The axial dipole strength and flow in the outer core

From 1590 to 1833, although extensive measurements of the geomagnetic field were made, no direct measurements of its intensity are available; field models for this period either make use of indirect (palaeomagnetic) intensity determinations or assume a similar change in dipole strength to that seen presently. Here, we examine the impact of different choices of field model strength on models of the core-surface flow, derived from the geomagnetic secular variation. We find that for a reasonable range of possible changes in the field strength, the variability that this unknown imposes on the flow structure is small compared with other choices in the modelling strategy, such as the degree to which the geomagnetic models are fit, the smoothness of the flows, and a priori information included to reduce non-uniqueness in the flow calculation. The flows predict variations in core angular momentum of similar magnitude to those seen in recent times. These variations are larger than seen in models from geodetic data, but do not seem inconsistent with the geodetic data themselves. We therefore suggest that it is reasonable to study core flows back at least until 1650, thereby doubling the time interval for which flows can be considered to constrain the behaviour of the geodynamo.