Articles | Volume 26, issue 4
https://doi.org/10.5194/nhess-26-1663-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/nhess-26-1663-2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
14 Apr 2026
Research article |
| 14 Apr 2026
| 14 Apr 2026
Simulating spatial multi-hazards with generative deep learning
Environmental Change Institute, University of Oxford, Oxford, UK
Yu Mo
Environmental Change Institute, University of Oxford, Oxford, UK
Jim W. Hall
Environmental Change Institute, University of Oxford, Oxford, UK
Related authors
No articles found.
Razi Sheikholeslami and Jim W. Hall
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2021-618, https://doi.org/10.5194/hess-2021-618, 2022
Manuscript not accepted for further review
Short summary
Short summary
In this study, we investigated the spatiotemporal variations in global freshwater nitrogen concentrations using a relatively parsimonious data-driven approach based on random forest method. We used the proposed model to identify several hotspots of nitrogen pollution in 115 major river basins of the world. Furthermore, we found that livestock population, nitrogen fertilizer application, temperature, and precipitation are the most influential predictors of nitrogen pollution of the river systems.
Cited articles
Abdelmoaty, H. M., Papalexiou, S. M., Mamalakis, A., Singh, S., Coia, V., Hairabedian, M., Szeftel, P., and Grover, P.: Generative Adversarial Networks for Downscaling Hourly Precipitation in the Canadian Prairies, Journal of Geophysical Research: Machine Learning and Computation, 2, e2025JH000678, https://doi.org/10.1029/2025JH000678, 2025. a
Alkhalidi, M., Al-Dabbous, A., Al-Dabbous, S., and Alzaid, D.: Evaluating the accuracy of the ERA5 model in predicting wind speeds across coastal and offshore regions, Journal of Marine Science and Engineering, 13, 149, https://doi.org/10.3390/jmse13010149, 2025. a
Arjovsky, M., Chintala, S., and Bottou, L.: Wasserstein generative adversarial networks, in: International Conference on Machine Learning, PMLR, 214–223, https://proceedings.mlr.press/v70/arjovsky17a.html (last access: 10 April 2026), 2017. a
Bader, B. and Yan, J.: eva: Extreme Value Analysis with Goodness-of-Fit Testing, r package version 0.2.6, https://doi.org/10.32614/CRAN.package.eva, 2020.
Bader, B., Yan, J., and Zhang, X.: Automated threshold selection for extreme value analysis via ordered goodness-of-fit tests with adjustment for false discovery rate, Ann. Appl. Stat., https://doi.org/10.1214/17-AOAS1092, 2018. a, b
Bailey, R., Benton, T. G., Challinor, A., Elliott, J., Gustafson, D., Hiller, B., Jones, A., Jahn, M., Kent, C., Lewis, K., Meacham, T., Rivington, M., Robson, D., Tiffin, R., and Wuebbles, D. J.: Extreme weather and resilience of the global food system (2015). Final Project Report from the UK–US Taskforce on extreme weather and global food system resilience, The Global Food Security Programme, https://www.foodsecurity.ac.uk/publications/extreme-weather-resilience-global-food-system.pdf (last access: 10 April 2026), 2015. a
Bates, P. D., Savage, J., Wing, O., Quinn, N., Sampson, C., Neal, J., and Smith, A.: A climate-conditioned catastrophe risk model for UK flooding, Nat. Hazards Earth Syst. Sci., 23, 891–908, https://doi.org/10.5194/nhess-23-891-2023, 2023. a, b
Becher, O., Pant, R., Verschuur, J., Mandal, A., Paltan, H., Lawless, M., Raven, E., and Hall, J.: A multi-hazard risk framework to stress-test water supply systems to climate-related disruptions, Earths Future, 11, https://doi.org/10.1029/2022EF002946, 2023. a
Bhatia, S., Jain, A., and Hooi, B.: ExGAN: Adversarial generation of extreme samples, in: Proceedings of the AAAI Conference on Artificial Intelligence, vol. 35, 6750–6758, https://doi.org/10.1609/aaai.v35i8.16834, 2021. a, b
Bunting, P., Rosenqvist, A., Hilarides, L., Lucas, R., Thomas, T., Tadono, T., Worthington, T., Spalding, M., Murray, N., and Rebelo, L.-M.: Global Mangrove Extent Change 1996–2020: Global Mangrove Watch Version 3.0, Remote Sens., https://doi.org/10.3390/rs14153657, 2022. a, b
Coles, S., Heffernan, J. E., and Tawn, J. A.: Dependence measures for extreme value analyses, Extremes, 2, 339–365, https://doi.org/10.1023/A:1009963131610, 1999. a, b
Cook, N.: Towards better estimation of extreme winds, J. Wind Eng. Ind. Aerod., 9, 295–323, 1982. a
Cook, N. J.: Consolidation of analysis methods for sub-annual extreme wind speeds, Meteorol. Appl., 21, 403–414, https://doi.org/10.1002/met.1355, 2014. a
Cook, N. J., Harris, R. I., and Whiting, R.: Extreme wind speeds in mixed climates revisited, J. Wind Eng. Ind. Aerod., 91, 403–422, https://doi.org/10.1016/S0167-6105(02)00397-5, 2003. a
Davison, A. C., Huser, R., and Thibaud, E.: Spatial extremes, in: Handbook of Environmental and Ecological Statistics, edited by: Gelfand, A. E., Fuentes, M., Hoeting, J. A., and Smith, R. L., CRC Press, ISBN 9781498752022, http://hdl.handle.net/10754/631198 (last access: 10 April 2026), 2019. a
de Fondeville, R. and Davison, A. C.: Functional peaks-over-threshold analysis, J. Roy. Stat. Soc. B, 84, 1392–1422, https://doi.org/10.1093/biomet/asy026, 2022. a
Engelke, S. and Ivanovs, J.: Sparse structures for multivariate extremes, Annu. Rev. Stat. Appl., 8, 241–270, https://doi.org/10.1146/annurev-statistics-040620-041554, 2021. a, b
Faraway, J., Marsaglia, G., Marsaglia, J., and Baddeley, A.: goftest: Classical Goodness-of-Fit Tests for Univariate Distributions, r package version 1.2-3, https://doi.org/10.32614/CRAN.package.goftest, 2021.
Gaupp, F., Hall, J., Mitchell, D., and Dadson, S.: Increasing risks of multiple breadbasket failure under 1.5 and 2 °C global warming, Agr. Syst., 175, 34–45, https://doi.org/10.1016/j.agsy.2019.05.010, 2019. a
Gilleland, E. and Katz, R. W.: extRemes 2.0: An Extreme Value Analysis Package in R, J. Stat. Softw., 72, 1–39, https://doi.org/10.18637/jss.v072.i08, 2016. a
Girard, S., Gobet, E., and Pachebat, J.: HTGAN: Heavy-tail GAN for multivariate dependent extremes via latent-dimensional control, International Journal of Computer Mathematics, 1–41, https://doi.org/10.1080/00207160.2025.2578391, 2025. a
Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A., and Bengio, Y.: Generative Adversarial Nets, in: Advances in Neural Information Processing Systems, vol. 27, arXiv [preprint], https://doi.org/10.48550/arXiv.1406.2661, 2014. a
Guillod, B. P., Jones, R. G., Dadson, S. J., Coxon, G., Bussi, G., Freer, J., Kay, A. L., Massey, N. R., Sparrow, S. N., Wallom, D. C. H., Allen, M. R., and Hall, J. W.: A large set of potential past, present and future hydro-meteorological time series for the UK, Hydrol. Earth Syst. Sci., 22, 611–634, https://doi.org/10.5194/hess-22-611-2018, 2018. a
Gulrajani, I., Ahmed, F., Arjovsky, M., Dumoulin, V., and Courville, A. C.: Improved training of Wasserstein GANs, in: Advances in Neural Information Processing Systems, vol. 30, arXiv [preprint], https://doi.org/10.48550/arXiv.1704.00028, 2017. a
Harris, I.: Generalised Pareto methods for wind extremes. Useful tool or mathematical mirage?, J. Wind Eng. Ind. Aerod., 93, 341–360, https://doi.org/10.1016/j.jweia.2005.02.004, 2005. a, b, c
Harris, R. I.: XIMIS, a penultimate extreme value method suitable for all types of wind climate, J. Wind Eng. Ind. Aerod., 97, 271–286, https://doi.org/10.1016/j.jweia.2009.06.011, 2009. a
Hersbach, H., Bell, B., Berrisford, P., Biavati, G., Horányi, A., Muñoz Sabater, J., Nicolas, J., Peubey, C., Radu, R., Rozum, I., Schepers, D., Simmons, A., Soci, C., Dee, D., and Thépaut, J.-N.: ERA5 hourly data on single levels from 1940 to present, Copernicus Climate Change Service (C3S) Climate Data Store (CDS) [data set], https://doi.org/10.24381/cds.adbb2d47, 2023a. a, b
Hersbach, H., Bell, B., Berrisford, P., Biavati, G., Horányi, A., Muñoz Sabater, J., Nicolas, J., Peubey, C., Radu, R., Rozum, I., Schepers, D., Simmons, A., Soci, C., Dee, D., and Thépaut, J.-N.: ERA5 hourly data on pressure levels from 1940 to present, Copernicus Climate Change Service (C3S) Climate Data Store (CDS) [data set], https://doi.org/10.24381/cds.bd0915c6, 2023b. a
Ho, J., Jain, A., and Abbeel, P.: Denoising diffusion probabilistic models, in: Advances in Neural Information Processing Systems, vol. 33, 6840–6851, arXiv, https://doi.org/10.48550/arXiv.2006.11239, 2020. a
Hollis, D., McCarthy, M., Kendon, M., Legg, T., and Simpson, I.: HadUK-Grid – A new UK dataset of gridded climate observations, Geosci. Data J., 6, 151–159, 2019. a
Hunt, K. M. R. and Bloomfield, H. C.: Landfalling tropical cyclones significantly reduce Bangladesh's energy security, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2025-4474, 2025. a
Huser, R. and Wadsworth, J. L.: Advances in statistical modeling of spatial extremes, Wiley Interdisciplinary Reviews: Computational Statistics, 14, e1537, https://doi.org/10.1002/wics.1537, 2022. a, b, c, d
Huser, R., Opitz, T., and Wadsworth, J. L.: Modeling of spatial extremes in environmental data science: Time to move away from max-stable processes, Environmental Data Science, 4, e3, https://doi.org/10.1017/eds.2024.54, 2025. a, b
Huster, T., Cohen, J., Lin, Z., Chan, K., Kamhoua, C., Leslie, N. O., Chiang, C.-Y. J., and Sekar, V.: Pareto GAN: Extending the representational power of GANs to heavy-tailed distributions, in: International Conference on Machine Learning, PMLR, 4523–4532, arXiv [preprint], https://doi.org/10.48550/arXiv.2101.09113, 2021. a, b, c, d, e, f, g
Islam, T. and Peterson, R. E.: Climatology of landfalling tropical cyclones in Bangladesh 1877–2003, Nat. Hazards, 48, 115–135, https://doi.org/10.1007/s11069-008-9252-4, 2009. a
Keef, C., Tawn, J., and Svensson, C.: Spatial dependence in extreme river flows and precipitation in Great Britain, J. Hydrol., 378, 240–252, https://doi.org/10.1016/j.jhydrol.2009.09.026, 2009. a
Kingma, D. P. and Welling, M.: Auto-Encoding Variational Bayes, arXiv [preprint], https://doi.org/10.48550/arXiv.1312.6114, 2013. a
Kobyzev, I., Prince, S. J., and Brubaker, M. A.: Normalizing flows: An introduction and review of current methods, IEEE T. Pattern Anal., 43, 3964–3979, arXiv, https://doi.org/10.48550/arXiv.1908.09257, 2021. a
Lamb, R., Keef, C., Tawn, J., Laeger, S., Meadowcroft, I., Surendran, S., Dunning, P., and Batstone, C.: A new method to assess the risk of local and widespread flooding on rivers and coasts, J. Flood Risk Manag., 3, 323–336, https://doi.org/10.1111/j.1753-318X.2010.01081.x, 2010. a, b
Lamb, R., Garside, P., Pant, R., and Hall, J. W.: A probabilistic model of the economic risk to Britain's railway network from bridge scour during floods, Risk Anal., 39, 2457–2478, https://doi.org/10.1111/risa.13370, 2019. a
Ledig, C., Theis, L., Huszár, F., Caballero, J., Cunningham, A., Acosta, A., Aitken, A., Tejani, A., Totz, J., Wang, Z., and Wenzhe, S.: Photo-realistic single image super-resolution using a generative adversarial network, in: Proceedings of the IEEE conference on computer vision and pattern recognition, 4681–4690, arXiv [preprint], https://doi.org/10.48550/arXiv.1609.04802, 2017. a
Lhaut, S., Rootzén, H., and Segers, J.: Wasserstein–Aitchison GAN for angular measures of multivariate extremes, arXiv [preprint], https://doi.org/10.48550/arXiv.2504.21438, 2025. a
Liu, B., Zhu, Y., Song, K., and Elgammal, A.: Towards faster and stabilized GAN training for high-fidelity few-shot image synthesis, in: International Conference on Learning Representations, https://openreview.net/forum?id=1Fqg133qRaI (last access: 10 April 2026), 2021. a
Ljung, G. M. and Box, G. E.: On a measure of lack of fit in time series models, Biometrika, 65, 297–303, https://doi.org/10.1093/biomet/65.2.297, 1978. a
Lloyd's: RDS 2025: Realistic Disaster Scenarios – Scenario Specification, Tech. Rep. EM 518 v1.0, Lloyd's Exposure Management, https://www.lloyds.com/market-resources/underwriting/realistic-disaster-scenarios (last access: 10 April 2026), 2025. a
Meiler, S., Vogt, T., Bloemendaal, N., Ciullo, A., Lee, C.-Y., Camargo, S. J., Emanuel, K., and Bresch, D. N.: Intercomparison of regional loss estimates from global synthetic tropical cyclone models, Nat. Commun., 13, 6156, https://doi.org/10.1038/s41467-022-33918-1, 2022. a, b
Metin, A. D., Dung, N. V., Schröter, K., Vorogushyn, S., Guse, B., Kreibich, H., and Merz, B.: The role of spatial dependence for large-scale flood risk estimation, Nat. Hazards Earth Syst. Sci., 20, 967–979, https://doi.org/10.5194/nhess-20-967-2020, 2020. a
Mo, Y., Simard, M., and Hall, J. W.: Tropical cyclone risk to global mangrove ecosystems: potential future regional shifts, Front. Ecol. Environ., 21, 269–274, https://doi.org/10.1002/fee.2650, 2023. a
Nelson, R.: An Introduction to Copulas, Springer, New York, ISBN 9780387286594, 2006. a
Obakrim, S., Benoit, L., and Allard, D.: A multivariate and space-time stochastic weather generator using a latent Gaussian framework, Stoch. Env. Res. Risk A., 39, 3677–3701, https://doi.org/10.1007/s00477-024-02897-8, 2025. a
Papalexiou, S. M., Serinaldi, F., and Porcu, E.: Advancing Space-Time Simulation of Random Fields: From Storms to Cyclones and Beyond, Water Resour. Res., 57, e2020WR029466, https://doi.org/10.1029/2020WR029466, 2021. a
Peard, A.: Supporting code and data: Simulating spatial multi-hazards with generative deep learning (0.2.0), Zenodo [data set], https://doi.org/10.5281/zenodo.19455813, 2026. a
Quinn, N., Bates, P. D., Neal, J., Smith, A., Wing, O., Sampson, C., Smith, J., and Heffernan, J.: The spatial dependence of flood hazard and risk in the United States, Water Resour. Res., 55, 1890–1911, https://doi.org/10.1029/2018WR024205, 2019. a, b, c
Radford, A., Metz, L., and Chintala, S.: Unsupervised representation learning with deep convolutional generative adversarial networks, arXiv [preprint], https://doi.org/10.48550/arXiv.1511.06434, 2015. a
Rani, S. I., Arulalan, T., George, J. P., Rajagopal, E., Renshaw, R., Maycock, A., Barker, D. M., and Rajeevan, M.: IMDAA: High-resolution satellite-era reanalysis for the Indian monsoon region, J. Climate, 34, 5109–5133, https://doi.org/10.1175/JCLI-D-20-0412.1, 2021. a
Schwarzwald, K. and Geil, K.: xagg: A Python package to aggregate gridded data onto polygons, Journal of Open Source Software, 9, 7239, https://doi.org/10.21105/joss.07239, 2024.
Serinaldi, F. and Kilsby, C. G.: Rainfall extremes: Toward reconciliation after the battle of distributions, Water Resour. Res., 50, 336–352, https://doi.org/10.1002/2013WR014211, 2014. a, b, c
Serinaldi, F., Bárdossy, A., and Kilsby, C. G.: Upper tail dependence in rainfall extremes: would we know it if we saw it?, Stoch. Env. Res. Risk A., 29, 1211–1233, https://doi.org/10.1007/s00477-014-0946-8, 2015. a
Southworth, H., Heffernan, J. E., and Metcalfe, P. D.: texmex: Statistical modelling of extreme values, r package version 2.4.9, https://doi.org/10.32614/CRAN.package.texmex, 2024. a
Speight, L., Hall, J., and Kilsby, C.: A multi-scale framework for flood risk analysis at spatially distributed locations, J. Flood Risk Manag., 10, 124–137, https://doi.org/10.1111/jfr3.12175, 2017. a
Stengel, K., Glaws, A., Hettinger, D., and King, R. N.: Adversarial super-resolution of climatological wind and solar data, P. Natl. Acad. Sci. USA, 117, 16805–16815, https://doi.org/10.1073/pnas.1918964117, 2020. a
Steptoe, H. and Economou, T.: Extreme wind return periods from tropical cyclones in Bangladesh: insights from a high-resolution convection-permitting numerical model, Nat. Hazards Earth Syst. Sci., 21, 1313–1322, https://doi.org/10.5194/nhess-21-1313-2021, 2021. a
Thompson, V., Mitchell, D., Melia, N., Bloomfield, H., Dunstone, N., and Kay, G.: Detecting rising wildfire risks for South East England, Climate Resilience and Sustainability, 4, e70002, https://doi.org/10.1002/cli2.70002, 2025. a, b
UNDRR: The Sendai Framework Terminology on Disaster Risk Reduction: Hazard, https://www.undrr.org/terminology/hazard (last access: 9 February 2026), 2017. a
Wiese, M., Knobloch, R., and Korn, R.: Copula & marginal flows: Disentangling the marginal from its joint, arXiv [preprint], https://doi.org/10.48550/arXiv.1907.03361, 2019. a, b, c
Wilks, D. S. and Wilby, R. L.: The weather generation game: a review of stochastic weather models, Prog. Phys. Geog., 23, 329–357, https://doi.org/10.1177/030913339902300302, 1999. a
Wilson, T., Tan, P.-N., and Luo, L.: DeepGPD: A Deep Learning Approach for Modeling Geospatio-Temporal Extreme Events, Proceedings of the AAAI Conference on Artificial Intelligence, 36, 4245–4253, https://doi.org/10.1609/aaai.v36i4.20344, 2022. a
Yin, J., Gentine, P., Slater, L., Gu, L., Pokhrel, Y., Hanasaki, N., Guo, S., Xiong, L., and Schlenker, W.: Future socio-ecosystem productivity threatened by compound drought–heatwave events, Nature Sustainability, 6, 259–272, https://doi.org/10.1038/s41893-022-01024-1, 2023. a
Zhang, S., Solari, G., Yang, Q., and Repetto, M. P.: Extreme wind speed distribution in a mixed wind climate, J. Wind Eng. Ind. Aerod., 176, 239–253, https://doi.org/10.1016/j.jweia.2018.03.019, 2018. a
Zhao, S., Liu, Z., Lin, J., Zhu, J.-Y., and Han, S.: Differentiable augmentation for data-efficient GAN training, in: Advances in Neural Information Processing Systems, vol. 33, 7559–7570, arXiv [preprint], https://doi.org/10.48550/arXiv.2006.10738, 2020. a, b, c, d
Zscheischler, J., Westra, S., van den Hurk, B. J. J. M., Seneviratne, S. I., Ward, P. J., Pitman, A., AghaKouchak, A., Bresch, D. N., Leonard, M., Wahl, T., and Zhang, X.: Future climate risk from compound events, Nat. Clim. Change, 8, 469–477, https://doi.org/10.1038/s41558-018-0156-3, 2018. a, b, c
Zscheischler, J., Martius, O., Westra, S., Bevacqua, E., Raymond, C., Horton, R. M., van den Hurk, B., AghaKouchak, A., Jézéquel, A., Mahecha, M. D., Maraun, D., Ramos, A. M., Ridder, N. N., Thiery, W., and Vignotto, E.: A typology of compound weather and climate events, Nature Reviews Earth & Environment, 1, 333–347, https://doi.org/10.1038/s43017-020-0060-z, 2020. a, b, c
Short summary
We developed a generative deep learning method combining generative adversarial networks and extreme value theory to simulate spatially resolved multi-hazards events across large regions. Tested on storms in the Bay of Bengal, the model captured spatial patterns of wind, pressure, and rainfall during storm events, enabling more realistic disaster risk assessments than traditional methods. This flexible framework can be applied to various hazards and regions for improved disaster planning.
We developed a generative deep learning method combining generative adversarial networks and...
Altmetrics
Final-revised paper
Preprint