Pith. sign in

REVIEW 4 major objections 6 minor 49 references

From Deterministic to Generative Deep Learning for Urban Air Quality Reconstruction from Sparse Observations

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A diffusion model trained entirely on chemistry-transport simulations can reconstruct full multi-pollutant air-quality fields over Paris from sparse station data, and its real-world fields retain the reference simulation's spatial spectrum.

desk verdict A solid but overclaimed benchmark: diffusion improves spectral realism on simulated and real data, yet 'realistic' rests on the same simulation used for training and on 9-28 stations, so the main claim needs an independent check. read the letter →

arxiv 2607.25687 v1 pith:JUDHSJHQ submitted 2026-07-28 cs.LG cs.AI

Abhishek A.Sabnis , Mihai Mitrea , Lya Lugon , Karine Sartelet , Marc Bocquet , Xiaoyuan Cheng , Shupeng Zhu , Sibo Cheng This is my paper · ORCID
classification cs.LGcs.AI
keywords airpollutionreconstructiondiffusionmodelsgenerativeAIsparseobservationsVoronoitessellationdataaugmentationmulti-pollutantmappingurbanquality
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks whether a generative diffusion model can reconstruct full maps of four air pollutants—NO2, O3, PM2.5, and PM10—over Paris from the readings of a sparse, noisy sensor network. The authors train deterministic baselines and a diffusion model on ten months of full-field chemistry-transport simulations, then evaluate on held-out simulations and on real November–December 2014 station data without retraining. Their central claim is that the diffusion model, conditioned on Voronoi tessellations of the sparse observations, produces reconstructions whose spatial structure matches the reference simulations far better than kriging or deterministic neural networks, and that this structural realism survives the shift from simulation to real data. They also show that ensembling the stochastic samples improves accuracy and that noise-based data augmentation reduces the simulation-to-reality gap. The contribution matters because full-field pollution maps are needed for exposure and health studies, and operational deployment would require this kind of training-on-simulation, inference-on-reality transfer.

What carries the argument

The load-bearing mechanism is a Voronoi-conditioned score-based diffusion model. The observation mask is turned into a Voronoi tessellation of the concentration field, so the network sees a piecewise-constant approximation of the true field rather than scattered point values; a transformer encoder embeds this tessellation and injects it into the denoiser's UNet through cross-attention. During reverse sampling, masked back-sampling re-noises the observed values at each step and reinserts them at sensor locations, pinning the sample to the data; finally, E samples are averaged. The paper demonstrates that this machinery generates fields coarse-to-fine in frequency, matching the reference simulation's power spectrum, and that ensemble averaging removes stochastic variance while preserving structural fidelity.

What would settle it

Take a period with a temporary dense sensor network (many more than the 9–28 permanent stations) or high-resolution satellite retrievals over Paris, reconstruct the field with the trained diffusion model, and compare the measured dense field's radially averaged power spectrum and pointwise values against the reconstruction; if the dense measurements show the simulation spectrum to be wrong at middle or high frequencies, or if the diffusion model's spectral alignment disappears when compared against the actual dense field rather than the simulation, the central claim fails.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that air-quality field reconstruction from sparse observations is better posed as conditional probabilistic generation than as deterministic regression. A score-based diffusion denoiser is conditioned on Voronoi tessellations of the current and past sparse observations through cross-attention, and at every reverse-sampling step the true sensor readings are re-noised and reinserted at observed locations via masked back-sampling. This yields an ensemble of plausible full fields; averaging E=20 samples with R=10 fast sampling steps outperforms the best deterministic model on simulation SSIM, and on real-world data the diffusion output's radially averaged power spectrum stays close to the chemistry-transport reference spectrum, while deterministic models show excess high-frequency power, which the authors read as sensitivity to noise and hallucinated texture. On the real station data the diffusion model reaches MRE 0.249, behind only CLSTM (0.228), and ahead of VUNet (0.264), ViTAE (0.300), and kriging (0.282).

Load-bearing premise

The whole transfer rests on the assumption that the chemistry-transport simulation fields are a faithful proxy for true Paris pollution fields, and that the hand-designed noise augmentations (Gaussian, Perlin, correlated, time-aware) fully capture the remaining difference between simulation and reality.

Editorial extensions

If this is right

  • Operationally, a model trained on simulations from January to October can be applied to real station data from November to December without retraining, so new time periods need only the monitoring data stream.
  • Inference is fast enough for near-real-time use: with the fast sampler at R=10 steps and E=20 ensemble members, the diffusion model produces a full multi-pollutant field quickly, unlike kriging or data assimilation which need error priors and are more expensive.
  • Because the generative model samples rather than regresses, its ensemble gives a set of plausible fields, not one map; the paper uses the average reconstruction for accuracy but positions the spread as useful for uncertainty quantification and data assimilation.
  • The frequency analysis implies a task-dependent choice: diffusion reconstructs the sharp fields of short-lived pollutants (NO2, O3) better, while deterministic models and the diffusion model trade off on smooth particulate fields, so pollutant-specific ensembling is a natural deployment strategy.
  • Adding any of the four noise augmentations during training lowers real-world MRE for every model, indicating the simulation-to-reality gap is reducible at training time without new simulations.

Reading between the lines

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

  • Because Voronoi conditioning and masked back-sampling make no Paris-specific assumption, the same training-on-simulation, inference-on-sensors pipeline should transfer to reconstructing other geophysical fields—temperature, soil moisture, water quality—whenever a full-field simulator and a sparse in situ network exist.
  • The spectral comparison suggests a general hallucination test for learned field reconstruction: if a method's radially averaged power spectrum drifts away from a trusted reference under input noise, its fine-scale textures are likely artifacts even if pointwise errors are small.
  • The paper's own numbers show a deployment fork: held-out station MRE favors CLSTM (0.228) over diffusion (0.249), while spectral realism favors diffusion; a careful reader should treat 'best model' as task-dependent, with the generative model's case resting on structure and uncertainty, not pointwise superiority.
  • A direct testable extension would be to calibrate the diffusion ensemble spread against observed station error; if the spread predicted actual reconstruction error, the samples could serve as a formal uncertainty product for exposure studies, something the paper does not yet demonstrate.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes a diffusion-based generative model for reconstructing full spatial fields of four air pollutants (NO2, O3, PM2.5, PM10) over Paris from sparse monitoring observations. Models are trained on ten months of Polyphemus/Polair3D simulation fields and evaluated on the remaining two months of simulation data as well as on real observations from 9 to 28 monitoring stations. The method conditions a UNet denoiser on Voronoi tessellations of the sparse observations and uses masked back-sampling to enforce observed values during reverse diffusion. Deterministic baselines (VUNet, ViTAE, CLSTM) and Kriging are compared. Data augmentation techniques (Gaussian, Perlin, correlated, time-aware Gaussian noise) are introduced to reduce the simulation-to-real distribution gap. The central claims are that the diffusion model achieves high structural similarity on simulated validation data, produces realistic spatial patterns on real-world observations as judged by power-spectrum analysis, and generalizes to real data without retraining.

Significance. If the transfer and realism claims were rigorously established, the study would be a useful contribution to urban air-quality mapping with potential operational value. Strengths include the evaluation on real observations, the systematic comparison of deterministic and generative approaches, the use of masked back-sampling to respect observational constraints, and the public availability of code and data. However, the current evaluation protocol contains circularities and selection-on-test-set issues that substantially weaken the evidence for the central claims. The paper demonstrates that diffusion models can be trained on simulation data and applied to real observations, but the claims of realistic spatial structure on real data and of unbiased generalization need stronger support.

major comments (4)
  1. [§2.1/Table 1 and §2.2/Figure 2] Model and hyperparameter selection is performed on the same Nov–Dec simulation holdout that is later used to report the headline metrics. The per-model optimal window length k is read from Table 1, and the ensemble size E and inference steps R are chosen from Figure 2 using the same holdout; Table C2 then reports these selected configurations as the final results. This double use of the same data for selection and reporting introduces positive selection bias, so the reported MRE/SSIM values are not unbiased estimates of generalization performance. Please split the simulation data into train/validation/test and report metrics on a test set that is untouched by any hyperparameter selection.
  2. [§4.2 and Figure 4] The augmentation mean shift μ_noise = μ_obs − μ_sim is computed from real observation statistics, and the noise standard deviation is selected by randomized search on the VUNet model. Because these hyperparameters are calibrated to the same real-data distribution used for the evaluation in Figure 4, the comparison partly measures how well the augmentation matches the target distribution rather than how well the model generalizes across distributions. Please state explicitly that the real observations used for μ_obs are restricted to the Jan–Oct training period, and select all augmentation hyperparameters on a validation period disjoint from the Nov–Dec test period.
  3. [§2.3/Figure 5a] The claim that the diffusion model produces realistic spatial patterns on real-world observations is supported by comparing the predicted power spectrum with the Polyphemus simulation spectrum. Since the model was trained on Polyphemus, this comparison demonstrates consistency with the training prior rather than with independent reality; the dashed lines in Figure 5a are not an independent reference. The stress-test concern that this is circular is therefore well founded. Please provide an independent full-field reference (e.g., a different simulation configuration, satellite retrievals, or a denser observational network) or revise the claim to say that the spatial structure is consistent with the simulation prior.
  4. [§2.3/Eq. (E7)] The real-data evaluation uses MRE computed at only one inner-city and one outer-city held-out station per time step, with 9 to 28 stations total per pollutant across the test period. This metric cannot constrain errors in spatial structure away from the stations. Moreover, no confidence intervals or significance tests are reported, so the model differences in Figure 4 (e.g., CLSTM 0.228 vs Diffusion 0.249 vs VUNet 0.264) may not be statistically distinguishable. Please report bootstrap or other interval estimates and discuss the limited spatial coverage of the evaluation explicitly.
minor comments (6)
  1. [§4.3.1, Eq. (1)] The text says the input is a concatenation of the Voronoi tessellation and its corresponding masks (z_t and Ω_t), but the equation uses x_t ⊙ Ω_t (the sparse observed values) rather than the binary mask Ω_t. Please reconcile the notation.
  2. [§4.3.2, Eq. (7)] The re-noising step uses x_t^0 (the clean field) plus σ_{r-1} ε, but in the reverse diffusion process the observation insertion at step r-1 should use the appropriate noise level; please clarify what x_t^0 denotes here (the observed values or the full simulation field).
  3. [§4.2 and Appendix D] The notation for the mask alternates between Ω^t (Eqs. D5 and 8) and Ω_t (Section 4.2); please use a single consistent convention.
  4. [Table 1] The first entry under 'MFB≊0' for k=1 appears as '0 0.012'; this formatting suggests a stray character. Please check the table layout.
  5. [Appendix F.4] The cross-correlation perturbation uses training-set statistics, but it is not stated whether these statistics are computed on the simulation training period only or on the full dataset; please clarify to avoid any look-ahead.
  6. [§2.2] The statement that Figure 2 plots the reference simulation against the predicted values is misleading; the figure shows metric curves and confidence ellipses, not a direct scatter of all grid points. Please rephrase.

Circularity Check

1 steps flagged · score 4.0 of 10

Realism on real-world data is scored against the simulation used for training, so the spatial-structure claim partially reduces to the training distribution; station MRE remains an independent check.

  1. self definitional [Section 2.3, Figure 5a; Appendix D.1, Eq. D3]
    "We plot the reference simulation data (considered as the ground truth when using synthetic observations) against the outputs of the generative and deterministic models evaluated on synthetic and real observations. ... By contrast, the diffusion model remains closely aligned with the simulation spectrum, highlighting its greater robustness to noisy observations and its ability to preserve realistic spatial structure."

    The 'realistic spatial structure' of real-world reconstructions is measured by spectral agreement with the Polyphemus simulation. The diffusion denoiser is trained by minimizing E||Dθ(x+n;σ)-x||² with x drawn from that same simulation (Eq. D3), so matching the simulation's power spectrum is a direct consequence of optimizing the training objective. Reporting that real-data outputs stay 'closely aligned with the simulation spectrum' therefore confirms that the model maps real inputs into its training distribution; it does not independently validate agreement with the true Parisian fields. The held-out station MRE is an external check, but with 9-28 stations per pollutant it cannot constrain the full power spectrum away from the stations, leaving the spatial-realism claim partially circular.

full rationale

The paper's central comparison against held-out real monitoring stations (Eq. E7) is external and independent: models trained on Polyphemus simulation fields are evaluated at real LCSQA/Geod'air stations withheld from the input, so the pointwise generalization claim does not reduce to the training data. The data-augmentation mean shift (μ_noise = μ_obs − μ_sim, Section 4.2) is transparent and does not by itself force the held-out MRE. The partial circularity is confined to the spatial-realism claim: Fig. 5a scores real-data reconstructions by their power-spectrum agreement with 'the reference simulation data,' which is the same distribution used as the training target in Eq. D3. A model trained to minimize reconstruction error against simulation will, by construction, tend to match the simulation spectrum, so 'realistic spatial structure on real-world observations' is partly a restatement that the model maps real inputs into its training distribution. Because only 9-28 stations are available, the station MRE cannot independently constrain the full-field spectrum, leaving that specific claim dependent on the very simulation used for training. The paper's Discussion acknowledges 'dependence on training with simulation data' as a limitation, which is consistent with this assessment. Self-citations to Refs. [28] and [41] (with overlapping authorship) are present but are used for architecture and uncertainty-quantification context, not as load-bearing proof of the main result.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central benchmark rests on two fitted quantities (augmentation mean and variance) plus several hand-chosen hyperparameters, all of which affect the reported real-world transfer. The most important assumption is that the simulation plus additive noise covers the real distribution. No new physical entities are introduced.

free parameters (6)
  • Augmentation noise standard deviation = not reported
    Controls perturbation strength; selected via randomized search on the VUNet model, not derived from first principles (Section 4.2, Appendix F).
  • Augmentation mean shift mu_noise = mu_obs - mu_sim
    Computed from all sensors and time steps in the training period to align simulated and observed distributions (Section 4.2).
  • Diffusion ensemble size E = 20
    Chosen from the validation plateau in Figure 2; a computational hyperparameter affecting output variance (Section 2.2).
  • Diffusion inference steps R = 10
    Chosen as a trade-off between quality and latency (Section 2.2).
  • Gaussian and time-aware Gaussian kernel sigmas = not reported
    Determine spatial and temporal smoothness of noise perturbations (Appendix F.1, F.2).
  • Perlin noise parameters (base resolution, octaves, persistence, lacunarity) = not reported
    Control coarseness and fractal structure of Perlin perturbations (Appendix F.3).
assumptions (5)
  • domain assumption Polyphemus/Polair3D simulation provides full-field ground truth representative of real Paris air quality at 2 km resolution.
    All training labels and the spectral reference on real data derive from this simulation (Sections 4.1, 2.1, Figure 5a); simulation bias would directly bias the learned reconstructions.
  • ad hoc to paper The noise augmentation family (Gaussian, Perlin, correlated, time-aware) is sufficient to close the simulation-to-real distribution gap.
    The paper assumes additive noise with tuned mean and variance makes simulated inputs statistically similar to real observations; there is no independent check that this captures all relevant distribution shifts (Section 4.2, Figure 8).
  • domain assumption Held-out station observations (one inner-city, one outer-city per time step) provide an unbiased estimate of full-field reconstruction error.
    Real-world MRE is computed at 9 to 28 held-out stations and extrapolated to claims about full-field quality (Section 4.2, Figure 7).
  • standard math Standard score-based diffusion SDE theory, EDM, and DPM-Solver++ apply correctly to 4x75x110 concentration fields with the stated training objective.
    Relied on for the generative model's formulation and fast sampling (Section 4.3.2, Appendix D).
  • domain assumption The 2014 simulation period and Geod'air observations are temporally and spatially comparable at the grid resolution used.
    Training on Jan-Oct simulation and testing on Nov-Dec real observations assumes consistent meteorology and emissions (Sections 4.1, 4.2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Deterministic to Generative Deep Learning for Urban Air Quality Reconstruction from Sparse Observations." pith.science (2026). https://pith.science/paper/JUDHSJHQ

@misc{pith2026260725687,
  author       = {Pith},
  title        = {Pith review of: From Deterministic to Generative Deep Learning for Urban Air Quality Reconstruction from Sparse Observations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JUDHSJHQ}},
  note         = {Machine review of arXiv:2607.25687}
}
read the original abstract

Full-field reconstruction of air pollution is essential for evaluating pollution exposure and supporting public health decision-making. However, the complex interactions among pollutants, hard-to-predict weather patterns, and limited monitoring station coverage make this a complex task. We apply deep learning techniques to provide fast and accurate reconstructions from sparse observations of four key pollutants: NO2, O3, PM2.5 and PM10. Models are trained on full-field simulation data and evaluated on real-world observations collected from 9 to 28 monitoring stations in the city of Paris. We introduce a diffusion-based generative framework for multi-pollutant reconstruction and benchmark its performance against deterministic deep learning models. Despite noisy observations and strong spatial variability, the models achieve high structural similarity on simulated validation data and produce realistic spatial patterns on real-world observations, as indicated by power-spectrum analysis. We introduce data augmentation methods that enable transfer to real-world observations without retraining, allowing the models to generalise beyond the training period. These findings highlight the potential of ML models for reliable real-world deployment in air pollution reconstruction tasks.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 37 canonical work pages

  1. [1]

    PloS one10(9), 0138146 (2015)

    Zheng, X.-y., Ding, H., Jiang, L.-n., Chen, S.-w., Zheng, J.-p., Qiu, M., Zhou, Y.-x., Chen, Q., Guan, W.-j.: Association between air pollutants and asthma emergency room visits and hospital admissions in time series studies: a systematic review and meta-analysis. PloS one10(9), 0138146 (2015)

    work page 2015

  2. [2]

    5 exposure and risks of ischemic heart disease and stroke events: review and meta-analysis

    Alexeeff, S.E., Liao, N.S., Liu, X., Van Den Eeden, S.K., Sidney, S.: Long-term pm2. 5 exposure and risks of ischemic heart disease and stroke events: review and meta-analysis. Journal of the American Heart Association10(1), 016890 (2021)

    work page 2021

  3. [3]

    New England Journal of Medicine376(26), 2513–2522 (2017) https://doi.org/10.1056/ NEJMoa1702747 https://www.nejm.org/doi/pdf/10.1056/NEJMoa1702747

    Di, Q., Wang, Y., Zanobetti, A., Wang, Y., Koutrakis, P., Choirat, C., Dominici, F., Schwartz, J.D.: Air pollution and mortality in the medicare population. New England Journal of Medicine376(26), 2513–2522 (2017) https://doi.org/10.1056/ NEJMoa1702747 https://www.nejm.org/doi/pdf/10.1056/NEJMoa1702747

    work page doi:10.1056/nejmoa1702747 2017

  4. [4]

    Science of The Total Environment858, 160064 (2023) https: //doi.org/10.1016/j.scitotenv.2022.160064 27

    Sicard, P., Agathokleous, E., Anenberg, S.C., De Marco, A., Paoletti, E., Calatayud, V.: Trends in urban air pollution over the last two decades: A global perspective. Science of The Total Environment858, 160064 (2023) https: //doi.org/10.1016/j.scitotenv.2022.160064 27

    work page arXiv 2023

  5. [5]

    Environmental science & technology50(1), 79–88 (2016)

    Brauer, M., Freedman, G., Frostad, J., Van Donkelaar, A., Martin, R.V., Dentener, F., Dingenen, R.v., Estep, K., Amini, H., Apte, J.S.,et al.: Ambi- ent air pollution exposure estimation for the global burden of disease 2013. Environmental science & technology50(1), 79–88 (2016)

    work page 2016

  6. [6]

    American Economic Review114(5), 1338–1381 (2024)

    Barwick, P.J., Li, S., Lin, L., Zou, E.Y.: From fog to smog: The value of pollution information. American Economic Review114(5), 1338–1381 (2024)

    work page 2024

  7. [7]

    Science of The Total Environment571, 416–425 (2016) https://doi.org/10.1016/j.scitotenv.2016.06.213

    Pascal, M., de Crouy Chanel, P., Wagner, V., Corso, M., Tillier, C., Bentayeb, M., Blanchard, M., Cochet, A., Pascal, L., Host, S., Goria, S., Le Tertre, A., Chatignoux, E., Ung, A., Beaudeau, P., Medina, S.: The mortality impacts of fine particles in France. Science of The Total Environment571, 416–425 (2016) https://doi.org/10.1016/j.scitotenv.2016.06.213

    work page doi:10.1016/j.scitotenv.2016.06.213 2016

  8. [8]

    Environment International199, 109474 (2025) https://doi.org/10.1016/j.envint.2025.109474

    Sartelet, K., Kerckhoffs, J., Athanasopoulou, E., Lugon, L., Vasilescu, J., Zhong, J., Hoek, G., Joly, C., Park, S.-J., Talianu, C., van den Elshout, S., Dugay, F., Gerasopoulos, E., Ilie, A., Kim, Y., Nicolae, D., Harrison, R.M., Petaja, T.: Air pollution mapping and variability over five European cities. Environment International199, 109474 (2025) https...

    arXiv 2025

Show all 49 references
  1. [9]

    Sustainable Cities and Society54, 101997 (2020) https://doi

    Shukla, K., Kumar, P., Mann, G.S., Khare, M.: Mapping spatial distribution of particulate matter using Kriging and Inverse Distance Weighting at supersites of megacity Delhi. Sustainable Cities and Society54, 101997 (2020) https://doi. org/10.1016/j.scs.2019.101997

    2020

  2. [10]

    Journal of Geophysical Research: Atmospheres 118(4), 2031–2040 (2013) https://doi.org/10.1002/jgrd.50233

    Tilloy, A., Mallet, V., Poulet, D., Pesin, C., Brocheton, F.: BLUE-based NO 2 data assimilation at urban scale. Journal of Geophysical Research: Atmospheres 118(4), 2031–2040 (2013) https://doi.org/10.1002/jgrd.50233

    2013 doi

  3. [11]

    Atmosphere15(12), 1523 (2024)

    Hayward, I., Martin, N.A., Ferracci, V., Kazemimanesh, M., Kumar, P.: Low- cost air quality sensors: biases, corrections and challenges in their comparability. Atmosphere15(12), 1523 (2024)

    2024

  4. [12]

    Environmental Monitoring and Assessment198(1), 40 (2026)

    Stowe, S., Bohra, R., Vilcassim, M.: Assessing the efficacy of a low-cost air pollu- tion monitoring device for environmental and occupational exposure assessments. Environmental Monitoring and Assessment198(1), 40 (2026)

    2026

  5. [13]

    Scientific Reports12(1), 12215 (2022)

    He, H., Sch¨ afer, B., Beck, C.: Spatial heterogeneity of air pollution statistics in europe. Scientific Reports12(1), 12215 (2022)

    2022

  6. [14]

    Geoscientific Model Development 10(10), 3695–3713 (2017)

    Hase, N., Miller, S.M., Maaß, P., Notholt, J., Palm, M., Warneke, T.: Atmospheric inverse modeling via sparse reconstruction. Geoscientific Model Development 10(10), 3695–3713 (2017)

    2017

  7. [15]

    Rodgers, C.D.: Inverse Methods for Atmospheric Sounding: Theory and Practice vol. 2. World scientific, Singapore (2000) 28

    2000

  8. [16]

    In: 2020 IEEE International Conference on Big Data and Smart Computing (BigComp), pp

    Le, V.-D., Bui, T.-C., Cha, S.-K.: Spatiotemporal deep learning model for citywide air pollution interpolation and prediction. In: 2020 IEEE International Conference on Big Data and Smart Computing (BigComp), pp. 55–62 (2020). IEEE

    2020

  9. [17]

    In: 2022 23rd IEEE Inter- national Conference on Mobile Data Management (MDM), pp

    Sasaki, Y., Harada, K., Yamasaki, S., Onizuka, M.: Airex: Neural network-based approach for air quality inference in unmonitored cities. In: 2022 23rd IEEE Inter- national Conference on Mobile Data Management (MDM), pp. 119–124 (2022). IEEE

    2022

  10. [18]

    arXiv preprint arXiv:2403.02354 (2024)

    Feng, Y., Wang, Q., Xia, Y., Huang, J., Zhong, S., Liang, Y.: Spatio-temporal field neural networks for air quality inference. arXiv preprint arXiv:2403.02354 (2024)

    2024 arXiv

  11. [19]

    Seinfeld, J.H., Pandis, S.N.: Atmospheric Chemistry and Physics: from Air Pol- lution to Climate Change, pp. 131–165. John Wiley & Sons, Hoboken, NJ (2016)

    2016

  12. [20]

    Nature Machine Intelligence5(11), 1317–1325 (2023)

    Santos, J.E., Fox, Z.R., Mohan, A., O’Malley, D., Viswanathan, H., Lubbers, N.: Development of the senseiver for efficient field reconstruction from sparse observations. Nature Machine Intelligence5(11), 1317–1325 (2023)

    2023

  13. [21]

    arXiv preprint arXiv:2011.13456 (2020)

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

    2020 arXiv

  14. [22]

    ACM computing surveys56(4), 1–39 (2023)

    Yang, L., Zhang, Z., Song, Y., Hong, S., Xu, R., Zhao, Y., Zhang, W., Cui, B., Yang, M.-H.: Diffusion models: A comprehensive survey of methods and applications. ACM computing surveys56(4), 1–39 (2023)

    2023

  15. [23]

    In: 2023 IEEE 39th International Conference on Data Engineering (ICDE), pp

    Liu, M., Huang, H., Feng, H., Sun, L., Du, B., Fu, Y.: Pristi: A conditional diffu- sion framework for spatiotemporal imputation. In: 2023 IEEE 39th International Conference on Data Engineering (ICDE), pp. 1927–1939 (2023). IEEE

    2023

  16. [24]

    ISPRS International Journal of Geo-Information15(4), 171 (2026)

    Zheng, G., Wan, Y., Zhang, X., Liu, X.: Noise-aware diffusion for city-scale air-quality reconstruction from sparse monitoring stations. ISPRS International Journal of Geo-Information15(4), 171 (2026)

    2026

  17. [25]

    Atmospheric Pollution Research13(5), 101365 (2022)

    Lugon, L., Kim, Y., Vigneron, J., Chr´ etien, O., Andr´ e, M., Andr´ e, J.-M., Moukhtar, S., Redaelli, M., Sartelet, K.: Effect of vehicle fleet composition and mobility on outdoor population exposure: A street resolution analysis in paris. Atmospheric Pollution Research13(5),...

    2022

  18. [26]

    Nature Machine Intelligence3(11) (2021)

    Fukami, K., Maulik, R., Ramachandra, N., Fukagata, K., Taira, K.: Global field reconstruction from sparse sensors with voronoi tessellation-assisted deep learning. Nature Machine Intelligence3(11) (2021)

    2021

  19. [27]

    In: International Conference on Medical Image Computing and Computer-assisted Intervention, pp

    Ronneberger, O., Fischer, P., Brox, T.: U-net: Convolutional networks for 29 biomedical image segmentation. In: International Conference on Medical Image Computing and Computer-assisted Intervention, pp. 234–241 (2015). Springer

    2015

  20. [28]

    Computer Physics Communications308(2025)

    Fan, H., Cheng, S., Nazelle, A.J., Arcucci, R.: Vitae-sl: A vision transformer-based autoencoder and spatial interpolation learner for field reconstruction. Computer Physics Communications308(2025)

    2025

  21. [29]

    In: Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks

    Gers, F.A., Schmidhuber, J.: Recurrent nets that time and count. In: Proceedings of the IEEE-INNS-ENNS International Joint Conference on Neural Networks. IJCNN 2000. Neural Computing: New Challenges and Perspectives for the New Millennium, vol. 3, pp. 189–194 (2000). https://d...

    2000 doi

  22. [30]

    Advances in neural information processing systems28(2015)

    Shi, X., Chen, Z., Wang, H., Yeung, D.-Y., Wong, W.-K., Woo, W.-c.: Convolu- tional lstm network: A machine learning approach for precipitation nowcasting. Advances in neural information processing systems28(2015)

    2015

  23. [31]

    European Journal of Operational Research192(3) (2009)

    Kleijnen, J.P.: Kriging metamodeling in simulation: A review. European Journal of Operational Research192(3) (2009)

    2009

  24. [32]

    Machine Intelligence Research, 1–22 (2025)

    Lu, C., Zhou, Y., Bao, F., Chen, J., Li, C., Zhu, J.: Dpm-solver++: Fast solver for guided sampling of diffusion probabilistic models. Machine Intelligence Research, 1–22 (2025)

    2025

  25. [33]

    arXiv preprint arXiv:2206.13397 (2022)

    Rissanen, S., Heinonen, M., Solin, A.: Generative modelling with inverse heat dissipation. arXiv preprint arXiv:2206.13397 (2022)

    2022 arXiv

  26. [34]

    https:// www.geodair.fr/donnees/consultation

    Laboratoire central de surveillance de la qualit’e de l’air: GEOD’AIR. https:// www.geodair.fr/donnees/consultation

  27. [35]

    5 and network activity during extreme pollution events

    Bashan, N.F., Li, W., Wang, Q.R.: Dynamics of pm2. 5 and network activity during extreme pollution events. npj Climate and Atmospheric Science7(1), 171 (2024)

    2024

  28. [36]

    Atmospheric Environment 41(29), 6116–6131 (2007) https://doi.org/10.1016/j.atmosenv.2007.04.024

    Sartelet, K.N., Debry, E., Fahey, K., Roustan, Y., Tombette, M., Sportisse, B.: Simulation of aerosols and gas-phase species over Europe with the Polyphemus system: Part I - Model-to-data comparison for 2001. Atmospheric Environment 41(29), 6116–6131 (2007) https://doi.org/10....

    2007 doi

  29. [37]

    Atmosphere11(5) (2020) https://doi.org/10.3390/atmos11050525

    Sartelet, K., Couvidat, F., Wang, Z., Flageul, C., Kim, Y.: SSH-Aerosol v1.1: A Modular Box Model to Simulate the Evolution of Primary and Secondary Aerosols. Atmosphere11(5) (2020) https://doi.org/10.3390/atmos11050525

    2020 doi

  30. [38]

    Atmospheric Environment241, 117752 (2020) https://doi.org/10.1016/j.atmosenv.2020.117752 30

    Andre, M., Sartelet, K., Moukhtar, S., Andre, J.M., Redaelli, M.: Diesel, petrol or electric vehicles: What choices to improve urban air quality in the Ile-de-France region? A simulation platform and case study. Atmospheric Environment241, 117752 (2020) https://doi.org/10.1016...

    2020

  31. [39]

    Neural computation 9(8), 1735–1780 (1997)

    Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural computation 9(8), 1735–1780 (1997)

    1997

  32. [40]

    Advances in neural information processing systems35, 26565–26577 (2022)

    Karras, T., Aittala, M., Aila, T., Laine, S.: Elucidating the design space of diffusion-based generative models. Advances in neural information processing systems35, 26565–26577 (2022)

    2022

  33. [41]

    Computer Methods in Applied Mechanics and Engineering435, 117623 (2025)

    Zhuang, Y., Cheng, S., Duraisamy, K.: Spatially-aware diffusion models with cross-attention for global field reconstruction with sparse observations. Computer Methods in Applied Mechanics and Engineering435, 117623 (2025)

    2025

  34. [42]

    arXiv preprint arXiv:2304.12891 (2023)

    Leinonen, J., Hamann, U., Nerini, D., Germann, U., Franch, G.: Latent diffu- sion models for generative precipitation nowcasting with accurate uncertainty quantification. arXiv preprint arXiv:2304.12891 (2023)

    2023 arXiv

  35. [43]

    arXiv preprint arXiv:2010.02502 (2020)

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

    2020 arXiv

  36. [44]

    IEEE transactions on image processing13(4), 600–612 (2004)

    Wang, Z., Bovik, A.C., Sheikh, H.R., Simoncelli, E.P.: Image quality assess- ment: from error visibility to structural similarity. IEEE transactions on image processing13(4), 600–612 (2004)

    2004

  37. [45]

    Pearson Education, Upper Saddle River, NJ (2008)

    Gonzalez, R.C., Woods, R.E.: Digital Image Processing, 3rd edn. Pearson Education, Upper Saddle River, NJ (2008)

    2008

  38. [46]

    Williams, C.K., Rasmussen, C.E.: Gaussian Processes for Machine Learning vol

  39. [47]

    MIT press Cambridge, MA, Cambridge, MA (2006)

    2006

  40. [48]

    ACM Siggraph Computer Graphics19(3), 287– 296 (1985)

    Perlin, K.: An image synthesizer. ACM Siggraph Computer Graphics19(3), 287– 296 (1985)

    1985

  41. [49]

    In: Central European Seminar on Computer Graphics, vol

    Man, P.: Generating and real-time rendering of clouds. In: Central European Seminar on Computer Graphics, vol. 1 (2006). Citeseer Cast´ a-Papiernicka, Slovakia 31

    2006

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.