintertime orographic GW activity in the Southern Hemisphere; introduces seasonal and hemispheric sampling bias.
  • PIREP flight-level bins = LOW 0-14,900 ft; MIDDLE 15,000-29,900 ft; HIGH ≥30,000 ft
    Hand-selected to match PBL, free atmosphere, and tropopause; affects the three classification tasks.
assumptions (3)
  • domain assumption ERA5 at 0.3° resolves gravity waves with wavelengths >150-200 km
    Invoked in GW Data Description to justify computing fluxes from reanalysis rather than storm-resolving simulation; if false, labels are aliased.
  • domain assumption Helmholtz decomposition plus removal of first 21 harmonics isolates GW momentum fluxes
    Used to construct Fu and Fv labels; an approximation from the literature but unvalidated within this paper.
  • domain assumption PERSIANN-CDR and IMERG Final are sufficiently consistent to pool as training reference
    Training record concatenates PERSIANN-CDR (pre-2000) with IMERG (post-2000); the paper notes they differ but does not quantify the impact.

how reviews work

0 comments
Cite this review

Pith. "Pith review of WxC-Bench: A Novel Dataset for Weather and Climate Downstream Tasks." pith.science (2026). https://pith.science/paper/LYUQCDGN

@misc{pith2026241202780,
  author       = {Pith},
  title        = {Pith review of: WxC-Bench: A Novel Dataset for Weather and Climate Downstream Tasks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LYUQCDGN}},
  note         = {Machine review of arXiv:2412.02780}
}
abstract

High-quality machine learning (ML)-ready datasets play a foundational role in developing new artificial intelligence (AI) models or fine-tuning existing models for scientific applications such as weather and climate analysis. Unfortunately, despite the growing development of new deep learning models for weather and climate, there is a scarcity of curated, pre-processed machine learning (ML)-ready datasets. Curating such high-quality datasets for developing new models is challenging particularly because the modality of the input data varies significantly for different downstream tasks addressing different atmospheric scales (spatial and temporal). Here we introduce WxC-Bench (Weather and Climate Bench), a multi-modal dataset designed to support the development of generalizable AI models for downstream use-cases in weather and climate research. WxC-Bench is designed as a dataset of datasets for developing ML-models for a complex weather and climate system, addressing selected downstream tasks as machine learning phenomenon. WxC-Bench encompasses several atmospheric processes from meso-$\beta$ (20 - 200 km) scale to synoptic scales (2500 km), such as aviation turbulence, hurricane intensity and track monitoring, weather analog search, gravity wave parameterization, and natural language report generation. We provide a comprehensive description of the dataset and also present a technical validation for baseline analysis. The dataset and code to prepare the ML-ready data have been made publicly available on Hugging Face -- https://huggingface.co/datasets/nasa-impact/WxC-Bench

Figures

Figures reproduced from arXiv: 2412.02780 by the authors.

Figure 1
Figure 1. Illustration showing the applicability of the proposed WxC-Bench in the overall AI/ML workflows after converting raw non-ML ready datasets to ML ready datasets. The ML-ready datasets can then be directly used for training new AI models or fine-tuning pre-trained AI models. The WxC-Bench dataset can be used with AI models relevant to downstream applications across multiple spatial and temporal scales and resolutions.… view at source ↗
Figure 2
Figure 2. Spatial distribution of PIREPs turbulence reports over CONUS from 2003-Present. Note that values are scaled logarithmically. The PIREP reports were first grouped by flight levels into bins of 0-14, 900 ft (LOW), 15,000-29,900 ft (MIDDLE), and 30,000 ft or greater (HIGH). These levels were selected as they correspond approximately to the planetary boundary layer (PBL), free atmosphere, and tropopause respectively. Fl… view at source ↗
Figure 3
Figure 3. Distribution of the coarse-grained input features (scaled background atmospheric state) and output labels (scaled GW momentum fluxes) for a single month (January 2010) from ERA5. Data Description For a given background state of the atmosphere, GW parameterizations in climate models predict the associated subgrid-scale momentum fluxes due to GWs [38]. Here, this problem is formulated as a regression task to predict t… view at source ↗
Figures from the paper (19 more)
Figure 4
Figure 4. Figure 4: Illustration of training images for (a) sea level pressure (b) temperature at 2m, for January 01, 2019. The above-mentioned approaches may be handled simultaneously by breaking the larger grids into sub-grids and encoding each separately with location information. Then…
Figure 5
Figure 5. Figure 5: Example of the input data for the Long-range precipitation forecast task. Each panel displays observations from one of the channels of the three observational datasets that form the input for the Long-range precipitation forecast. Panels (a), (b), and (c) show visible …
Figure 6
Figure 6. Figure 6: Tracks of named hurricanes in the Atlantic Ocean by their respective categories, ranging from tropical storms (TS) to Category 5 (Cat 5) hurricanes for the year 2017 [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Hurricane counts by categories and years (1980-2022) for North Atlantic and East Pacific hurricanes. 10/28 [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: (a) Illustration of distribution of words in the forecast reports, (b) Illustration of training data for (b) surface temperature and (c) corresponding forecast report for December 31, 2020. Weather data are from the HRRR 21Z analysis on December 31, 2020. • Removing nu…
Figure 9
Figure 9. Figure 9: Monthly mean normalized (a) zonal and (c) meridional components of the GW fluxes at 200 hPa (upper troposphere) provided by WxC-Bench for May 2015. (b) and (d) respectively show the predicted zonal and meridional fluxes from a convolutional neural network trained on Wx…
Figure 10
Figure 10. Figure 10: Illustration of the baseline model architecture for deep learning based weather analog search [PITH_FULL_IMAGE:figures/full_fig_p014_10.png]
Figure 11
Figure 11. Figure 11: Illustration of the top-5 images for temperature along with their date of occurrence in the past, with their structural similarity score with respect to the query image. All components of the model, except the stems and head, consist of ResNeXt [84] blocks. The exact …
Figure 12
Figure 12. Figure 12: The architecture of the convolutional, auto-regressive forecast model used to implement the machine learning baseline model for the long-range precipitation forecast [PITH_FULL_IMAGE:figures/full_fig_p016_12.png]
Figure 13
Figure 13. Figure 13: Reference and predicted mean daily precipitation for forecasts initialized on 10 January 2019. The first row shows the satellite-based reference precipitation estimates from the IMERG Final product. Rows two and three show the corresponding forecasts from the control …
Figure 14
Figure 14. Figure 14: Accuracy of daily accumulated precipitation forecasts. Panel (a) shows the forecast bias defined as the area-weighted mean of the difference between predicted and reference precipitation. Panel (b) shows the area-weighted linear correlation coefficient. Panel (c) show…
Figure 15
Figure 15. Figure 15: Time series of the correlation and root mean square error of wind speed (m/s), SLP (Pa), T2M (K) and Z500 (m) for the hurricane Michael (2018) over the hurricane Michael domain (105◦W, 60◦W, 20◦N, 50◦N ). To validate this task, FourCastNet has been used as the baselin…
Figure 16
Figure 16. Figure 16: Spatial distribution of sea level pressure (Pa) for Hurricane Michael (2018) for 24H, 48H and 72H forecast from the ground truth, prediction, and the difference between the prediction and the ground truth. Fourier domain using Fast Fourier Transforms (FFTs). This inno…
Figure 17
Figure 17. Figure 17: Illustration of high-level architecture for generating natural language-based weather forecast captions. The captions are generated from an encoder-decoder architecture, pre-trained in vision and language models, and fine-tuned to the proposed dataset [PITH_FULL_IMAG…
Figure 18
Figure 18. Figure 18: Results show the performance of the natural language-based weather forecasting trained on the proposed dataset. The top row shows an illustration of three channels stacked from temperature, u, and v components of the wind; the second row shows the generated caption, w…
Figure 19
Figure 19. Figure 19: Illustration of (a) Box plots representing the count of sentences, words, stop words, and characters in the forecast report; (b) Histograms representing the distribution of text in the labels. • Count and density distribution of characters, words, and stop words [PIT…
Figure 20
Figure 20. Figure 20: Density distribution of words with respect to part of speech (POS). (ADJ: adjective, ADP: adpositions, ADV: adverb, CONJ, conjunctions, DET: determiners and NOUN: nouns.) 27/28 [PITH_FULL_IMAGE:figures/full_fig_p027_20.png]
Figure 21
Figure 21. Figure 21: Illustration of histogram for top-10 stop words from the forecast reports. • Grammatical categorization and POS tagging: Grammatical or POS tagging maps a word with a corresponding part of speech (nouns, adjective, conjunction, etc.) and plays a key role in understand…
Figure 22
Figure 22. Figure 22: shows the loss curve for fine-tuning the captioning model with the number of steps [PITH_FULL_IMAGE:figures/full_fig_p028_22.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. The Rise of AI in Weather and Climate Information and its Impact on Global Inequality

    physics.ao-ph 2026-03 conditional novelty 4.0 of 10

    AI weather and climate tools inherit Northern-controlled data and compute, risking worse forecasts and maladaptation for the Global South rather than democratizing climate information.

Reference graph

Works this paper leans on

90 extracted references · 37 canonical work pages · cited by 1 Pith paper

  1. [1]

    Smith, A. B. Us billion-dollar weather and climate disasters, 1980-present (2022)

    2022

  2. [2]

    Kurth, T. et al. Fourcastnet: Accelerating global high-resolution weather forecasting using adaptive fourier neural operators. In Proceedings of the platform for advanced scientific computing conference, 1–11 (2023)

    2023

  3. [3]

    Lam, R. et al. Graphcast: Learning skillful medium-range global weather forecasting. arXiv preprint arXiv:2212.12794 (2022)

    arXiv 2022

  4. [4]

    Chen, K. et al. Fengwu: Pushing the skillful global medium-range weather forecast beyond 10 days lead. arXiv preprint arXiv:2304.02948 (2023)

    arXiv 2023

  5. [5]

    Bi, K. et al. Pangu-weather: A 3d high-resolution model for fast and accurate global weather forecast. arXiv preprint arXiv:2211.02556 (2022)

    arXiv 2022

  6. [6]

    Sønderby, C. K. et al. Metnet: A neural weather model for precipitation forecasting. arXiv preprint arXiv:2003.12140 (2020)

    arXiv 2020

  7. [7]

    A., Durran, D

    Weyn, J. A., Durran, D. R., Caruana, R. & Cresswell-Clay, N. Sub-Seasonal Forecasting With a Large Ensemble of Deep-Learning Weather Prediction Models. J. Adv. Model. Earth Syst. 13, e2021MS002502, 10.1029/2021MS002502 (2021)

    doi:10.1029/2021ms002502 2021

  8. [8]

    Mansfield, L. A. et al. Updates on Model Hierarchies for Understanding and Simulating the Climate System: A Focus on Data-Informed Methods and Climate Change Impacts. J. Adv. Model. Earth Syst. 15, e2023MS003715, 10.1029/ 2023MS003715 (2023)

    2023

Show all 90 references
  1. [9]

    Rasp, S., Pritchard, M. S. & Gentine, P. Deep learning to represent subgrid processes in climate models. Proc. Natl. Acad. Sci. U.S.A. 115, 9684–9689, 10.1073/pnas.1810286115 (2018)

    2018 doi

  2. [10]

    & Bolton, T

    Zanna, L. & Bolton, T. Data-Driven Equation Discovery of Ocean Mesoscale Closures. Geophys. Res. Lett. 47, e2020GL088376, 10.1029/2020GL088376 (2020)

    2020 doi

  3. [11]

    I., Sheshadri, A., Cain, G

    Espinosa, Z. I., Sheshadri, A., Cain, G. R., Gerber, E. P. & DallaSanta, K. A Deep Learning Parameterization of Gravity Wave Drag Coupled to an Atmospheric Global Climate Model. Geophys. Res. Lett. (submitted, 2021)

    2021

  4. [12]

    & O’Gorman, P

    Wang, P., Yuval, J. & O’Gorman, P. A. Non-Local Parameterization of Atmospheric Subgrid Processes With Neural Networks. J. Adv. Model. Earth Syst. 14, e2022MS002984, 10.1029/2022MS002984 (2022)

    2022 doi

  5. [13]

    Bretherton, C. S. et al. Correcting Coarse-Grid Weather and Climate Models by Machine Learning From Global Storm- Resolving Simulations. J. Adv. Model. Earth Syst. 14, e2021MS002794, 10.1029/2021MS002794 (2022)

    2022 doi

  6. [14]

    Davenport, F. V . & Diffenbaugh, N. S. Using Machine Learning to Analyze Physical Causes of Climate Change: A Case Study of U.S. Midwest Extreme Precipitation. Geophys. Res. Lett. 48, e2021GL093787, 10.1029/2021GL093787 (2021)

    2021 doi

  7. [15]

    Diffenbaugh, N. S. & Barnes, E. A. Data-driven predictions of the time remaining until critical global warming thresholds are reached. Proc. Natl. Acad. Sci. 120, e2207183120, 10.1073/pnas.2207183120 (2023)

    2023 doi

  8. [16]

    Roy, S. et al. Clifford neural operators on atmospheric data influenced partial differential equations. In 12th International Conference on Learning Representations (2024)

    2024

  9. [17]

    Schmude, J. et al. Prithvi wxc: Foundation model for weather and climate (2024). 2409.13598

    2024 arXiv

  10. [18]

    Nguyen, T., Brandstetter, J., Kapoor, A., Gupta, J. K. & Grover, A. ClimaX: A foundation model for weather and climate. arXiv (2023). 2301.10343

    2023 arXiv

  11. [19]

    Bodnar, C. et al. Aurora: A foundation model of the atmosphere. arXiv preprint arXiv:2405.13063 (2024)

    2024 arXiv

  12. [20]

    Jakubik, J. et al. Foundation models for generalist geospatial artificial intelligence (2023). 2310.18660

    2023 arXiv

  13. [21]

    Bommasani, R. et al. On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258 (2021)

    2021 arXiv

  14. [22]

    Rasp, S. et al. Weatherbench: a benchmark data set for data-driven weather forecasting. J. Adv. Model. Earth Syst. 12, e2020MS002203 (2020). 21/28

    2020

  15. [23]

    Rasp, S. et al. Weatherbench 2: A benchmark for the next generation of data-driven global weather models (2024). 2308.15560

    2024 arXiv

  16. [24]

    Climatenet: An expert-labelled open dataset and deep learning architecture for enabling high-precision analyses of extreme weather

    Prabhat et al. Climatenet: An expert-labelled open dataset and deep learning architecture for enabling high-precision analyses of extreme weather. Geosci. Model. Dev. Discuss. 2020, 1–28 (2020)

    2020

  17. [25]

    Watson-Parris, D. et al. Climatebench: A benchmark dataset for data-driven climate projections. Authorea Prepr. (2022)

    2022

  18. [26]

    Yu, S. et al. Climsim: A large multi-scale dataset for hybrid physics-ml climate emulation. Adv. Neural Inf. Process. Syst. 36 (2024)

    2024

  19. [27]

    & Grover, A

    Nguyen, T., Jewik, J., Bansal, H., Sharma, P. & Grover, A. Climatelearn: Benchmarking machine learning for weather and climate modeling. Adv. Neural Inf. Process. Syst. 36 (2024)

    2024

  20. [28]

    Shinde, R. et al. Windset: Weather insights and novel data for systematic evaluation and testing. In The Twelfth International Conference on Learning Representations (ICLR) (2024)

    2024

  21. [29]

    & Yoshino, K

    Ito, J., Niino, H. & Yoshino, K. Large Eddy Simulation on Horizontal Convective Rolls that Caused an Aircraft Accident during its Landing at Narita Airport. Geophys. Res. Lett. 47, 10.1029/2020gl086999 (2020)

    2020 doi

  22. [30]

    Turbulence and Its Impact on Commercial Aviation

    Golding, W. Turbulence and Its Impact on Commercial Aviation. J. Aviat. Educ. & Res. 10.15394/jaaer.2002.1301 (2000)

    2000

  23. [31]

    Emara, M. et al. Machine Learning Enabled Turbulence Prediction using Flight Data for Safety Analysis. In 32nd Congress of the International Council of the Aeronautical Sciences (Shanghai, China, 2021)

    2021

  24. [32]

    Williams, J. K. Using random forests to diagnose aviation turbulence.Mach. Learn. 95, 51–70, 10.1007/s10994-013-5346-7 (2014)

    2014 doi

  25. [33]

    K., Ng, C

    Hon, K. K., Ng, C. W. & Chan, P. W. Machine learning based multi-index prediction of aviation turbulence over the Asia-Pacific. Mach. Learn. with Appl. 2, 100008, 10.1016/j.mlwa.2020.100008 (2020)

    2020

  26. [34]

    Gelaro, R. et al. The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2). J. Clim. 30, 5419–5454, 10.1175/jcli-d-16-0758.1 (2017)

    2017 doi

  27. [36]

    Wolff, J. K. & Sharman, R. D. Climatology of Upper-Level Turbulence over the Contiguous United States. J. Appl. Meteorol. Climatol. 47, 2198–2214, 10.1175/2008jamc1799.1 (2008)

    2008 doi

  28. [37]

    Fritts, D. C. & Alexander, M. J. Gravity wave dynamics and effects in the middle atmosphere. Rev. Geophys. 41, 10.1029/2001RG000106 (2003)

    2003 doi

  29. [38]

    Kim, Y .-J., Eckermann, S. D. & Chun, H.-Y . An overview of the past, present and future of gravity-wave drag parametriza- tion for numerical climate and weather prediction models. Atmosphere-Ocean 41, 65–98, 10.3137/ao.410105 (2003)

    2003 doi

  30. [39]

    Achatz, U. et al. Atmospheric Gravity Waves: Processes and Parameterization. J. Atmospheric Sci. -1, 10.1175/ JAS-D-23-0210.1 (2023)

    2023

  31. [40]

    Improving Non-local Parameterization of the Convective Boundary Layer

    Sorbjan, Z. Improving Non-local Parameterization of the Convective Boundary Layer. Boundary-Layer Meteorol 130, 57–69, 10.1007/s10546-008-9331-9 (2009)

    2009 doi

  32. [41]

    Chen, T.-C., Yau, M. K. & Kirshbaum, D. J. Assessment of Conditional Symmetric Instability from Global Reanalysis Data. J. Atmospheric Sci. 75, 2425–2443, 10.1175/JAS-D-17-0221.1 (2018)

    2018 doi

  33. [42]

    & Lott, F

    Plougonven, R., de la Cámara, A., Hertzog, A. & Lott, F. How does knowledge of atmospheric gravity waves guide their parameterizations? Q. J. Royal Meteorol. Soc. 146, 1529–1543, 10.1002/qj.3732 (2020)

    2020 doi

  34. [43]

    F., Shepherd, T

    McLandress, C., Scinocca, J. F., Shepherd, T. G., Reader, M. C. & Manney, G. L. Dynamical Control of the Mesosphere by Orographic and Nonorographic Gravity Wave Drag during the Extended Northern Winters of 2006 and 2009.J. Atmos. Sci. 70, 2152–2169, 10.1175/JAS-D-12-0297.1 (2012)

    2012 doi

  35. [44]

    & Hitchcock, P

    de la Cámara, A., Abalos, M. & Hitchcock, P. Changes in Stratospheric Transport and Mixing During Sudden Stratospheric Warmings. J. Geophys. Res. Atmospheres 123, 3356–3373, 10.1002/2017JD028007 (2018). 22/28

    2018 doi

  36. [45]

    Gupta, A. et al. Estimates of Southern Hemispheric Gravity Wave Momentum Fluxes Across Observations, Reanalyses, and Kilometer-scale Numerical Weather Prediction Model. J. Atmospheric Sci. -1, 10.1175/JAS-D-23-0095.1 (2024)

    2024 doi

  37. [46]

    Kruse, C. G. et al. Observed and Modeled Mountain Waves from the Surface to the Mesosphere near the Drake Passage. J. Atmospheric Sci. 79, 909–932, 10.1175/JAS-D-21-0252.1 (2022)

    2022 doi

  38. [47]

    S., Bölöni, G., Zängl, G

    Kim, Y .-H., V oelker, G. S., Bölöni, G., Zängl, G. & Achatz, U. Crucial Role of Obliquely Propagating Gravity Waves in Tropical Stratospheric Circulation (2023). 2309.15301

    2023 arXiv

  39. [48]

    Gupta, A., Sheshadri, A., Alexander, M. J. & Birner, T. Insights on Lateral Gravity Wave Propagation in the Extratropical Stratosphere from 44 Years of ERA5 Data. Geophys. Res. Lett. 10.1029/2024GL108541 (2024)

    2024 doi

  40. [49]

    A Helmholtz decomposition of structure functions and spectra calculated from aircraft data

    Lindborg, E. A Helmholtz decomposition of structure functions and spectra calculated from aircraft data. J. Fluid Mech. 762, R4, 10.1017/jfm.2014.685 (2015)

    2015 doi

  41. [50]

    Hersbach, H. et al. The ERA5 global reanalysis. Q. J. Royal Meteorol. Soc. 146, 1999–2049, 10.1002/qj.3803 (2020)

    2020 doi

  42. [51]

    Skamarock, W. C. Evaluating Mesoscale NWP Models Using Kinetic Energy Spectra.Mon. Weather. Rev.132, 3019–3032, 10.1175/MWR2830.1 (2004)

    2004 doi

  43. [52]

    & Polichtchouk, I

    Gupta, A., Birner, T., Dörnbrack, A. & Polichtchouk, I. Importance of Gravity Wave Forcing for Springtime Southern Polar V ortex Breakdown as Revealed by ERA5.Geophys. Res. Lett. 48, e2021GL092762, 10.1029/2021GL092762 (2021)

    2021 doi

  44. [53]

    A., Fu, Q., Wallace, J

    Pahlavan, H. A., Fu, Q., Wallace, J. M. & Kiladis, G. N. Revisiting the Quasi-Biennial Oscillation as Seen in ERA5. Part I: Description and Momentum Budget. J. Atmospheric Sci. 78, 673–691, 10.1175/JAS-D-20-0248.1 (2021)

    2021 doi

  45. [54]

    & Hassanzadeh, P

    Chattopadhyay, A., Nabizadeh, E. & Hassanzadeh, P. Analog forecasting of extreme-causing weather patterns using deep learning. J. Adv. Model. Earth Syst. 12, e2019MS001958 (2020)

    2020

  46. [55]

    A new look at weather forecasting through analogues

    Van den Dool, H. A new look at weather forecasting through analogues. Mon. weather review 117, 2230–2247 (1989)

    1989

  47. [56]

    & Alessandrini, S

    Yang, D. & Alessandrini, S. An ultra-fast way of searching weather analogs for renewable energy forecasting. Sol. Energy 185, 255–261 (2019)

    2019

  48. [57]

    & Furlanello, C

    Franch, G., Jurman, G., Coviello, L., Pendesini, M. & Furlanello, C. Mass-umap: Fast and accurate analog ensemble search in weather radar archives. Remote. Sens. 11, 2922 (2019)

    2019

  49. [58]

    Ahn, H. et al. Searching similar weather maps using convolutional autoencoder and satellite images. ICT Express 9, 69–75 (2023)

    2023

  50. [59]

    & Lawrence, B

    Raoult, B., Di Fatta, G., Pappenberger, F. & Lawrence, B. Fast retrieval of weather analogues in a multi-petabytes archive using wavelet-based fingerprints. In International Conference on Computational Science, 697–710 (Springer, 2018)

    2018

  51. [60]

    & Bruzzone, L

    Demir, B. & Bruzzone, L. Hashing-based scalable remote sensing image search and retrieval in large archives. IEEE transactions on geoscience remote sensing 54, 892–904 (2015)

    2015

  52. [61]

    & Delle Monache, L

    Hu, W., Cervone, G., Young, G. & Delle Monache, L. Machine learning weather analogs for near-surface variables. Boundary-Layer Meteorol. 186, 711–735 (2023)

    2023

  53. [62]

    Dool, H. M. v. d. A New Look at Weather Forecasting through Analogues. Mon. Weather. Rev. 117, 2230–2247, 10.1175/1520-0493(1989)117<2230:anlawf>2.0.co;2 (1989)

    1989 doi

  54. [63]

    Vitart, F. et al. Outcomes of the wmo prize challenge to improve subseasonal to seasonal predictions using artificial intelligence. Bull. Am. Meteorol. Soc. 103, E2878 – E2886, 10.1175/BAMS-D-22-0046.1 (2022)

    2022 doi

  55. [64]

    W., Vitart, F

    Robertson, A. W., Vitart, F. & Camargo, S. J. Subseasonal to seasonal prediction of weather to climate with application to tropical cyclones. J. Geophys. Res. Atmospheres 125, e2018JD029375, https://doi.org/10.1029/2018JD029375 (2020). E2018JD029375 10.1029/2018JD029375, https...

    2020 doi

  56. [65]

    Vitart, F., Robertson, A. W. & Anderson, D. L. Subseasonal to seasonal prediction project: Bridging the gap between weather and climate. Bull. World Meteorol. Organ. 61, 23 (2012). 23/28

    2012

  57. [66]

    de Witt, C. S. et al. Rainbench: towards global precipitation forecasting from satellite imagery. arXiv preprint arXiv:2012.09670 (2020)

    2020 arXiv

  58. [67]

    Knapp, K. R. et al. Globally gridded satellite observations for climate studies. Bull. Am. Meteorol. Soc. 92, 893 – 907, 10.1175/2011BAMS3039.1 (2011)

    2011 doi

  59. [68]

    J., Philips, C., Heidinger, A

    Foster, M. J., Philips, C., Heidinger, A. K. & Program, N. C. Noaa climate data record (cdr) of advanced very high resolution radiometer (avhrr) and high-resolution infra-red sounder (hirs) reflectance, brightness temperature, and cloud products from pathfinder atmospheres - e...

    2021 doi

  60. [69]

    D., Berg, W

    Kummerow, C. D., Berg, W. K., Kuo, C.-P., & Program, N. C. Noaa climate data record (cdr) of advanced very high resolution radiometer (avhrr) and high-resolution infra-red sounder (hirs) reflectance, brightness temperature, and cloud products from pathfinder atmospheres - exte...

    2022 doi

  61. [70]

    Ashouri, H. et al. Persiann-cdr: Daily precipitation climate data record from multisatellite observations for hydrological and climate studies. Bull. Am. Meteorol. Soc. 96, 69–83 (2015)

    2015

  62. [71]

    & Jackson, T

    Huffman, G., Stocker, E., Bolvin, D., Nelkin, E. & Jackson, T. Gpm imerg final precipitation l3 1 day 0.1 degree x 0.1 degree v07, greenbelt, md, goddard earth sciences data and information services center (ges disc).[accessed 2023 sep 14]. DOI: https://doi. org/10.5067/GPM/IM...

    2023 doi

  63. [72]

    Assessing the present and future probability of hurricane harvey’s rainfall

    Emanuel, K. Assessing the present and future probability of hurricane harvey’s rainfall. Proc. Natl. Acad. Sci. 114, 12681–12684 (2017)

    2017

  64. [73]

    Bromirski, P. D. & Kossin, J. P. Increasing hurricane wave power along the us atlantic and gulf coasts. J. Geophys. Res. Ocean. 113 (2008)

    2008

  65. [74]

    Increasing destructiveness of tropical cyclones over the past 30 years

    Emanuel, K. Increasing destructiveness of tropical cyclones over the past 30 years. Nature 436, 686–688 (2005)

    2005

  66. [75]

    & Lillo, S

    Burg, T. & Lillo, S. P. Tropycal: A python package for analyzing tropical cyclones and more. In 34th Conference on Hurricanes and Tropical Meteorology (AMS, 2021)

    2021

  67. [76]

    & Kittredge, R

    Goldberg, E., Driedger, N. & Kittredge, R. Using natural-language processing to produce weather forecasts. IEEE Expert. 9, 45–53, 10.1109/64.294135 (1994)

    1994 doi

  68. [77]

    J., Barro, S

    Ramos-Soto, A., Bugarin, A. J., Barro, S. & Taboada, J. Linguistic descriptions for automatic generation of textual short-term weather forecasts on real prediction data. IEEE Transactions on Fuzzy Syst. 23, 44–57 (2014)

    2014

  69. [78]

    & Davy, I

    Reiter, E., Sripada, S., Hunter, J., Yu, J. & Davy, I. Choosing words in computer-generated weather forecasts. Artif. Intell. 167, 137–169 (2005)

    2005

  70. [79]

    & Lv, Z.-L

    Zhang, H.-P., Wu, H.-P., Gao, J., Zhao, Y .-P. & Lv, Z.-L. Meteorological bulletin automatic generation based on spatio- temporal reasoning. In 2011 International Conference on Machine Learning and Cybernetics, vol. 4, 1927–1931 (IEEE, 2011)

    2011

  71. [80]

    Oktay, O. et al. Attention U-Net: Learning Where to Look for the Pancreas, 10.48550/arXiv.1804.03999 (2018). 1804. 03999

    2018 doi arXiv

  72. [81]

    Gupta, A. et al. Machine Learning Global Simulation of Nonlocal Gravity Wave Propagation (2024). 2406.14775

    2024 arXiv

  73. [82]

    Geller, M. A. et al. A Comparison between Gravity Wave Momentum Fluxes in Observations and Climate Models.J. Clim. 26, 6383–6405, 10.1175/JCLI-D-12-00545.1 (2013)

    2013 doi

  74. [83]

    & Beghdadi, A

    Bouzerdoum, A., Havstad, A. & Beghdadi, A. Image quality assessment using a neural network approach. In Proceedings of the Fourth IEEE International Symposium on Signal Processing and Information Technology, 2004., 330– 333 (IEEE, 2004)

    2004

  75. [84]

    Xie, S., Girshick, R., Dollar, P., Tu, Z. & He, K. Aggregated Residual Transformations for Deep Neural Networks. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 5987–5995, 10.1109/CVPR.2017.634 (IEEE, Honolulu, HI, 2017)

    2017 doi

  76. [85]

    Pfreundschuh, S. et al. A neural network approach to estimating a posteriori distributions of bayesian retrieval problems. Atmospheric Meas. Tech. 11, 4627–4643 (2018). 24/28

    2018

  77. [86]

    & Hutter, F

    Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101 (2017)

    2017 arXiv

  78. [87]

    Vitart, F. et al. The subseasonal to seasonal (s2s) prediction project database. Bull. Am. Meteorol. Soc. 98, 163 – 173, 10.1175/BAMS-D-16-0017.1 (2017)

    2017 doi

  79. [88]

    M., Coelho, C

    de Andrade, F. M., Coelho, C. A. & Cavalcanti, I. F. Global precipitation hindcast quality assessment of the subseasonal to seasonal (s2s) prediction project models. Clim. Dyn. 52, 5451–5475 (2019)

    2019

  80. [89]

    Dosovitskiy, A. et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929 (2020)

    2020 arXiv

  81. [90]

    Radford, A. et al. Language models are unsupervised multitask learners. OpenAI blog 1, 9 (2019)

    2019

  82. [91]

    /WxC-Bench/aviation_turbulence/dataset.py

    Lin, C.-Y . ROUGE: A package for automatic evaluation of summaries. In Text Summarization Branches Out, 74–81 (Association for Computational Linguistics, Barcelona, Spain, 2004). Acknowledgements This work was supported by NASA’s Office of Chief Science Data Officer and Earth ...

    2004

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.