Predicting Effective Angular Momentum Function Forecast Errors

Time-variations in the orientation of the solid Earth are largely governed by the exchange of angular momentum with the surface geophysical fluids of atmosphere, oceans, and the land surface. Modelled fields of atmospheric winds, atmospheric surface pressure, ocean currents, ocean bottom pressure, and terrestrial water storage allow to calculate effective angular momentum (EAM) functions that provide highly reliable information about the orientation changes of the Earth. EAM forecasts derived from model forecast runs support substantially short-term Earth Orientation Parameters (EOP) Predictions. So far, routinely available EAM forecasts do not include any error information needed for rigorous combination of EAM forecasts with EOP predictions from various geodetic techniques. Based on hindcast experiments, we analysed the EAM forecast error and trained a neural network to predict EAM forecast errors. As expected, EAM forecast errors increase with the prediction horizon but we found also irregular large variation in the EAM forecast error that seem to be well predictable with machine learning methods.