Evidence for rapid geomagnetic field intensity variations in Western Europe over the past 800 years from new French archeointensity data

The number of reliable archeointensity determinations obtained from Western Europe for the past millennium remains limited. Moreover, the large scatter between different datasets available is puzzling. The present study analyzed 31 new groups of baked clay (ceramic or brick) fragments sampled in France (29 groups) and in Belgium (2 groups). These groups contain several fragments collected from different artefacts and are precisely dated principally from historical constraints between the XIIIth and the XIXth centuries. Additionally, we re-evaluated 14 intensity values that we previously obtained from the same time period. The fragments were analyzed using two different thermal methods: (1) the in field-zero field (IZ) or the IZZI version of the classical Thellier and Thellier method and (2) the Triaxe protocol that involves high-temperature magnetization measurements. Data were corrected for the anisotropy of thermoremanent magnetization (TRM) and the dependence of TRM acquisition on the cooling rate was taken into account in the different protocols. Archeointensity data obtained on twin specimens sampled from the same fragment and using both experimental techniques generally show a good agreement (i.e. within 5%) at the fragment and at the site level. All retained site-level averaged intensity results (43 of 45 groups) have standard deviations of less than 5 PT. Furthermore, groups of approximately the same age have very consistent archeointensity. Altogether, the data presented herein recover a detailed and smoothed geomagnetic field intensity variation curve characterized by two peaks in intensity, the first during the second half of the XIVth century and the second around AD 1600, followed by a significant decreasing trend in intensity during most the XVIIth and XVIIIth centuries. This evolution does not satisfactorily fit with the expected intensity values for France derived from geomagnetic field models relying on a different evolution of the axial dipole moment. Our results lead us to propose that the axial dipole moment decreased from AD 1600 to the end of the XVIIIth century, then slightly increased up to similar to AD 1850 before decreasing again to present day. (C) 2009 Elsevier B.V. All rights reserved.