The reliability of emplacement temperature estimates using palaeomagnetic methods: A case study from Santorini, Greece

Abstract

The mode of origin of volcaniclastic deposits can be difficult to determine from field constraints, and the palaeomagnetic technique of emplacement temperature (T(e)) determination provides a powerful discriminatory test for primary volcanic origin. This technique requires that the low-blocking-temperature (T(b)) component of remanence in the direction of the Earth's field in inherited lithic clasts is of thermal origin and was acquired during transport and cooling in a hot pyroclastic flow; otherwise, the T(e) determination may be inaccurate. If the low-T(b) component is not a thermal origin it may be a viscous remanent magnetization (VRM) or a chemical remanent magnetization (CRM). The acquisition of a VRM depends on the duration of exposure to an applied magnetic field, and thus the laboratory unblocking temperature (T(ub)) of a VRM of a certain age imposes a minimum T(e) that can be determined for that deposit. Palaeointensity experiments were carried out to assess the magnetic origin (pTRM, CRM, or a combination of both) of the low-T(b) component in a number of samples from pyroclastic deposits from Santorini, Greece. Seven of the 24 samples used in these experiments passed the stringent tests for reliable palaointensity determination. These values demonstrated, for six of the samples, that the low-T(b) component was of thermal origin and therefore that the estimate of T(e) was valid. In the other 17 samples, valuable information was gained about the characteristics of the magnetic alteration that occurred during the palaeointensity experiments, allowing assessment of the reliability of T(e) estimates in these cases. These cases showed that if a CRM is present it has a direction parallel to the applied field, and not parallel to the direction of the parent grain. They also show that, even if a CRM is present, it does not necessarily affect the estimate of T(e). Two samples used in these experiments displayed curvature between their two components of magnetization. Data from this study indicate that the curvature in these vector plots is probably due to the presence of MD grains and not CRM, although CRM is not ruled out as the cause of curvature in other samples with two vector components.