The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2629
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T08:06:09.940287Z
status: published
lens: g_space
sector: climate
headline: Forecast-based attribution of extratropical cyclones using AI weather models
dek: Abstract Extratropical cyclones are among the most damaging weather systems in midlatitudes; however, quantifying the influence of anthropogenic climate change (ACC) on individual storms remains challenging due to the interplay between thermodynamic and dynamical processes. Here, we assess the capacity of artificial intelligence-based weather prediction (AIWP) models to both forecast and attribute two high-impact European extratropical cyclones with contrasting characteristics: storm Ciarán (November 2023), whic…
gain_title: In a study of two European storms, four AI weather models reproduced the large-scale evolution several days ahead and supported event-specific, computationally efficient attribution of human-driven climate change within the forecast window.
problem_title: (none)
trace_subject: (none)
gain_reading: In a study of two European storms, four AI weather models reproduced the large-scale evolution several days ahead and supported event-specific, computationally efficient attribution of human-driven climate change within the forecast window.
gain_evidence: All AIWP systems skillfully reproduce the large-scale evolution of both storms several days in advance | AIWP models can support event-specific, computationally efficient ACC attribution within the forecast window
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers tested four AI-based weather prediction models on two damaging European extratropical cyclones: storm Ciarán in November 2023, which had explosive cyclogenesis and extreme winds, and storm Claudia in November 2025, marked by an intense atmospheric river. They benchmarked the models against the ECMWF Integrated Forecast System.

All four AI models reproduced the large-scale evolution of both storms several days in advance, though performance varied by event. The team then compared forecasts started from ERA5 data with counterfactual runs using a pseudo-global warming perturbation derived from CMIP6 historical simulations.

The attribution showed coherent climate-change signals in sea-level pressure, low-level winds, moisture and precipitation. Moisture signals held across models for both storms, while wind and circulation responses depended more on the model. Precipitation results from the ECMWF AI forecasting system showed regional increases consistent with near-Clausius-Clapeyron scaling.
limitation: AI model performance was event-dependent, and wind and circulation attribution responses varied more between models than moisture signals did.
tag: Evidence-backed gain
key_points: Two storms were studied: Ciarán (November 2023, explosive cyclogenesis, extreme winds) and Claudia (November 2025, intense atmospheric river). | Four AI weather prediction models were benchmarked against ECMWF's operational Integrated Forecast System. | All AI systems reproduced the large-scale evolution of both storms several days ahead, with event-dependent performance. | Counterfactual simulations used a pseudo-global warming perturbation derived from CMIP6 historical simulations. | Moisture-related climate-change signals were robust across models for both storms. | Wind and circulation responses varied more between models, reflecting less robust dynamical responses. | ECMWF's AI system showed regional precipitation increases consistent with near-Clausius-Clapeyron scaling.
rundown: The study applies a forecast-based storyline attribution framework. Factual forecasts are initialized from ERA5 and compared with counterfactual simulations in which a warming perturbation derived from CMIP6 historical simulations is applied.

The two storms were chosen for contrasting characteristics: Ciarán was dominated by explosive deepening and extreme winds, while Claudia was defined by moisture transport in an atmospheric river.

Results differed by process. Thermodynamic moisture signals were consistent across the four models, while wind and circulation responses showed greater model dependence. Precipitation increases from the ECMWF AI system were modulated by event-specific dynamics affecting moisture availability and transport.
sources:
- peer_reviewed | Environmental Research: Climate | https://doi.org/10.1088/2752-5295/ae714b | 2026-05-21
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

e24afc9271281e5f375395d5441684462f0137192a537e534e837bc60725e4b8

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000