Historical measurements of the Earth's magnetic field compared with remanence directions from lava flows in Italy over the last four centuries

Direct measurements of the Earth's magnetic field in Italy since 1640 a.d. have been used to check the remanence directions derived from historically dated volcanic rocks of Etna and Vesuvius. Direct measurements consist of the records of L'Aquila and Pola geomagnetic observatories, the repeat stations of the Italian Magnetic Network and the data base of the Historical Italian Geomagnetic Data Catalogue. All have been relocated to the same reference site (Viterbo - lat. 42.45°N, long. 12.03°E) in order to draw a reference secular variation (SV) curve. The direction of the Earth's field at Viterbo has also been calculated from the historical records (2000-1600) of ref. [Jackson, A., Jonkers, A.R.T., Walker, M.R., 2000. Four centuries of geomagnetic secular variation from historical records. Phil. Trans. R. Soc. London, Ser. A 358, 957-990] database. The remanence directions from Etna show a general agreement with the trend of the SV curve, although their inclination is usually lower than that from the direct measurement. The directions from Vesuvius are more scattered. Large discrepancies occur at both volcanoes and in some cases have been ascribed in the literature to poor geographic information, making it difficult to identify the flows actually emplaced during the eruptions reported in the chronicles. Closer examination shows that the great majority of the best-defined remanence directions (semi-angle of confidence α95 < 2.5°) deviate significantly from the geomagnetic direction measured at the time of the emplacement, the angle between the two directions being larger than the α95 value. The value of 2.5-3.0° can thus be regarded as a conservative evaluation of the error when dealing with dating Etna and Vesuvius lava flows older than 17th century, even when the accuracy attained in remanence measurements is higher. In default of a SV curve for Italy derived from archaeological artefacts, a further error in dating is introduced when reference is made to SV curves of other countries, even if well-established, as these are from regions too far from Italy (>600 km) to confidently relocate magnetic directions.