Statistical forecasting techniques applied to observatory data for core field modelling

Modelling of the geomagnetic field is a complex challenge, hindered by noisy and incomplete ground and satellite observations, and the extent to which we can separate the contributions of the various field sources in these data. Forecasting of the core field and its time variations (secular variation), an activity of key interest for academic and applied studies of geomagnetism and space weather, is further complicated by an incomplete knowledge of the physics controlling magnetic field generation. Predictive core field models often rely on simple mathematical extrapolation to produce short term (