A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation

Abstract Local climate information is crucial for impact assessment and decision-making, yet coarse global climate simulations cannot capture small-scale phenomena. Current statistical downscaling methods infer these phenomena as temporally decoupled spatial patches. However, to preserve physical pr...

Full description

Saved in:
Bibliographic Details
Main Authors: Jonathan Schmidt, Luca Schmidt, Felix M. Strnad, Nicole Ludwig, Philipp Hennig
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:npj Climate and Atmospheric Science
Online Access:https://doi.org/10.1038/s41612-025-01157-y
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849387862005383168
author Jonathan Schmidt
Luca Schmidt
Felix M. Strnad
Nicole Ludwig
Philipp Hennig
author_facet Jonathan Schmidt
Luca Schmidt
Felix M. Strnad
Nicole Ludwig
Philipp Hennig
author_sort Jonathan Schmidt
collection DOAJ
description Abstract Local climate information is crucial for impact assessment and decision-making, yet coarse global climate simulations cannot capture small-scale phenomena. Current statistical downscaling methods infer these phenomena as temporally decoupled spatial patches. However, to preserve physical properties, estimating spatio-temporally coherent high-resolution weather dynamics for multiple variables across long time horizons is crucial. We present a novel generative framework that uses a score-based diffusion model trained on high-resolution reanalysis data to capture the statistical properties of local weather dynamics. After training, we condition on coarse climate model data to generate weather patterns consistent with the aggregate information. As this predictive task is inherently uncertain, we leverage the probabilistic nature of diffusion models and sample multiple trajectories. We evaluate our approach with high-resolution reanalysis information before applying it to the climate model downscaling task. We then demonstrate that the model generates spatially and temporally coherent weather dynamics that align with global climate output.
format Article
id doaj-art-a145346239dc4c41b48cd037107b7ccf
institution Kabale University
issn 2397-3722
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series npj Climate and Atmospheric Science
spelling doaj-art-a145346239dc4c41b48cd037107b7ccf2025-08-20T03:42:27ZengNature Portfolionpj Climate and Atmospheric Science2397-37222025-07-018111210.1038/s41612-025-01157-yA Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate SimulationJonathan Schmidt0Luca Schmidt1Felix M. Strnad2Nicole Ludwig3Philipp Hennig4University of TübingenUniversity of TübingenUniversity of TübingenUniversity of TübingenUniversity of TübingenAbstract Local climate information is crucial for impact assessment and decision-making, yet coarse global climate simulations cannot capture small-scale phenomena. Current statistical downscaling methods infer these phenomena as temporally decoupled spatial patches. However, to preserve physical properties, estimating spatio-temporally coherent high-resolution weather dynamics for multiple variables across long time horizons is crucial. We present a novel generative framework that uses a score-based diffusion model trained on high-resolution reanalysis data to capture the statistical properties of local weather dynamics. After training, we condition on coarse climate model data to generate weather patterns consistent with the aggregate information. As this predictive task is inherently uncertain, we leverage the probabilistic nature of diffusion models and sample multiple trajectories. We evaluate our approach with high-resolution reanalysis information before applying it to the climate model downscaling task. We then demonstrate that the model generates spatially and temporally coherent weather dynamics that align with global climate output.https://doi.org/10.1038/s41612-025-01157-y
spellingShingle Jonathan Schmidt
Luca Schmidt
Felix M. Strnad
Nicole Ludwig
Philipp Hennig
A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
npj Climate and Atmospheric Science
title A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
title_full A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
title_fullStr A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
title_full_unstemmed A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
title_short A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulation
title_sort generative framework for probabilistic spatiotemporally coherent downscaling of climate simulation
url https://doi.org/10.1038/s41612-025-01157-y
work_keys_str_mv AT jonathanschmidt agenerativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT lucaschmidt agenerativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT felixmstrnad agenerativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT nicoleludwig agenerativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT philipphennig agenerativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT jonathanschmidt generativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT lucaschmidt generativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT felixmstrnad generativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT nicoleludwig generativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation
AT philipphennig generativeframeworkforprobabilisticspatiotemporallycoherentdownscalingofclimatesimulation