Pith. sign in

REVIEW 4 major objections 5 minor 53 references

RainShift: A Benchmark for Precipitation Downscaling Across Geographies

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read RainShift, a global benchmark built from reanalysis and satellite rainfall, shows that state-of-the-art precipitation downscaling models degrade by up to 30% when applied to unseen regions, and that quantile-mapping input alignment…

desk verdict RainShift is a genuinely useful benchmark for cross-geography downscaling and deserves reviewer time, but the 'controlled setup' framing is untested and the headline OOD drops likely mix geographic shift with regional product biases. read the letter →

arxiv 2507.04930 v1 pith:7K6KLKFV submitted 2025-07-07 cs.CV

classification cs.CV
keywords precipitationdownscalinggeographicgeneralizationdistributionshiftdeeplearningbenchmarkdatasetERA5IMERGquantilemapping
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

RainShift is a dataset and benchmark for testing whether deep-learning precipitation downscaling models trained in data-rich regions, mostly in the Global North, can be applied to data-scarce regions, mostly in the Global South. The paper pairs coarse ERA5 reanalysis fields with high-resolution IMERG satellite rainfall across 12 training and 6 evaluation regions, and evaluates a deterministic ResNet, a Wasserstein GAN, and a diffusion model. Its central finding is that every learned model beats bilinear interpolation in unseen regions, but out-of-distribution performance drops by up to 30% relative to training on the target region, and the drop remains up to 17% even with the largest training domain. The paper also shows that aligning the input rainfall distributions of target regions to the training region with quantile mapping improves or matches performance in most regions, suggesting that distribution shift, not model architecture, is the main barrier to geographic generalization. If right, this gives the community a standard way to measure and improve transfer of downscaling models to regions with scarce observations.

What carries the argument

The load-bearing object is the benchmark itself: a set of 18 fixed 20-degree-by-20-degree patches, 12 training and 6 evaluation, built from paired hourly ERA5 reanalysis inputs (nine atmospheric variables plus land-sea mask and orography) and IMERG satellite precipitation targets at 2.5x upsampling. Training configurations A1 through A4 are hierarchical subsets of the training patches, simulating increasing observational coverage. Evaluation uses pixel-wise CRPS with eight samples, against bilinear interpolation as a lower bound and in-region training as an upper bound. The paper's corrective mechanism is multiplicative quantile mapping, which maps the target region's historical input CDF onto the training region's CDF and applies that transfer function to future inputs, aligning distributions before normalization.

What would settle it

Measure ERA5-minus-IMERG residuals against independent gauge or radar data separately in each training and evaluation region; if those residual differences across regions are comparable in size to the reported out-of-distribution CRPS drops, the benchmark's geographic shift is partly an artifact of data-product bias rather than pure generalization.

Watch

Extended reading notes

Core claim

The paper's central claim is that geographic distribution shift is the dominant obstacle to using learned downscaling models in new regions, and that this shift is measurable and partly correctable. Across six target regions and four hierarchical training configurations, all three learned models improve on bilinear interpolation, yet CRPS relative to in-distribution training falls by up to 30% for the single-region setup and up to 17% for the full Global North setup. Adding more training regions helps in some areas but not uniformly, and on-target training is not always best. The paper further shows that applying multiplicative quantile mapping to the low-resolution precipitation inputs before inference reduces the distributional mismatch and improves or equals the unaligned baseline in all target regions except Cape Horn, even stabilizing a ResNet that otherwise produces numerically unreliable CRPS on the Tibetan Plateau. The benchmark is offered as a standardized zero-shot task: train on the training regions, evaluate on the evaluation regions, with CRPS as the headline metric.

Load-bearing premise

The load-bearing premise is that ERA5 reanalysis and IMERG satellite rainfall are globally consistent enough that a model trained in one region and tested in another isolates geographic distribution shift rather than regional data-product bias.

Editorial extensions

If this is right

  • All learned downscaling models beat bilinear interpolation in every evaluation region, so the learned coarse-to-fine mapping transfers at least partially even when the target region is unseen.
  • Out-of-distribution CRPS drops of up to 30% in the single-region setup and up to 17% in the largest setup show that simply adding training regions does not eliminate the geographic gap, meaning zero-shot use in data-sparse regions carries a measurable accuracy penalty.
  • Probabilistic models, both GAN and diffusion, reliably outperform the deterministic ResNet, with the diffusion model more robust to a small training domain and the GAN benefiting more from added regions.
  • Quantile-based input alignment improves or matches unaligned performance in all target regions except Cape Horn, including stabilizing a ResNet that otherwise produces numerically unreliable scores on the Tibetan Plateau.
  • Performance differences between GAN and diffusion are small relative to differences between geographic regions, so regional climate variation is the dominant factor limiting generalization.

Reading between the lines

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

  • If the same splits were evaluated with a local radar or gauge target instead of IMERG, the out-of-distribution drops could be larger, because regional product biases would add to the geographic shift.
  • Extending quantile mapping from the single precipitation input to all nine ERA5 variables might close more of the remaining gap; the paper reports only the precipitation-input variant.
  • The strong correlation between mean precipitation and error suggests region difficulty could be predicted from climatology, allowing future benchmarks to stratify results by expected difficulty rather than treating regions as exchangeable.
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

4 major / 5 minor

Summary. The paper introduces RainShift, a large-scale benchmark dataset and evaluation framework for studying geographic generalization of deep-learning-based precipitation downscaling. The dataset pairs ERA5 reanalysis fields (coarse input) with IMERG satellite precipitation (high-resolution target) over 12 training regions and 6 evaluation regions, and defines four hierarchical training scenarios (A1–A4) to simulate increasing availability of high-resolution observations. The authors evaluate a deterministic ResNet, a WGAN-GP, a diffusion model, and bilinear interpolation, using CRPS as the primary metric, and supplement the main results with in-region training baselines and a quantile-mapping input-alignment experiment. The headline findings are that all learned models outperform bilinear interpolation, generative models outperform the deterministic ResNet, out-of-distribution performance drops by up to 30% (A1) and 17% (A4) relative to in-distribution training, expanding the training domain helps only partially, and quantile mapping improves performance in most target regions.

Significance. If the benchmark's premise holds, RainShift fills a real gap: existing downscaling benchmarks largely evaluate within a single region or on a limited set of regions, whereas cross-geography generalization is a central operational problem for global applicability. The paper's strengths include a substantial public dataset (300 GB of Zarr data), a clear temporal train/validation/test split, multiple model classes, in-region upper-bound comparisons, and a concrete domain-alignment proposal with public code. The quantitative results, while preliminary in places, provide a useful reference point for future method development. However, the central interpretative claim — that the benchmark measures geographic/climatic distribution shift rather than regionally varying data-product biases — is not yet supported by direct evidence, and several reporting issues (an unexplained extreme outlier, lack of repeated-seed statistics, and an overstatement of quantile-mapping gains) limit the confidence one can place in the specific numerical conclusions.

major comments (4)
  1. [Methods (Input data, Target data, Data processing); Table 4] The claim that 'globally consistent satellite and reanalysis data enables a controlled benchmark setup' is not substantiated. Both ERA5 and IMERG have known, region-dependent error characteristics: ERA5 is constrained by a heterogeneous assimilation network and is less reliable in data-sparse and complex-terrain regions, while IMERG Final Run incorporates gauge calibration whose station density varies geographically. The six evaluation regions (Amazon, Tibetan Plateau, Melanesia, etc.) are precisely regions where such product errors are expected to be largest. Furthermore, the preprocessing step clips ERA5 precipitation using IMERG-derived min/max thresholds, injecting target-product statistics into the input stream. Consequently, the reported out-of-distribution CRPS drops (Figures 5–6, Table 4) may partly reflect regional product-bias artifacts rather than pure geographic/climatic shift. To support the benchmark's central premise, the authors should provide diagnostic evidence of product consistency, for example by comparing ERA5 and IMERG against available local gauge or radar measurements in the evaluation regions, and by reporting the sensitivity of the main results to the clipping step (e.g., omitting clipping or varying thresholds).
  2. [Table 4, Results ('Probabilistic models outperform deterministic ones')] The ResNet A1 entry for the Tibetan Plateau (E5) is reported as CRPS = 18113.558, which is several orders of magnitude larger than all other values. The text attributes this to numerical instabilities during inference, but the numerical value is still included in the table, and the caption describes it as showing mean pixel-wise CRPS. This outlier should either be excluded and replaced with a placeholder (e.g., 'N/A' or 'unstable') or analyzed quantitatively so that the reader can understand whether it is a single divergent sample, an overflow in the CRPS calculation, or a genuine model failure. As reported, the number obscures the comparison among models for that cell and could distort any aggregate analysis computed from Table 4.
  3. [Evaluation ('Quantitative evaluation'), Training details] All results in Table 4 and Figures 5–6 appear to be based on a single training run per model and training scenario; no repeated-seed statistics or confidence intervals are reported. Given that GANs and diffusion models are stochastic both in training and in sampling, and that many performance differences in Table 4 are small (e.g., GAN vs. diffusion differences of 0.01–0.02 mm/h), it is not clear that the qualitative conclusions (generative models outperform ResNet; expansion from A3 to A4 helps in some regions) are robust to training seed variability. The authors should provide mean and standard deviation (or at least seed-level results) across several training runs, or otherwise justify why the 8-sample CRPS ensemble is sufficient to establish the reported patterns.
  4. [Methods ('Quantile mapping for geographical generalization'), Table 1, Results ('Geographical factors dominate…] The quantile mapping results are overstated relative to the data. The text says quantile mapping 'greatly improves performance' and 'across nearly all regions,' while Table 1 shows that the diffusion model improves only in E3 and E5, degrades in E1 and E6, and is unchanged in E2 and E4; the GAN improves in E3, E5, and E6, but degrades in E1 and E3 (E3: 0.093 to 0.080, actually improvement; E1: 0.075 to 0.093 is a degradation; E3 improves; E5 improves; E6 improves; E2 unchanged; E4 unchanged). The caption itself notes 'except Cape Horn,' but the main text's phrasing is stronger than the evidence. The paper should either soften the claim to 'improves performance for some models and regions' or provide a more nuanced per-model characterization.
minor comments (5)
  1. [General] The manuscript contains several typographical errors, including 'atomospheric' (Methods, RainShift dataset), 'particulary' (Methods, Data processing), 'high-reslution' (Background), and 'Y et' (Abstract). These should be corrected.
  2. [Results ('Probabilistic models outperform deterministic ones')] The text refers to 'absolute (see Figure 4)' when reporting A4 GAN and diffusion performance, but Figure 4 is a map of training and evaluation regions, not a performance plot. The absolute CRPS values are in Table 4; the citation should be corrected.
  3. [Figure 5 and Figure 6] The heatmaps in Figures 5 and 6 are described in the text as showing percentage improvements/drops, but the color scale and exact numerical values are not defined in the captions or in the text. Adding a colorbar with units and describing how the percentages are computed (e.g., relative to which baseline) would improve interpretability.
  4. [Table 1 caption] The caption states that 'Applying quantile mapping equals or improves performance across most models and regions (except Cape Horn),' but the table shows that for the diffusion model, Melanesia (E6) also degrades (0.295 to 0.310), and for the GAN, E1 degrades (0.075 to 0.093). The caption should be updated to reflect the full pattern of results.
  5. [Data availability] The repository is said to be 'made available upon acceptance,' which is acceptable for a submission, but the authors should clarify the planned license for the benchmark code and dataset, as well as whether the dataset can be downloaded directly from Hugging Face without additional steps.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: RainShift is an empirical benchmark evaluation against external baselines and held-out target data.

full rationale

This paper makes no analytic derivation claim; it constructs a dataset and measures CRPS of models (ResNet, WGAN, diffusion) on held-out regions against IMERG. The headline findings ('learned models improve over bilinear interpolation'; OOD CRPS drops up to 30%/17%; quantile mapping helps) are all empirical comparisons measured against externally fixed baselines and test data, not quantities fitted from the test set. The quantile-mapping experiment fits its transfer function to historical CDFs (2001-2020) of training and target inputs and applies it to 2021-2022 inputs before evaluating on 2021-2022 IMERG; future target values are not used in fitting the mapping, so the reported improvement is not forced by construction. Architectures are adopted from prior work (Harris et al. 2022; Watt & Mansfield 2024), but the reported CRPS values are measured against IMERG, and the prior-work citations are architectural provenance, not uniqueness claims or forbidden alternatives. The only concern raised by the skeptic—that ERA5/IMERG may have region-dependent biases so 'geographic shift' is partly product-bias—is a validity/interpretability concern about the benchmark's controlled-setup premise, not a circularity: no result is equivalent to an input by definition. The paper itself notes the limitation by stating 'a promising avenue for future work is to expand the benchmark to incorporate additional, more localized sources'. Self-citations are minor (e.g., Harder et al. 2023) and non-load-bearing. Therefore the circularity score is 0.

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

No physical entities are introduced. The benchmark relies on standard statistical downscaling assumptions; the central claims are empirical and are not derived from fitted constants.

free parameters (3)
  • Number of quantiles in quantile mapping = 1000
    Chosen by hand in Methods; QM results depend on this discretization but no sensitivity analysis is reported.
  • Number of ensemble samples for CRPS = 8
    Evaluation uses 8 samples per location to estimate CRPS; no convergence or variance analysis is reported.
  • Patch size and upsampling factor = N=2.5; 80x80 to 200x200
    Fixed task design choice; the benchmark results may depend on this resolution ratio.
assumptions (3)
  • domain assumption ERA5 and IMERG are globally consistent enough that paired training/evaluation measures geographic transfer rather than product-specific artifacts.
    Methods: 'The use of globally consistent satellite and reanalysis data enables a controlled benchmark setup'; the benchmark's validity depends on this.
  • domain assumption The target relationship p(y|x) is comparable across regions, so zero-shot evaluation is meaningful.
    The downscaling task formulation assumes pE(·|xE) is learnable from training region pA; this is the benchmark premise stated in Methods, 'Geographical generalization'.
  • domain assumption Pixel-wise CRPS over 8 samples is a reliable ranking metric without repeated seeds or uncertainty intervals.
    Evaluation section describes CRPS over 8 samples; no variance or multiple-seed analysis is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of RainShift: A Benchmark for Precipitation Downscaling Across Geographies." pith.science (2026). https://pith.science/paper/7K6KLKFV

@misc{pith2026250704930,
  author       = {Pith},
  title        = {Pith review of: RainShift: A Benchmark for Precipitation Downscaling Across Geographies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7K6KLKFV}},
  note         = {Machine review of arXiv:2507.04930}
}
read the original abstract

Earth System Models (ESM) are our main tool for projecting the impacts of climate change. However, running these models at sufficient resolution for local-scale risk-assessments is not computationally feasible. Deep learning-based super-resolution models offer a promising solution to downscale ESM outputs to higher resolutions by learning from data. Yet, due to regional variations in climatic processes, these models typically require retraining for each geographical area-demanding high-resolution observational data, which is unevenly available across the globe. This highlights the need to assess how well these models generalize across geographic regions. To address this, we introduce RainShift, a dataset and benchmark for evaluating downscaling under geographic distribution shifts. We evaluate state-of-the-art downscaling approaches including GANs and diffusion models in generalizing across data gaps between the Global North and Global South. Our findings reveal substantial performance drops in out-of-distribution regions, depending on model and geographic area. While expanding the training domain generally improves generalization, it is insufficient to overcome shifts between geographically distinct regions. We show that addressing these shifts through, for example, data alignment can improve spatial generalization. Our work advances the global applicability of downscaling methods and represents a step toward reducing inequities in access to high-resolution climate information.

Figures

Figures reproduced from arXiv: 2507.04930 by the authors.

Figure 1
Figure 1. Map of ground-based radar stations. The map shows the availability of precipitation data, with each blue dot representing a station. Coverage is relatively high in the Global North and comparatively low across the Global South. Image from the Tropical Globe radar database20 . Added value of downscaling models All learned models demonstrate some generalization to unseen target regions. As shown in [PITH_FULL_IMAGE:f… view at source ↗
Figure 2
Figure 2. Graphical summary of RainShift setup. The inputs of the downscaling model are a combination of ERA5 time series data and geographical features. The downscaling model is then able to generate probabilistic samples. For training, we sample from geographic areas T1,...,T12 and years 2001–2020, and compare the generated samples with the ground truth target IMERG to compute the loss. For evaluation, we use areas E1,...,E… view at source ↗
Figure 3
Figure 3. Illustration of training configurations. The training configurations A1,...,A4 are composed of progressively larger subsets of the 12 selected training regions located in the Global North. The choice of regions is guided by availability of high-resolution observational data and inspired by existing works14, 23 . 180° 180° 120°W 120°W 60°W 60°W 0° 0° 60°E 60°E 120°E 120°E 180° 180° 60°S 60°S 40°S 40°S 20°S 20°S 0° 0°… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Illustration of location splits and training configurations. Patches T1,...,T12 represent training regions and patches E1,...,E6 correspond to evaluation areas that are used within 4 sub-tasks, simulating different scenarios that correspond to varying levels of data av…
Figure 5
Figure 5. Figure 5: Heatmap of % improvement relative to interpolation. Change in CRPS (lower better) in [%] for each model relative to bilinear interpolation. A1,...,A4 represent hierarchical training scenarios with progressively more high-resolution data, from training on a single regio…
Figure 6
Figure 6. Figure 6: Heatmap of % performance drop between in and out-of-distribution training. Change in CRPS (lower better) in [%] for each model relative to training the model directly on the target regions. A1,...,A4 represent hierarchical training scenarios with progressively more hig…
Figure 7
Figure 7. Figure 7: Qualitative comparison of downscaled precipitation fields. This plot shows a sample, one time step from the evaluation set in the Cape Horn area in the first two columns and aggregated features in the last column. The random sample includes the matching input (ERA5) an…
Figure 8
Figure 8. Figure 8: Cumulative distribution functions (CDFs) of precipitation inputs for training and target regions. For each target region, a mapping is constructed between the historic CDF of the training region and the historic precipitation inputs of the target region. This mapping i…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

53 extracted references · 38 canonical work pages

  1. [1]

    Seneviratne, S. I. et al. Weather and climate extreme events in a changing climate. In Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, 2021)

  2. [2]

    D., Rhoades, A

    Ombadi, M., Risser, M. D., Rhoades, A. M. & Varadharajan, C. A warming-induced reduction in snow fraction amplifies rainfall extremes. Nature 619, 305–310 (2023)

  3. [3]

    & Yang, Y

    Xiong, J. & Yang, Y . Climate change and hydrological extremes.Curr. Clim. Chang. Reports 11, 1 (2024)

  4. [4]

    Maraun, D. et al. Precipitation downscaling under climate change: Recent developments to bridge the gap between dynamical models and the end user. Rev. geophysics 48 (2010). 13/17

  5. [5]

    G., Knutti, R., Lehner, F., Deser, C

    Pendergrass, A. G., Knutti, R., Lehner, F., Deser, C. & Sanderson, B. M. Precipitation variability increases in a warmer climate. Sci. reports 7, 17966 (2017)

  6. [6]

    Fosser, G. et al. Convection-permitting climate models offer more certain extreme rainfall projections. NPJ Clim. atmospheric science 7, 51 (2024)

  7. [7]

    Deepsd: Generating high resolution climate change projections through single image super-resolution

    Vandal, T.et al. Deepsd: Generating high resolution climate change projections through single image super-resolution. Assoc. for Comput. Mach. 1663–1672, DOI: 10.1145/3097983.3098004 (2017)

  8. [8]

    J., West, G

    Sha, Y ., Gagne, D. J., West, G. & Stull, R. Deep-learning-based gridded downscaling of surface meteorological variables in complex terrain. part i: Daily maximum and minimum 2-m temperature. J. Appl. Meteorol. Climatol. (2020)

Show all 53 references
  1. [9]

    & Westermann, R

    Höhlein, K., Kern, M., Hewson, T. & Westermann, R. A comparative study of convolutional neural network models for wind field downscaling. Meteorol. Appl. 27, e1961, DOI: https://doi.org/10.1002/met.1961 (2020). https://rmets. onlinelibrary.wiley.com/doi/pdf/10.1002/met.1961

  2. [10]

    Liu, Y ., Ganguly, A. R. & Dy, J. Climate downscaling using ynet: A deep convolutional network with skip connections and fusion. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , KDD ’20, 3145–3153, DOI: 10.1145/3394486.34033...

  3. [11]

    & Netto, M

    Rocha Rodrigues, E., Oliveira, I., Cunha, R. & Netto, M. Deepdownscale: A deep learning strategy for high-resolution weather forecast. In 2018 IEEE 14th International Conference on e-Science (e-Science) , 415–422, DOI: 10.1109/eScience. 2018.00130 (2018)

  4. [12]

    Goodfellow, I. J. et al. Generative adversarial networks. Commun. ACM 63, 139 – 144 (2014)

  5. [13]

    Harris, L., McRae, A. T. T., Chantry, M., Dueben, P. D. & Palmer, T. N. A generative deep learning approach to stochastic downscaling of precipitation forecasts. J. Adv. Model. Earth Syst. 14 (2022)

  6. [14]

    C., McRae, A

    Cooper, F. C., McRae, A. T. T., Chantry, M., Antonio, B. & Palmer, T. N. Further analysis of cgan: A system for generative deep learning post-processing of precipitation (2023). 2309.15689

  7. [15]

    N., Weiss, E

    Sohl-Dickstein, J. N., Weiss, E. A., Maheswaranathan, N. & Ganguli, S. Deep unsupervised learning using nonequilibrium thermodynamics. ArXiv abs/1503.03585 (2015)

  8. [16]

    Mardani, M. et al. Generative residual diffusion modeling for km-scale atmospheric downscaling. ArXiv abs/2309.15214 (2023)

  9. [17]

    Y .et al

    Wan, Z. Y .et al. Debias coarsely, sample conditionally: Statistical downscaling through optimal transport and probabilistic diffusion models. ArXiv abs/2305.15618 (2023)

  10. [18]

    & Watson, P

    Addison, H., Kendon, E., Ravuri, S., Aitchison, L. & Watson, P. A. Machine learning emulation of precipitation from km-scale regional climate simulations using a diffusion model (2024). 2407.14158

  11. [19]

    Ling, F., Lu, Z., Luo, J. J. et al. Diffusion model-based probabilistic downscaling for 180-year East Asian climate reconstruction. npj Clim. Atmospheric Sci. 7, 131, DOI: 10.1038/s41612-024-00679-1 (2024)

  12. [20]

    Tropical globe radar database

    Tropical Globe. Tropical globe radar database. https://tropicalglobe.com/radar_database/ (2025). Accessed: 2025-01-26

  13. [21]

    & Bottou, L

    Arjovsky, M., Chintala, S. & Bottou, L. Wasserstein GAN. ArXiv abs/1701.07875 (2017)

  14. [22]

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

  15. [23]

    Prasad, A. et al. Evaluating the transferability potential of deep learning models for climate downscaling. ICML Work. Mach. Learn. for Earth Syst. Model. (2024)

  16. [24]

    & Rußwurm, M

    Klemmer, K., Rolf, E., Robinson, C., Mackey, L. & Rußwurm, M. Satclip: Global, general-purpose location embeddings with satellite imagery. ArXiv abs/2311.17179 (2023)

  17. [25]

    Exploring the potential of multi-source unsupervised domain adaptation in crop mapping using sentinel-2 images

    Wang, Y .et al. Exploring the potential of multi-source unsupervised domain adaptation in crop mapping using sentinel-2 images. GIScience & Remote. Sens. 59, 2247–2265, DOI: 10.1080/15481603.2022.2156123 (2022). https://doi.org/10. 1080/15481603.2022.2156123

  18. [26]

    & Lobell, D

    Rußwurm, M., Wang, S., Körner, M. & Lobell, D. Meta-learning for few-shot land cover classification. Prepr. arXiv 2004.13390 (2020)

  19. [28]

    Hersbach, H. et al. The era5 global reanalysis. Q. J. Royal Meteorol. Soc. 146, 1999–2049, DOI: https://doi.org/10.1002/ qj.3803 (2020). 14/17

  20. [29]

    & Pillosu, F

    Hewson, T. & Pillosu, F. M. A low-cost post-processing technique improves weather forecasts around the world. Commun. Earth & Environ. 2 (2020)

  21. [30]

    Huffman, G. et al. Integrated Multi-satellitE Retrievals for GPM (IMERG), version 4.4. NASA’s Precipitation Processing Center (2014). Accessed: 31 March, 2015

  22. [31]

    Huffman, G. J. et al. IMERG V07 Release Notes. https://gpm.nasa.gov/resources/documents/imerg-v07-release-notes (2024). Accessed: 2025-01-26

  23. [32]

    Miles, A. et al. zarr-developers/zarr-python: v2.4.0, DOI: 10.5281/zenodo.3773450 (2020)

  24. [33]

    N., Beighley, E

    Seyyedi, H., Anagnostou, E. N., Beighley, E. & McCollum, J. Hydrologic evaluation of satellite and reanalysis precipitation datasets over a mid-latitude basin. Atmospheric Res. 164, 37–48 (2015)

  25. [34]

    Evaluation of imerg and era5 precipitation products over the mongolian plateau

    Xin, Y .et al. Evaluation of imerg and era5 precipitation products over the mongolian plateau. Sci. reports 12, 21776 (2022)

  26. [35]

    Rampal, N. et al. Enhancing regional climate downscaling through advances in machine learning. Artif. Intell. for Earth Syst. DOI: 10.1175/AIES-D-23-0066.1 (2024)

  27. [36]

    Hard-constrained deep learning for climate downscaling

    Harder, P.et al. Hard-constrained deep learning for climate downscaling. J. Mach. Learn. Res. 24, 1–40 (2023)

  28. [37]

    & Berne, A

    Leinonen, J., Nerini, D. & Berne, A. Stochastic super-resolution for downscaling time-evolving atmospheric fields with a generative adversarial network. IEEE Transactions on Geosci. Remote. Sens. 59, 7211–7223 (2020)

  29. [38]

    & Ganguli, S

    Sohl-Dickstein, J., Weiss, E., Maheswaranathan, N. & Ganguli, S. Deep unsupervised learning using nonequilibrium thermodynamics. In Bach, F. & Blei, D. (eds.) Proceedings of the 32nd International Conference on Machine Learning , vol. 37 of Proceedings of Machine Learning Rese...

  30. [39]

    Price, I., Sanchez-Gonzalez, A., Alet, F. et al. Probabilistic weather forecasting with machine learning. Nature 637, 84–90, DOI: 10.1038/s41586-024-08252-9 (2025)

  31. [40]

    & Laine, S

    Karras, T., Aittala, M., Aila, T. & Laine, S. Elucidating the design space of diffusion-based generative models. In Proc. NeurIPS (2022)

  32. [41]

    & Smith, L

    Broecker, J. & Smith, L. A. Increasing the Reliability of Reliability Diagrams , vol. 22 (Weather and Forecasting, 2007)

  33. [42]

    & Rolnick, D

    Tseng, G., Kerner, H. & Rolnick, D. Timl: Task-informed meta-learning for agriculture (2022). 2202.02124

  34. [43]

    Teng, M. et al. Satbird: a dataset for bird species distribution modeling using remote sensing and citizen science data. In Oh, A. et al. (eds.) Advances in Neural Information Processing Systems , vol. 36, 75925–75950 (Curran Associates, Inc., 2023)

  35. [44]

    Cole, E. et al. Spatial Implicit Neural Representations for Global-Scale Species Mapping. In ICML (2023)

  36. [45]

    & Zhou, Q

    Zhu, H. & Zhou, Q. Advancing satellite-derived precipitation downscaling in data-sparse area through deep transfer learning. IEEE Transactions on Geosci. Remote. Sens. 62, 1–13, DOI: 10.1109/TGRS.2024.3367332 (2024)

  37. [46]

    Sha, Y ., II, D. J. G., West, G. & Stull, R. Deep-learning-based gridded downscaling of surface meteorological variables in complex terrain. part ii: Daily precipitation. J. Appl. Meteorol. Climatol. 59, 2075 – 2092, DOI: 10.1175/JAMC-D-20-0058. 1 (2020)

  38. [47]

    Prasad, A. et al. Evaluating the transferability potential of deep learning models for climate downscaling (2024). 2407.12517

  39. [48]

    Rasp, S. et al. Weatherbench 2: A benchmark for the next generation of data-driven global weather models (2023). 2308.15560

  40. [49]

    Watson-Parris, D. et al. Climatebench v1.0: A benchmark for data-driven climate projections. J. Adv. Model. Earth Syst. 14 (2022)

  41. [50]

    & Grover, A

    Nguyen, T., Jewik, J., Bansal, H., Sharma, P. & Grover, A. Climatelearn: Benchmarking machine learning for weather and climate modeling. ArXiv abs/2307.01909 (2023)

  42. [51]

    Kaltenborn, J. et al. Climateset: A large-scale climate model dataset for machine learning. ArXiv abs/2311.03721 (2023)

  43. [52]

    Schroeder de Witt, C. et al. Rainbench: Towards data-driven global precipitation forecasting from satellite imagery. Proc. AAAI Conf. on Artif. Intell. 35, 14902–14910 (2021)

  44. [53]

    Chen, X. et al. Rainnet: A large-scale imagery dataset and benchmark for spatial precipitation downscaling. In Neural Information Processing Systems (2020)

  45. [54]

    J., Sobie, S

    Cannon, A. J., Sobie, S. R. & Murdock, T. Q. Bias correction of gcm precipitation by quantile mapping: how well do methods preserve changes in quantiles and extremes? J. Clim. 28, 6938–6959 (2015). 15/17 Figure legends Figure 1. Map of ground-based radar stations. The map show...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.