Archaeomagnetic investigations in Bolgar (Tatarstan)

Abstract

The objective of this study is to provide a well-dated point for a future palaeosecular variation (PSV) reference curve for western Russia. For this purpose archaeomagnetic and magnetic property analyses were carried out on a pottery kiln unearthed at the UNESCO World Heritage site of ancient Bolgar, having a rather precise age dating. The archaeological context provided an age between 1340 and 1360 C.E. The characteristic remanence vector was determined through alternating field demagnetisation and Thellier-Thellier palaeointensity experiments. Some innovations were introduced regarding palaeointensity. The check testing the equality of blocking and unblocking temperature was redefined. This allowed waiving the commonly used additional zero-field cooling steps during the Thellier-Thellier experiment. Another innovation concerns the calculation of archaeointensity at structure level. A Bayesian approach was introduced for averaging individual specimen archaeointensities using a prior probability distribution of unknown uncertainties. Next, an additional prior probability distribution was used to correct for cooling rate effects. This resulted in a lower uncertainty compared to common practice and in eluding time consuming cooling rate experiments. The complex magnetic mineralogy consists of maghaemite, multi-domain haematite and Al-substituted haematite. Some samples contained also some non-stoichiometric magnetite. The magnetic mineralogy was determined through hysteresis loops, backfield and remanence decay curves, measurements of the frequency-dependent magnetic susceptibility and through low temperature magnetisation curves. Accompanying hightemperature thermomagnetic analyses revealed an excellent thermo-chemical stability of the studied specimens. Directions obtained from alternating field demagnetisation and those extracted from archaeointensity experiments are congruent and have low uncertainties. The obtained archaeomagnetic results are fairly in agreement with global geomagnetic field models and contemporary PSV data of the wider area. The

Bias Corrected Estimation of Paleointensity (BiCEP): An Improved Methodology for Obtaining Paleointensity Estimates

“…In this way, the BiCEP method is quite different from the recently published Bayesian approach of Kosareva et al. (2020). When tested on a compilation of sites for which an approximate paleointensity is known a priori , our method yields levels of accuracy and precision similar to, or better than restrictive paleointensity criteria, whilst accepting as many results as permissive criteria. Our method generates some slightly inaccurate paleointensity estimates with high levels of precision, but these can generally be explained with inaccuracies in the expected field (see Section 4.4).…”

“…In this way, the BiCEP method is quite different from the recently published Bayesian approach of Kosareva et al. (2020).…”

“…Instead of excluding specimens from the paleointensity analysis in the traditional (binary) selection criteria based approach, our method predicts an amount of bias for each specimen, using the curvature of the Arai plot as a metric of nonlinearity and a predictor of bias. In this way, the BiCEP method is quite different from the recently published Bayesian approach of Kosareva et al (2020). 3.…”

“…We discuss this workflow in more detail in Section 4.3. The property of increasing precision with number of specimens is inherent to Bayesian models and can also be found in the method of Kosareva et al (2020), although their method does not include the bias correction found in our method.…”