Examining fast radiative feedbacks using machine-learning-based emulators of weather

The response of the Earth system to radiative perturbations is governed by a combination of fast and slow feedbacks.  Slow feedbacks are typically activated in response to changes in ocean temperatures on decadal timescales and often manifest as changes in Earth-system state with no recent analogue.  On the other hand, fast feedbacks can be activated in response to rapid atmospheric physical processes on timescales of weeks and are already operative in the present-day weather system. This distinction implies that the physics of fast radiative feedbacks is present in the historical reanalyses that have served as the training data for many of the most successful recent machine-learning-based emulators of weather.  In addition, these feedbacks are functional under the historical boundary conditions pertaining to the top-of-atmosphere radiative balance and sea-surface temperatures.    We discuss our work using historically trained weather emulators to characterize and quantify fast radiative feedbacks without the need to retrain for prospective Earth system  conditions.