Caltech team uses AI to forecast tipping points in simulations
Researchers demonstrated a recurrent neural operator that predicts abrupt transitions in climate-system and aircraft-wing airflow simulations.
Event / data period: 2026-10-15
Published: · Updated:
Researchers led by Caltech reported on 15 October that a deep-learning model called a recurrent neural operator could forecast tipping points in simulations of simple systems. Tests covered a climate-change system and turbulent airflow over an aircraft wing; the study appeared in PNAS.
The team proposes using the model’s detection that learned physics no longer describes the dynamics as a warning of an approaching abrupt transition. The results are from simulations and do not establish a forecasting system validated in the real climate or in flight. The research involved scientists from Caltech, the University of Colorado Boulder and NVIDIA.
Data
Research record
Research announcement published 15 October
- · Occurred
Neural-operator simulations
Forecasting transitions in climate-system and aircraft-wing airflow simulations.
Methodology
Findings are simulation results reported in the PNAS paper and Caltech account; they do not show operational forecasting performance in the real world.
Updated:
Subjects and places
Subject proposals
- Artificial intelligenceSuggested
- Subject suggested from the title or summary: AIwww.caltech.edu ↗
Geographic references
Only source-provided geographic references are shown. Opening an external map is optional. Optional story maps use OpenStreetMap. Opening a map sends your IP address and this site’s address to the map provider; location data remains readable as text without loading the map.
- United States · Place role: Coverage · Place type: Country or area · Location precision: RegionalEvidence
- Explicit preserved editorial country/region; no city or national centroid inferredwww.caltech.edu ↗