Model Training, Data Assimilation, and Forecast Experiments with a Hybrid Atmospheric Model
This paper investigates the performance of a unique proof-of-concept hybrid model in a cycling data assimilation scheme. This previously published model combines the Simplified Parameterization, primitive-Equation Dynamics model (SPEEDY) with an ML-based component that itself is capable of modeling the global atmosph...