MétaCan
Menu
Back to cohort
Record W4408427947 · doi:10.5194/egusphere-egu25-12640

Weather Generator Based on Generative AI for Interdisciplinary Probabilistic Downscaling Using Convection-Permitting Model Outputs and Potential Utility in Equitable, Community-focused Climate Scenario-ing

2025· preprint· en· W4408427947 on OpenAlex

Why this work is in the frame

A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.

aboutThe title or abstract carries a Canadian signal from the geographic lexicon.
no affNo Canadian affiliation: this work is invisible to an affiliation-only frame.
No Canadian affiliation. An affiliation-only frame, the usual design, would never have seen this work. It is one of the works that make the case for inverting the frame.

Bibliographic record

Venuenot available
Typepreprint
Languageen
FieldEnvironmental Science
TopicHydrological Forecasting Using AI
Canadian institutionsnot available
Fundersnot available
KeywordsDownscalingProbabilistic logicClimate modelGenerator (circuit theory)Generative grammarClimatologyComputer scienceMeteorologyEnvironmental scienceClimate changeArtificial intelligenceGeographyGeologyOceanography

Abstract

fetched live from OpenAlex

Convection-permitting model outputs offer significant opportunities for training statistical downscaling approaches. The Coordinated Regional Climate Downscaling Experiment (CORDEX) on the urban environment and regional climate change ensemble simulations provide valuable insights into the uncertainties of numerical atmospheric models. Traditional weather generators, based on the Maximum Likelihood for the Generalised Linear Model approach, have been instrumental in modelling precipitation occurrence and amount. This study advances the statistical downscaling method by integrating Generative AI approaches, using deep learning to create stochastic precipitation ensembles.Compared to deterministic simulations, this new probabilistic approach allows for an exploration of the nonstationary statistical properties influenced by regional climate conditions through more feasible nonlinear representation for the weather generator parameters by deep learning. Emphasis is placed on the importance of probabilistic and agnostic methods in exploring, interpreting, and explaining uncertainties.Findings related to temperature variations for daily precipitation extremes attribute the roles of sensible and latent heat, which are further interpreted through regional processes. The integration of generative AI highlights the stochastic uncertainties in weather generators, emphasising the need for consistency between deterministic convection-permitting model outputs and observational data. By examining scaling relationships, the interpretability and explainability of model outputs, particularly concerning energy balance processes, are demonstrated.Through interpretable and explainable statistical downscaling, the approach to modelling precipitation extremes based on maximum likelihood theory fosters international collaboration in the Climate Collaboratorium* project (IIRCC; ‘Exploring climate solutions with interactive theatre’)This includes contributions from Canada, Germany, the UK, and the US, aimed at providing accessible science that can inform climate decisions in partnership with social science/arts and humanities researchers, tailored to place-based user needs. Advocacy for responsible AI in atmospheric and water sciences facilitates interdisciplinary climate adaptation and mitigation with Taiwanese and Brazilian communities. This approach promotes transparency and fairness through explainable and interpretable climate scenarios. By incorporating immersive experiences and smart decision-making processes, the pathway for human oversight remains central to fair climate action to achieve Sustainable Development Goal 13.*https://www.ukri.org/publications/international-science-partnerships-fund-iircc-initiative-funded-projects/international-joint-initiative-for-research-in-climate-change-adaptation-and-mitigation-project-overview/

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.003
metaresearch head score (Gemma)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesMeta-epidemiology (narrow), Science and technology studies
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.212
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0030.000
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.000
Science and technology studies0.0010.000
Scholarly communication0.0000.000
Open science0.0000.005
Research integrity0.0010.002
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.057
GPT teacher head0.320
Teacher spread0.263 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same topicHydrological Forecasting Using AIFrench-language works237,207