{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,31]],"date-time":"2026-10-31T23:14:24Z","timestamp":1793488464067,"version":"4.5.0"},"reference-count":81,"publisher":"National Academy of Sciences","issue":"17","license":[{"start":{"date-parts":[[2025,4,25]],"date-time":"2025-04-25T00:00:00Z","timestamp":1745539200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/"}],"content-domain":{"domain":["www.pnas.org"],"crossmark-restriction":true},"short-container-title":["Proc. Natl. Acad. Sci. U.S.A."],"published-print":{"date-parts":[[2025,4,29]]},"abstract":"<jats:p>Regional high-resolution climate projections are crucial for many applications, such as agriculture, hydrology, and natural hazard risk assessment. Dynamical downscaling, the state-of-the-art method to produce localized future climate information, involves running a regional climate model (RCM) driven by an Earth System Model (ESM), but it is too computationally expensive to apply to large climate projection ensembles. We propose an approach combining dynamical downscaling with generative AI to reduce the cost and improve the uncertainty estimates of downscaled climate projections. In our framework, an RCM dynamically downscales ESM output to an intermediate resolution, followed by a generative diffusion model that further refines the resolution to the target scale. This approach leverages the generalizability of physics-based models and the sampling efficiency of diffusion models, enabling the downscaling of large multimodel ensembles. We evaluate our method against dynamically downscaled climate projections from the Coupled Model Intercomparison Project 6 (CMIP6) ensemble. Our results demonstrate its ability to provide more accurate uncertainty bounds on future regional climate than alternatives such as dynamical downscaling of smaller ensembles, or traditional empirical statistical downscaling methods. We also show that dynamical-generative downscaling results in significantly lower errors than popular statistical downscaling techniques, and captures more accurately the spectra, tail dependence, and multivariate correlations of meteorological fields. These characteristics make the dynamical-generative framework a flexible, accurate, and efficient way to downscale large ensembles of climate projections, currently out of reach for pure dynamical downscaling.<\/jats:p>","DOI":"10.1073\/pnas.2420288122","type":"journal-article","created":{"date-parts":[[2025,4,25]],"date-time":"2025-04-25T13:44:25Z","timestamp":1745588665000},"update-policy":"https:\/\/doi.org\/10.1073\/pnas.cm10313","source":"Crossref","is-referenced-by-count":63,"title":["Dynamical-generative downscaling of climate model ensembles"],"prefix":"10.1073","volume":"122","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7255-5895","authenticated-orcid":false,"given":"Ignacio","family":"Lopez-Gomez","sequence":"first","affiliation":[{"name":"Google Research","place":["Mountain View"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zhong Yi","family":"Wan","sequence":"additional","affiliation":[{"name":"Google Research","place":["Mountain View"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Leonardo","family":"Zepeda-N\u00fa\u00f1ez","sequence":"additional","affiliation":[{"name":"Google Research","place":["Mountain View"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-5687-2287","authenticated-orcid":false,"given":"Tapio","family":"Schneider","sequence":"additional","affiliation":[{"name":"Google Research","place":["Mountain View"]},{"name":"Department of Environmental Science and Engineering","place":["Pasadena"]},{"name":"California Institute of Technology","place":["Pasadena"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"John","family":"Anderson","sequence":"additional","affiliation":[{"name":"Google Research","place":["Mountain View"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0006-1733-5482","authenticated-orcid":false,"given":"Fei","family":"Sha","sequence":"additional","affiliation":[{"name":"Google Research","place":["Mountain View"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"341","published-online":{"date-parts":[[2025,4,25]]},"reference":[{"key":"e_1_3_4_1_2","doi-asserted-by":"publisher","DOI":"10.1038\/nclimate2563"},{"key":"e_1_3_4_2_2","doi-asserted-by":"publisher","DOI":"10.1038\/nature05095"},{"key":"e_1_3_4_3_2","doi-asserted-by":"publisher","DOI":"10.1038\/nclimate2258"},{"key":"e_1_3_4_4_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0506042102"},{"key":"e_1_3_4_5_2","doi-asserted-by":"publisher","DOI":"10.1175\/BAMS-D-19-0113.1"},{"key":"e_1_3_4_6_2","doi-asserted-by":"publisher","DOI":"10.1175\/MWR2996.1"},{"key":"e_1_3_4_7_2","doi-asserted-by":"publisher","DOI":"10.1007\/s00382-009-0650-4"},{"key":"e_1_3_4_8_2","doi-asserted-by":"publisher","DOI":"10.1002\/qj.2484"},{"key":"e_1_3_4_9_2","doi-asserted-by":"publisher","DOI":"10.1038\/s43016-020-00181-w"},{"key":"e_1_3_4_10_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.jhydrol.2012.05.052"},{"key":"e_1_3_4_11_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-021-21785-1"},{"key":"e_1_3_4_12_2","doi-asserted-by":"publisher","DOI":"10.1175\/JCLI-D-12-00539.1"},{"key":"e_1_3_4_13_2","doi-asserted-by":"publisher","DOI":"10.1088\/1748-9326\/ab83a7"},{"key":"e_1_3_4_14_2","doi-asserted-by":"publisher","DOI":"10.1029\/2018JD030094"},{"key":"e_1_3_4_15_2","doi-asserted-by":"publisher","DOI":"10.1126\/science.aaa0629"},{"key":"e_1_3_4_16_2","doi-asserted-by":"publisher","DOI":"10.1007\/s00382-013-2011-6"},{"key":"e_1_3_4_17_2","doi-asserted-by":"publisher","DOI":"10.1029\/2023GL105979"},{"key":"e_1_3_4_18_2","unstructured":"President\u2019s Council of Advisors on Science and Technology (PCAST) \u201cExtreme weather risk in a changing climate: Enhancing prediction and protecting communities\u201d (Report to the president Washington D.C. 2023)."},{"key":"e_1_3_4_19_2","unstructured":"F. Doblas-Reyes \u201cLinking global to regional climate change\u201d 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 V. Masson-Delmotte Ed. (Cambridge University Press 2021) pp. 1363\u20131512."},{"key":"e_1_3_4_20_2","doi-asserted-by":"publisher","DOI":"10.1175\/1520-0493(1998)126<1306:TGHPSA>2.0.CO;2"},{"key":"e_1_3_4_21_2","doi-asserted-by":"publisher","DOI":"10.1029\/2005GL024954"},{"key":"e_1_3_4_22_2","doi-asserted-by":"publisher","DOI":"10.1038\/nclimate1344"},{"key":"e_1_3_4_23_2","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.aba1323"},{"key":"e_1_3_4_24_2","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.0900094106"},{"key":"e_1_3_4_25_2","doi-asserted-by":"publisher","DOI":"10.1175\/BAMS-D-23-0100.1"},{"key":"e_1_3_4_26_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-023-38906-7"},{"key":"e_1_3_4_27_2","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.aaw1976"},{"key":"e_1_3_4_28_2","doi-asserted-by":"publisher","DOI":"10.1175\/JAMC-D-20-0057.1"},{"key":"e_1_3_4_29_2","doi-asserted-by":"publisher","DOI":"10.1029\/2023EF003531"},{"key":"e_1_3_4_30_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-17-229-2024"},{"key":"e_1_3_4_31_2","doi-asserted-by":"publisher","DOI":"10.1007\/s00382-022-06343-9"},{"key":"e_1_3_4_32_2","doi-asserted-by":"publisher","DOI":"10.1029\/2022EF003291"},{"key":"e_1_3_4_33_2","doi-asserted-by":"publisher","DOI":"10.1029\/2022MS003593"},{"key":"e_1_3_4_34_2","doi-asserted-by":"publisher","DOI":"10.1029\/2024GL111828"},{"key":"e_1_3_4_35_2","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.adk4489"},{"key":"e_1_3_4_36_2","doi-asserted-by":"crossref","unstructured":"M. Mardani Residual corrective diffusion modeling for km-scale atmospheric downscaling. Commun Earth Environ 6 124 (2025).","DOI":"10.1038\/s43247-025-02042-5"},{"key":"e_1_3_4_37_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-024-08252-9"},{"key":"e_1_3_4_38_2","first-page":"47749","article-title":"Debias coarsely, sample conditionally: Statistical downscaling through optimal transport and probabilistic diffusion models","volume":"36","author":"Wan Z. Y.","year":"2023","unstructured":"Z. Y. Wan , Debias coarsely, sample conditionally: Statistical downscaling through optimal transport and probabilistic diffusion models. Adv. Neural Inf. Process. Syst. 36, 47749\u201347763 (2023).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"e_1_3_4_39_2","doi-asserted-by":"publisher","DOI":"10.1029\/2022MS003120"},{"key":"e_1_3_4_40_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41612-024-00679-1"},{"key":"e_1_3_4_41_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-17-2265-2024"},{"key":"e_1_3_4_42_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-9-3461-2016"},{"key":"e_1_3_4_43_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.gloenvcha.2016.05.009"},{"key":"e_1_3_4_44_2","unstructured":"W. Skamarock \u201cA description of the advanced research WRF model version 4\u201d (Tech. Rep. NCAR\/TN-556+STR NCAR 2021)."},{"key":"e_1_3_4_45_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-12-4823-2019"},{"key":"e_1_3_4_46_2","doi-asserted-by":"crossref","unstructured":"A. W. Wood E. P. Maurer A. Kumar D. P. Lettenmaier Long-range experimental hydrologic forecasting for the eastern United States. J. Geophys. Res. Atmos. 107 ACL 6\u20131\u2013ACL 6\u201315 (2002).","DOI":"10.1029\/2001JD000659"},{"key":"e_1_3_4_47_2","doi-asserted-by":"publisher","DOI":"10.1023\/B:CLIM.0000013685.99609.9e"},{"key":"e_1_3_4_48_2","doi-asserted-by":"publisher","DOI":"10.5194\/hess-16-3309-2012"},{"key":"e_1_3_4_49_2","doi-asserted-by":"publisher","DOI":"10.1029\/2023EF004107"},{"key":"e_1_3_4_50_2","doi-asserted-by":"crossref","unstructured":"S. Basile A. R. Crimmins C. W. Avery B. D. Hamlington K. E. Kunkel \u201cAppendix 3. Scenarios and datasets\u201d in Fifth National Climate Assessment A. Crimmins Eds. (U.S. Global Change Research Program 2023).","DOI":"10.7930\/NCA5.2023.A3"},{"key":"e_1_3_4_51_2","unstructured":"B. Barthel Sorensen A probabilistic framework for learning non-intrusive corrections to long-time climate simulations from short-time training data. arXiv [Preprint] (2024). https:\/\/doi.org\/10.48550\/arXiv.2408.02688 (Accessed 1 September 2024)."},{"key":"e_1_3_4_52_2","doi-asserted-by":"publisher","DOI":"10.1175\/1520-0434(2000)015<0559:DOTCRP>2.0.CO;2"},{"key":"e_1_3_4_53_2","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9469.2005.00483.x"},{"key":"e_1_3_4_54_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41558-021-01076-9"},{"key":"e_1_3_4_55_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.jhydrol.2005.01.025"},{"key":"e_1_3_4_56_2","doi-asserted-by":"publisher","DOI":"10.1029\/2009RG000314"},{"key":"e_1_3_4_57_2","doi-asserted-by":"publisher","DOI":"10.1175\/BAMS-89-3-347"},{"key":"e_1_3_4_58_2","doi-asserted-by":"publisher","DOI":"10.1038\/nclimate2657"},{"key":"e_1_3_4_59_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-023-40112-4"},{"key":"e_1_3_4_60_2","doi-asserted-by":"publisher","DOI":"10.1029\/2024GL108160"},{"key":"e_1_3_4_61_2","doi-asserted-by":"publisher","DOI":"10.1002\/joc.2257"},{"key":"e_1_3_4_62_2","doi-asserted-by":"publisher","DOI":"10.1002\/2015GL067081"},{"key":"e_1_3_4_63_2","doi-asserted-by":"publisher","DOI":"10.1175\/WAF-D-15-0141.1"},{"key":"e_1_3_4_64_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-15-2973-2022"},{"key":"e_1_3_4_65_2","doi-asserted-by":"publisher","DOI":"10.1029\/2018GL080261"},{"key":"e_1_3_4_66_2","doi-asserted-by":"publisher","DOI":"10.1029\/2020EF001783"},{"key":"e_1_3_4_67_2","doi-asserted-by":"publisher","DOI":"10.1038\/s41558-018-0156-3"},{"key":"e_1_3_4_68_2","doi-asserted-by":"publisher","DOI":"10.1088\/1748-9326\/ad6d82"},{"key":"e_1_3_4_69_2","doi-asserted-by":"publisher","DOI":"10.1029\/2019MS001739"},{"key":"e_1_3_4_70_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-12-2727-2019"},{"key":"e_1_3_4_71_2","doi-asserted-by":"publisher","DOI":"10.1071\/ES19040"},{"key":"e_1_3_4_72_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-12-3241-2019"},{"key":"e_1_3_4_73_2","doi-asserted-by":"publisher","DOI":"10.5194\/gmd-13-6165-2020"},{"key":"e_1_3_4_74_2","doi-asserted-by":"publisher","DOI":"10.1029\/2020MS002353"},{"key":"e_1_3_4_75_2","doi-asserted-by":"publisher","DOI":"10.1002\/qj.3803"},{"key":"e_1_3_4_76_2","first-page":"6840","article-title":"Denoising diffusion probabilistic models","volume":"33","author":"Ho J.","year":"2020","unstructured":"J. Ho, A. Jain, P. Abbeel, Denoising diffusion probabilistic models. Adv. Neural Inf. Process. Syst. 33, 6840\u20136851 (2020).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"e_1_3_4_77_2","doi-asserted-by":"publisher","DOI":"10.1162\/NECO_a_00142"},{"key":"e_1_3_4_78_2","first-page":"26565","article-title":"Elucidating the design space of diffusion-based generative models","volume":"35","author":"Karras T.","year":"2022","unstructured":"T. Karras, M. Aittala, T. Aila, S. Laine, Elucidating the design space of diffusion-based generative models. Adv. Neural Inf. Process. Syst. 35, 26565\u201326577 (2022).","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"e_1_3_4_79_2","unstructured":"J. Ho T. Salimans Classifier-free diffusion guidance. arXiv [Preprint] (2022). https:\/\/doi.org\/10.48550\/arXiv.2207.12598 (Accessed 7 August 2024)."},{"key":"e_1_3_4_80_2","unstructured":"I. Lopez-Gomez Z. Y. Wan Supporting software for \u201cDynamical-generative downscaling of climate model ensembles\u201d. GitHub. https:\/\/github.com\/google-research\/swirl-dynamics\/tree\/main\/swirl_dynamics\/projects\/probabilistic_diffusion\/downscaling\/gcm_wrf. Deposited 20 February 2025."},{"key":"e_1_3_4_81_2","unstructured":"I. Lopez-Gomez Supporting data for \u201cDynamical-generative downscaling of climate model ensembles\u201d. Google Cloud. https:\/\/console.cloud.google.com\/storage\/browser\/dynamical_generative_downscaling. Deposited 14 January 2025."}],"container-title":["Proceedings of the National Academy of Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/pnas.org\/doi\/pdf\/10.1073\/pnas.2420288122","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,25]],"date-time":"2025-04-25T13:45:11Z","timestamp":1745588711000},"score":1,"resource":{"primary":{"URL":"https:\/\/pnas.org\/doi\/10.1073\/pnas.2420288122"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,25]]},"references-count":81,"journal-issue":{"issue":"17","published-print":{"date-parts":[[2025,4,29]]}},"alternative-id":["10.1073\/pnas.2420288122"],"URL":"https:\/\/doi.org\/10.1073\/pnas.2420288122","relation":{},"ISSN":["0027-8424","1091-6490"],"issn-type":[{"value":"0027-8424","type":"print"},{"value":"1091-6490","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,4,25]]},"assertion":[{"value":"2024-10-02","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-03-19","order":2,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2025-04-25","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e2420288122"}}