New archaeomagnetic data recovered from the study of celtiberic remains from central Spain (Numantia and Ciadueña, 3rd-1st centuries BC). Implications on the fidelity of the Iberian paleointensity database

Abstract : Variations of geomagnetic field in the Iberian Peninsula prior to roman times are poorly constrained. Here we report new archaeomagnetic results from four ceramic collections and two combustion structures recovered in two pre-roman (celtiberic) archaeological sites in central Spain. The studied materials have been dated by archaeological evidences and supported by five radiocarbon dates. Rock magnetic experiments indicate that the characteristic remanent manetization (ChRM) is carried by a low coercivity magnetic phase with Curie temperatures of 530–575 °C, most likely Ti-poor titanomagnetite/titanomaghemite. Archaeointensity determinations were carried out by using the classical Thellier-Thellier protocol including tests and corrections for magnetic anisotropy and cooling rate dependency. Two magnetic behaviours were depicted during the laboratory treatment. Black potsherds and poor heated samples from the kilns, presented two magnetization components, alterations or curved Arai plots and were therefore rejected. In contrast, well heated specimens (red ceramic fragments and well heated samples from the kilns) show one single well defined component of magnetization going through the origin and linear Arai plots providing successful archaeointensity determinations. The effect of anisotropy of the thermoremanent magnetization (ATRM) on paleointensity analysis was systematically investigated obtaining very high ATRM corrections on fine pottery specimens. In some cases, differences between the uncorrected and ATRM corrected paleointensity values reached up to 86 %. The mean intensity values obtained from three selected set of samples were 64.3 ± 5.8 μT; 56.8 ± 3.8 and 56.7 ± 4.6 μT (NUS2, CI2 and CIA, respectively), which contribute to better understand the evolution of the palaeofield intensity in central Iberia during the 3rd-1st centuries BC. The direction of the field at first century BC has also been determined from oriented samples from CIA kilns (D = 357.2°; I = 62.2°; N = 10, α95 = 2.7°). The new archaeointensity data disagrees with previous results from Iberian ceramics which were not corrected for the ATRM effect. On the contrary, they are in agreement with the most recent French paleointensity curve and the latest European intensity model; both based on a selection of high quality paleointensity data. This result reinforces the idea that the puzzling scatter often observed in the global paleointensity database is likely due to differences in the laboratory protocols. Further data from well-established laboratory protocols are still necessary to delineate confidently the evolution of the geomagnetic palaeofield during the first millennium BC.