Reproducibility of archaeointensity determinations with a multimethod approach on archaeological material reproductions

Abstract

Archaeointensity determinations on burnt archaeological material are complex and reliable data scarce, although this kind of material can be of great interest in archaeological investigations. With the goal of analysing the reliability of archaeointensity determinations, an interlaboratory comparison study has been performed combining different experimental protocols on present-day reproductions of Pre-Columbian Mesoamerican archaeological artefacts and two brick samples. Samples were baked in an original kiln from an artisan workshop in western Mexico. The ambient magnetic field at the site during the experiment was measured and continuous temperature data were recorded at four different positions in the kiln during the heating–cooling procedure.

Archaeointensity determinations were carried out with four different methods at four different palaeomagnetic laboratories: Thellier–Coe (Burgos, Spain), microwave (Liverpool, UK), multispecimen (Morelia, Mexico) and multispecimen with the extended protocols for fraction and domain-state correction (Montpellier, France). 26 conventional resistive heating determinations with the Thellier–Coe protocol yielded a 100 per cent success rate, while 7 out of 8 microwave-heating determinations with the Thellier–Coe protocol also provided successful results. Also, two multispecimen determinations performed with both multispecimen methods provided statistically reliable results. In all cases, a good agreement between the determined archaeointensities and the ambient field at the production site could be observed.

Highly reversible magnetization-versus-temperature curves yielded slightly Al, Mg or Ti-substituted magnetite as the main ferromagnetic (s.l.) phase. In addition, in several samples, a thermally stable low Curie-temperature phase displaying a high coercivity behaviour could be observed in thermomagnetic curves and by thermal demagnetization of saturation isothermal remanent magnetization. This phase is interpreted as ε-Fe2O3. To our knowledge, its occurrence has never been reported through the experimental recreation of burnt archaeological materials. No correlation could be observed between the proxies of domain-state behaviour and deviation of palaeointensity determinations from the expected result.

Results obtained on clay samples heated in this type of ancient kiln can be considered a good source for determining the geomagnetic field strength variation in the past. Matching palaeointensity results obtained with different methods based on different principles can be taken as a quality criterion for result reliability and consistency.