Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

TAUDiff: Highly efficient kilometer-scale downscaling using generative diffusion models

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 A two-stage generative pipeline downscales atmospheric wind to kilometer resolution in about four minutes per year of data.

desk verdict Novel two-stage diffusion downscaler with a real speed win, but the km-scale extreme-event claim rests on an unvalidated single-sample assumption. read the letter →

arxiv 2412.13627 v2 pith:EHR5DJO2 submitted 2024-12-18 cs.LG

Rahul Sundar , Yucong Hu , Nishant Parashar , Antoine Blanchard , Boyko Dodov This is my paper · ORCID
classification cs.LG
keywords downscalingdiffusionmodelsspatio-temporalpredictionwindfieldskilometer-scaleextremeweatherspectralbiasvideo
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 proposes TAUDiff, a two-component generative downscaling pipeline for atmospheric wind fields. A deterministic spatio-temporal model (SimVP with a temporal attention unit) predicts the mean large-scale field, while a comparatively small diffusion model adds the fine-scale stochastic features that deterministic regression misses. The authors show that this separation recovers spatial and temporal spectra better than either a pure mean-field model or an end-to-end diffusion model under a fixed training budget, and that it preserves the dynamical consistency of storm fronts. They then extend the approach to kilometer-scale downscaling from ERA5 (0.25 degrees) to CERRA (0.0625 degrees) by running the diffusion at the coarse resolution and finishing with a deterministic UNet, reaching about four minutes of inference per year of data.

What carries the argument

The key machinery is the two-stage 'mean field plus correction' design: a SimVP-based video-prediction model with a Temporal Attention Unit (TAU) handles the deterministic spatio-temporal mean, trained with physics-based losses on advection, vorticity, and divergence; a smaller diffusion model (O(1) to O(10) million parameters) trained with score matching then generates the residual fine-scale stochastic component. In the kilometer-scale extension, the diffusion component runs at 0.25-degree resolution and a separate deterministic UNet regression model (O($10^{2}$) thousand parameters) maps its output up to 0.0625 degrees, which is what cuts inference time from roughly 76 minutes to roughly 4 minutes per year.

What would settle it

Generate many (say 100) TAUDiff samples for the same input period and compare a tail-sensitive statistic, such as the 99th percentile of wind speed or local storm counts, from a single sample versus the full ensemble against CERRA ground truth; if the single-sample value falls outside the ensemble spread, or the ensemble fails to cover the reference tail, the single-sample representativeness assumption is refuted.

Watch

Extended reading notes

Core claim

On atmospheric wind velocity fields over Europe, TAUDiff produces dynamically consistent, spectrally accurate downscaled fields at a fraction of the inference cost of running a diffusion model at full resolution. The central claim is that a small correction diffusion model is sufficient once a strong deterministic mean-field model accounts for the resolved dynamics: the diffusion model only needs to generate the remaining stochastic fine-scale content. As a proof of concept, chaining the coarse-resolution TAUDiff to a deterministic UNet downscales ERA5 to CERRA resolution (0.25 degrees to 0.0625 degrees) in about four minutes per year on a single H100 GPU, whereas feeding 0.0625-degree inputs directly to the diffusion model would take about 76 minutes. The authors also report that using a single diffusion sample at the coarse stage is intended to be representative of the field statistics, which is what makes the speedup possible.

Load-bearing premise

The speed advantage depends on a single diffusion sample at 0.25 degrees being representative of the full field statistics, so that no ensemble spread is needed before the deterministic fine-scale UNet step.

Editorial extensions

If this is right

  • A year of kilometer-scale wind downscaling takes roughly four minutes on a single GPU, making it practical to generate thousand-member event sets for tail-risk estimation.
  • Because the diffusion model only needs to generate residual stochastic content, its parameter count stays at O(1) to O(10) million, cutting training cost and carbon footprint relative to end-to-end diffusion.
  • Temporal modeling in the mean component preserves the propagation of storm fronts, avoiding the vorticity distortion that end-to-end diffusion introduces.
  • Spectral recovery is better than both a pure deterministic model, which underrepresents high frequencies, and an end-to-end diffusion model, which overrepresents them.

Reading between the lines

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

  • A direct test of the single-sample assumption would be to compare tail metrics (e.g., 99th-percentile wind speed or storm counts) from one TAUDiff pass against those from a multi-sample ensemble; if the tails diverge, a small ensemble for extreme-event runs could recover the statistics at a fraction of the cost of full-resolution diffusion.
  • The mean-plus-correction decomposition is not limited to wind: applying TAUDiff to temperature or precipitation would test whether the size of the needed diffusion corrector scales with the amount of variance left unexplained by the deterministic model.
  • The reported 4-minute figure assumes a single H100 and 20 reverse-diffusion steps; the paper's preliminary TensorRT result suggests up to a further 3x speedup, but that remains an unbenchmarked claim.
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 proposes TAUDiff, a two-component downscaling method for atmospheric wind fields. The first component is a deterministic spatio-temporal mean-field model (SimVP with a UNet backbone and temporal attention units) trained with physics-based losses (advection, vorticity, divergence). The second component is a small correction diffusion model (about 1M parameters) that generates stochastic fine-scale residuals conditioned on the mean-model output. The method is evaluated in two settings: (i) downscaling ERA5-derived coarse fields and bias-corrected CAM4 GCM fields to 0.25° resolution, compared against a deterministic mean-only model and an end-to-end diffusion model; and (ii) a proof-of-concept kilometer-scale extension from 0.25° ERA5 to 0.0625° CERRA, where TAUDiff runs at 0.25° and a deterministic UNet upsamples the result, yielding an inference time of about 4 minutes per year on a single H100 GPU. The central claims are that TAUDiff achieves dynamically consistent downscaling with accurate spatio-temporal spectra, enables fast extreme-event simulation, and that the km-scale pipeline preserves physically consistent fields.

Significance. If the claims hold, TAUDiff would be a practically useful contribution: it combines temporal modeling with generative correction, reports spectral recovery on held-out years and on out-of-training-distribution GCM inputs, and demonstrates a dramatic inference speedup over prior diffusion-based downscaling (about 4 minutes per year versus about two hours reported for the ensemble-diffusion baseline). The use of physics-based losses and vorticity-based qualitative evaluation are thoughtful, and the evaluation on CAM4 data that was not used in training provides external validity beyond the reanalysis-training regime. The paper is explicit about its limitations regarding inference-time assumptions and proof-of-concept status. However, the strength of the quantitative claims is tempered by missing uncertainty estimates, an unfair baseline comparison in the main validation, and an unvalidated single-sample assumption in the km-scale pipeline. These gaps prevent the paper from fully supporting its central claim that the method reliably captures extreme-event statistics at low inference cost.

major comments (3)
  1. [Section 3.1, Figures 2(b)-2(d)] The comparison against the end-to-end diffusion baseline is not ceteris paribus: the end-to-end diffusion model receives only a single time instance as input, whereas the mean model and TAUDiff receive a temporal sequence of snapshots. The reported qualitative and quantitative advantages of TAUDiff over the end-to-end diffusion model (e.g., less vorticity distortion, better temporal spectrum) could therefore be due to the additional temporal context rather than to the proposed architecture. Please retrain or re-evaluate the end-to-end diffusion baseline with the same temporal input, or explicitly justify why a single-frame input is the appropriate comparison.
  2. [Section 3.2, paragraph starting 'Methodology extension'] The km-scale inference speed claim rests on the assertion that 'a single ensemble member of the diffusion model should be representative of the field-statistics,' but this assertion is not validated anywhere in the paper. For a generative model, one draw is a single realization whose pointwise values, spatial spectra, and especially tail quantiles and storm counts are random variables. The paper provides no ensemble experiment: no multiple samples per date, no comparison of single-sample storm counts or 99th/99.9th percentile winds against ensemble statistics or CERRA, and no analysis of whether the deterministic UNet upsampler preserves stochastic sub-grid variability. Because the approximately 4-minute-per-year speed advantage comes precisely from generating only one coarse sample, the extreme-event claim is unsupported unless single-sample representativeness is demonstrated. Please add an ensemble-consistency experiment or qualify the extreme-event claims accordingly.
  3. [Section 3.1, Figures 2 and 3] The quantitative evaluation reports spectra and storm counts without any uncertainty estimates or error bars. The temporal spectra are computed over only three validation years (2021-2023) and the storm-count comparisons over CAM4-driven outputs rely on a simple quantile mapping; without confidence intervals or spread across GCM ensemble members, it is difficult to assess whether the reported agreement with ERA5 is statistically meaningful or whether apparent discrepancies are within noise. Please provide uncertainty quantification (e.g., bootstrap intervals, ensemble spread, or at least per-location variability).
minor comments (6)
  1. [Section 3.2, Figure 5] The km-scale validation shows only a single vorticity snapshot and aggregate spectra. Since the paper emphasizes extreme-event statistics, consider also showing storm-count maps or tail-quantile comparisons for the ERA5-to-CERRA experiment, as done for the CAM4 experiment in Figure 3(d).
  2. [Throughout] Figure captions contain a typo: 'V orticity' should be 'Vorticity'.
  3. [Section 2, Component 1] The weights for the MAE, MSE, advection, vorticity, and divergence losses are not specified. Please report these values or state that they are equal, since they directly affect the dynamical-consistency claim.
  4. [Section 2, Component 2] The text says the correction diffusion model uses a SimVP architecture but with a residual dense UNet as the spatial backbone; it would be clearer to state explicitly which components of SimVP (TAU, Inception modules, etc.) are retained and which are replaced.
  5. [Section 3.2, Experimental protocol] The training protocol for the km-scale experiment uses 10 years of ERA5/CERRA (2011-2020) and testing on 2010, which is a non-contiguous temporal split. Please clarify whether this was deliberate (e.g., to avoid overlapping weather regimes) and whether any care was taken to prevent leakage.
  6. [Section 3.2, Testing] The inference-time comparison of 76 minutes (at 0.0625° resolution) versus 4 minutes (at 0.25° resolution) does not state whether both figures use the same hardware and batching setup; please specify this to make the speedup interpretable.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TAUDiff is validated on held-out periods and on CAM4 GCM fields not used in training; the single-sample claim in Section 3.2 is an unvalidated modeling assumption, not a circular derivation.

full rationale

The paper's central results are empirical evaluations, not derivations that reduce to their inputs. The first experiment trains on ERA5 (1980-2020) and validates on 2021-23, then tests on bias-corrected CAM4 fields; neither validation set is used for fitting, so the spectral and storm-count agreement is independent evidence. The km-scale experiment trains on 2011-2020 ERA5/CERRA pairs and evaluates on 2010, so the Figure 5 agreement is also out-of-sample. The only self-referential elements are two citations to co-authored prior work ([9], [27]) used for context and spherical wavelet filtering; they do not carry the load of the main claim. The statement in Section 3.2 that 'a single ensemble member of the diffusion model should be representative of the field-statistics' is an assumption used to justify fast coarse-to-fine sampling; it is not derived from, nor equivalent to, the paper's inputs or definitions. It may be an unsupported statistical claim and a correctness risk for extreme-event tails, but that is not circularity. No fitted parameter is renamed as a prediction and no uniqueness theorem is imported from the authors. Therefore the circularity score is 0.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The empirical claims rest on several unstated or underspecified modeling choices: coarse-filtered ERA5 as GCM proxy, residual diffusion sufficiency, single-sample representativeness, and deterministic fine upsampling. These are domain choices, not new physical entities.

free parameters (4)
  • Loss weights for MAE, MSE, advection, vorticity, divergence
    Chosen by hand; exact values not reported; changes dynamical consistency of the mean model.
  • Reverse diffusion sampling steps = 20
    Selected to balance sample quality and inference speed; affects spectral fidelity.
  • Training epochs and parameter budgets = 50 epochs; O(10M) parameters
    Fixed across the three compared models; may disadvantage the end-to-end diffusion baseline, which receives no temporal conditioning.
  • Per-grid-cell quantile maps for CAM4 bias correction = fitted to 40-year distributions
    Used to put CAM4 on an ERA5 footing; fitted to historical data and not part of the network, but needed for the GCM test.
assumptions (5)
  • domain assumption Wavelet-filtered ERA5 coarse fields are representative of bias-corrected GCM coarse fields.
    Section 3.1: 'This approach closely mirrors real-world scenarios where bias-corrected GCM data lacks fine-scale spatio-temporal features.'
  • domain assumption The residual between mean-model output and truth is mostly fine-scale stochastic content capturable by a roughly 1M parameter diffusion model.
    Section 2: 'To capture the residual stochastic fine scale features... a relatively small correction diffusion model ( ~ O(1) million parameters)'.
  • domain assumption A single diffusion sample is representative of the field statistics.
    Section 3.2: 'a single ensemble member of the diffusion model should be representative of the field-statistics.'
  • ad hoc to paper A deterministic UNet regression at 0.0625 degrees preserves the stochastic fine-scale content generated by TAUDiff at 0.25 degrees.
    Section 3.2 proposes chaining 'the generated samples at coarser resolution can then be downscaled using deterministic UNet based regression model'; no ablation against a full-resolution diffusion is given.
  • standard math Score-based reverse diffusion sampling from the trained model produces samples from the target conditional distribution.
    Standard background (Song et al. 2020; Karras et al. 2022) relied on for sampling.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TAUDiff: Highly efficient kilometer-scale downscaling using generative diffusion models." pith.science (2026). https://pith.science/paper/EHR5DJO2

@misc{pith2026241213627,
  author       = {Pith},
  title        = {Pith review of: TAUDiff: Highly efficient kilometer-scale downscaling using generative diffusion models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EHR5DJO2}},
  note         = {Machine review of arXiv:2412.13627}
}
read the original abstract

Deterministic regression-based downscaling models for climate variables often suffer from spectral bias, which can be mitigated by generative models like diffusion models. To enable efficient and reliable simulation of extreme weather events, it is crucial to achieve rapid turnaround, dynamical consistency, and accurate spatio-temporal spectral recovery. We propose an efficient correction diffusion model, TAUDiff, that combines a deterministic spatio-temporal model for mean field downscaling with a smaller generative diffusion model for recovering the fine-scale stochastic features. We demonstrate the efficacy of this approach on downscaling atmospheric wind velocity fields obtained from coarse GCM simulations. We then extend TAUDiff for computationally efficient kilometer-scale downscaling of atmospheric wind velocity fields. Owing to low inference times, our approach can ensure quicker simulation of extreme events necessary for estimating associated risks and economic losses.

Figures

Figures reproduced from arXiv: 2412.13627 by the authors.

Figure 1
Figure 1. Schematic of the TAUDiff model a residual dense UNet as the spatial backbone. Once the model is trained, a data sample can be generated by solving a stochastic differential equation modelling a reverse diffusion process [17, 19]. Since the conditional input to the diffusion model is the mean model output (for a single time instance), the TAU morphs into a Channel Attention Unit (CAU) where the dynamical attention un… view at source ↗
Figure 2
Figure 2. (a) European region used for training the downscaling models, with select locations used for [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Assessment of downscaling performance on bias corrected CAM4 data: (a) Vorticity [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Schematic of the km-scale downscaling pipeline. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Assessment of ERA5 to CERRA downscaling performance: (a) Vorticity contours at UTC: [PITH_FULL_IMAGE:figures/full_fig_p005_5.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. Boosting Ensembles for Statistics of Tails at Conditionally Optimal Advance Split Times

    physics.ao-ph 2025-07 conditional novelty 7.0 of 10

    Ensemble boosting with a thresholded-entropy rule for choosing the advance split time accurately samples tail probabilities of extreme tracer fluctuations in a quasigeostrophic model.

Reference graph

Works this paper leans on

30 extracted references · 18 canonical work pages · cited by 1 Pith paper

  1. [1]

    P. Hoeppe. Trends in weather related disasters–consequences for insurers and society.Weather and Climate Extremes, 11:70–79, 2016

    work page 2016

  2. [2]

    Houser, S

    T. Houser, S. Hsiang, R. Kopp, K. Larsen, M. Delgado, A. Jina, M. Mastrandrea, S. Mohan, R. Muir-Wood, D. J. Rasmussen, et al. Economic risks of climate change: an American prospectus . Columbia University Press, 2015

    work page 2015

  3. [3]

    C. B. Field, V . Barros, T. F. Stocker, and Q. Dahe.Managing the risks of extreme events and disasters to advance climate change adaptation: Special report of the Intergovernmental Panel on Climate Change . Cambridge University Press, 2012

    work page 2012

  4. [4]

    SEC proposes rules to enhance and standardize climate-related disclosures for investors

    The Securities and Exchange Commission. SEC proposes rules to enhance and standardize climate-related disclosures for investors. https://www.sec.gov/news/press-release/2022-46 , March 21, 2022

    work page 2022

  5. [5]

    P. J. Robinson, W. J. W. Botzen, S. Duijndam, and A. Molenaar. Risk communication nudges and flood insurance demand. Climate Risk Management, 34:100366, 2021

    work page 2021

  6. [6]

    Hersbach, B

    H. Hersbach, B. Bell, P. Berrisford, S. Hirahara, A. Horányi, J. Muñoz-Sabater, J. Nicolas, C. Peubey, R. Radu, D. Schepers, et al. The ERA5 global reanalysis. Quarterly Journal of the Royal Meteorological Society, 146(730):1999–2049, 2020

    1999

  7. [7]

    C. Wang, L. Zhang, S.-K. Lee, L. Wu, and C. R. Mechoso. A global perspective on CMIP5 climate model biases. Nature Climate Change, 4(3):201–205, 2014

    work page 2014

  8. [8]

    S. Park, K. Singh, A. Nellikkattil, E. Zeller, and M. Mai, T. D.and Cha. Downscaling earth system models with deep learning. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 3733–3742, 2022

    work page 2022

Show all 30 references
  1. [9]

    Blanchard, N

    A. Blanchard, N. Parashar, B. Dodov, C. Lessig, and T. Sapsis. A multi-scale deep learning framework for projecting weather extremes. arXiv preprint arXiv:2210.12137, 2022

    2022 arXiv

  2. [10]

    Daust and A

    K. Daust and A. Monahan. Capturing climatic variability: Using deep learning for stochastic downscaling. arXiv preprint arXiv:2406.02587, 2024

    2024 arXiv

  3. [11]

    Z. J. Xu, Y . Zhang, and T. Luo. Overview frequency principle/spectral bias in deep learning.arXiv preprint arXiv:2201.07395, 2022

    2022 arXiv

  4. [12]

    Rampal, P

    N. Rampal, P. B. Gibson, S. Sherwood, G. Abramowitz, and S. Hobeichi. A robust generative adversarial network approach for climate downscaling and weather generation. Authorea Preprints, 2024

    2024

  5. [13]

    Li and G

    G. Li and G. Cao. Generative Adversarial Models for Extreme Downscaling of Climate Datasets. arXiv preprint arXiv:2402.14049, 2024

    2024 arXiv

  6. [14]

    Mardani, N

    M. Mardani, N. Brenowitz, Y . Cohen, J. Pathak, C.Y . Chen, C.C. Liu, A. Vahdat, K. Kashinath, J. Kautz, and M. Pritchard. Generative residual diffusion modeling for km-scale atmospheric downscaling. arXiv preprint arXiv:2309.15214, 2023

    2023 arXiv

  7. [15]

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

    2024 arXiv

  8. [16]

    J. Song, C. Meng, and S. Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020

    2010 arXiv

  9. [17]

    J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models. Advances in Neural Information Processing Systems, 33:6840–6851, 2020

    2020

  10. [18]

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole. Score-based generative modeling through stochastic differential equations. arXiv preprint arXiv:2011.13456, 2020

    2011 arXiv

  11. [19]

    Karras, M

    T. Karras, M. Aittala, T. Aila, and S. Laine. Elucidating the design space of diffusion-based generative models. Advances in Neural Information Processing Systems , 35:26565–26577, 2022

    2022

  12. [20]

    Merizzi, A

    F. Merizzi, A. Asperti, and S. Colamonaco. Wind speed super-resolution and validation: from era5 to cerra via diffusion models. arXiv preprint arXiv:2401.15469, 2024. 6

    2024 arXiv

  13. [21]

    C. Tan, S. Li, Z. Gao, W. Guan, Z. Wang, Z. Liu, L. Wu, and S. Z. Li. Openstl: A comprehensive benchmark of spatio-temporal predictive learning. Advances in Neural Information Processing Systems , 36:69819–69831, 2023

    2023

  14. [22]

    D. Yoon, M. Seo, D. Kim, Y . Choi, and D. Cho. Deterministic guidance diffusion model for probabilistic weather forecasting. arXiv preprint arXiv:2312.02819, 2023

    2023 arXiv

  15. [23]

    Z. Gao, C. Tan, L. Wu, and S. Z Li. Simvp: Simpler yet better video prediction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 3170–3180, 2022

    2022

  16. [24]

    C. Tan, Z. Gao, L. Wu, Y . Xu, J. Xia, S. Li, and S. Z. Li. Temporal attention unit: Towards efficient spatiotemporal predictive learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 18770–18782, 2023

    2023

  17. [25]

    Song and S

    Y . Song and S. Ermon. Improved techniques for training score-based generative models. Advances in Neural Information Processing Systems, 33:12438–12448, 2020

    2020

  18. [26]

    Schröder and W

    P. Schröder and W. Sweldens. Spherical wavelets: Efficiently representing functions on the sphere. In Proceedings of the 22nd Annual Conference on Computer Graphics and Interactive Techniques , pages 161–172, 1995

    1995

  19. [27]

    C. C. da Silva, C. Lessig, B. Dodov, H. Dijkstra, and T. Sapsis. A local spectral exterior calculus for the sphere and application to the shallow water equations. arXiv preprint arXiv:2005.03598, 2020

    2005 arXiv

  20. [28]

    B Neale, J

    R. B Neale, J. H. Richter, AJ Conley, S. Park, P. H. Lauritzen, A. Gettelman, D.L. Williamson, P.J. Rasch, S. J. Vavrus, M. A. Taylor, et al. Description of the NCAR Community Atmosphere Model (CAM 4.0), NCAR Tech. Note, TN–485, 212, 2010

    2010

  21. [29]

    D. Maraun. Bias correction, quantile mapping, and downscaling: Revisiting the inflation issue. Journal of Climate, 26(6):2137–2143, 2013

    2013

  22. [30]

    Cerra sub-daily regional reanalysis data for europe on single levels from 1984 to present

    S Schimanke, M Ridal, P Le Moigne, L Berggren, P Undén, R Randriamampianina, U Andrea, E Bazile, A Bertelsen, P Brousseau, et al. Cerra sub-daily regional reanalysis data for europe on single levels from 1984 to present. Copernicus Climate Change Service (C3S) Climate Data Sto...

    1984

Pith tools

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