High-quality microwave archaeointensity determinations from an early 18th century ad English brick kiln

Abstract

SUMMARY Samples of bricks from the Dogmersfield Park post-medieval brick kiln (Hampshire, UK), have been used to determine the intensity of the geomagnetic field during the early 18th century ad. Extensive magnetic mineralogical analyses have been carried out revealing weathered zones and multidomain dominated areas within the bricks. Likewise, statistical data treatment has been used to identify mineralogical heterogeneities. Once the most suitable parts for the study were selected, archaeointensity determinations were carried out applying Coe's double-step variant of the Thellier method. The 14 GHz microwave system used in this study was developed at Liverpool University and the thermal demagnetizer is a commercial instrument. The former uses microwaves instead of heat to demagnetize the sample, the latter uses heat and was used for comparison. Data from the microwave system were of very high quality and uncertainties assigned to single measurements are smaller than the intrinsic variability of the archaeointensity stored within the bricks. Both pTRM and pTRM tail checks rarely fail, and even pass when the sample is almost fully demagnetized. Microwave intensities group around a central value (∼53 μT), slightly higher than the present-day field intensity, and agree well with previous studies and geomagnetic field models. Thermal results exhibit higher and more scattered values; some samples show evidence of laboratory thermal alteration. The different behaviour of samples when subjected to microwave and thermal experiments appears to be linked with the magnetic mineralogy (high-Ti titanomagnetite): easy microwave demagnetization without alteration is achieved whereas sample alteration is almost unavoidable during thermal demagnetization. High-Ti titanomagnetite seems to be the ideal mineralogy for microwave experiments instead of pure magnetite, preferred in conventional Thellier experiments. The last part of this paper discusses the usefulness of conventional alteration tests applied to microwave experiments. It appears that for this technique negative alteration checks can originate from poor reproducibility of the absorbed microwave power.