Timezone: »

Improving the predictions of ML-corrected climate models with novelty detection
Clayton Sanford · Anna Kwa · Oliver Watt-Meyer · Spencer K. Clark · Noah Brenowitz · Jeremy McGibbon · Christopher S. Bretherton
Event URL: https://www.climatechange.ai/papers/neurips2022/10 »

While previous works have shown that machine learning (ML) can improve the prediction accuracy of coarse-grid climate models, these ML-augmented methods are more vulnerable to irregular inputs than the traditional physics-based models they rely on. Because ML-predicted corrections feed back into the climate model’s base physics, the ML-corrected model regularly produces out of sample data, which can cause model instability and frequent crashes. This work shows that adding semi-supervised novelty detection to identify out-of-sample data and disable the ML-correction accordingly stabilizes simulations and sharply improves the quality of predictions. We design an augmented climate model with a one-class support vector machine (OCSVM) novelty detector that provides better temperature and precipitation forecasts in a year-long simulation than either a baseline (no-ML) or a standard ML-corrected run. By improving the accuracy of coarse-grid climate models, this work helps make accurate climate models accessible to researchers without massive computational resources.

Author Information

Clayton Sanford (Columbia University)
Anna Kwa (Allen Institute for Artificial Intelligence)
Oliver Watt-Meyer (Allen Institute for Artificial Intelligence)
Spencer K. Clark (Allen Institute for Artificial Intelligence)
Noah Brenowitz (Allen Institute for AI)
Jeremy McGibbon (Allen Institute for Artificial Intelligence)
Christopher S. Bretherton (Allen Institute for Artificial Intelligence)

More from the Same Authors