The effect of thermal core–mantle interactions on the palaeomagnetic secular variation

We compare a numerical dynamo model with a model of the palaeosecular variation and find that, although there is good agreement in the meridional distribution of the secular variation, the amplitude of secular variation in the dynamo model is too small. Increasing the Rayleigh number does little to increase the amplitude of the secular variation, while it increases the axial dipole component of the field to geophysically unreasonably large values. However, by introducing lateral variations in heat flux across the core–mantle boundary into the dynamo model, we are able to match the amplitude of the secular variation. In addition, we find that the dynamo model then exhibits persistently anomalous behaviour beneath the Pacific region.