Synchronous Coupling of Large Climate Models for Improved Climate Change Projection

Models used internationally to project climate change give divergent predictions in regard to the magnitude of globally averaged warming and in regard to specific regional changes. An ``interactive ensemble' of three different global climate models (GCMs) has been formed using inter-model data assimilation. Even with crude assignment of assimilation weights based on historical data, the resulting 'supermodel' promises improved performance as compared to any single model or weighted average of model outputs.