ALBERT

Description: Methodological aspects in obtaining reliable absolute palaeointensity estimates have attracted renewed attention in recent years. Obtaining a reliable palaeointensity from lavas, however, still is notoriously difficult: in many cases lavas have been shown to be a non-ideal recorder of the palaeofield. Here we assess the influence of various thermomagnetic effects on the outcome of palaeointensity experiments on a suite of recent lavas from Mt Etna. Several palaeointensity studies have targeted recent flows of Mt Etna before; often palaeointensity underestimates were found. We apply both Thellier- and multispecimen-style palaeointensity experiments and obtain estimates generally 20–35 per cent below the expected reference value—in line with previous studies. We explored subtle chemical alteration, viscous magnetic behaviour and differences between the natural and laboratory cooling rates used during the palaeointensity experiments as possible causes for the observed underestimates of the palaeofield. For the latter, a furnace with controlled, programmable cooling rates was constructed. It can cool samples very slowly, with rates that are in the same order of magnitude as natural cooling rates typical of Mt Etna lava flows (~10 K/day). Both viscous magnetic behaviour and differences in cooling rate appear to influence the outcome of the palaeointensity experiments. However, the magnitude of the changes observed is not sufficient to explain the observed underestimates. We, therefore, hypothesize that the observed underestimates may be primarily explained by transdomain processes occurring during the heating steps of the palaeointensity experiments.