Multidomain behavior during Thellier paleointensity experiments: a phenomenological model

Abstract

The most commonly used method for paleointensity determination, the Thellier method, requires the absence of multidomain magnetic behavior. The influence of multidomain remanence on paleointensity determination and its recognition during the experiment are, however, only partially understood. Here, we present a phenomenological model of TRM acquisition and its application to the Thellier experiment. This model is an extension of the approach of [Earth Planet. Sci. Lett. 188 (2001) 45] and is used to investigate, beside other multidomain characteristics, the effect of pTRM*-tails. Furthermore, the appropriateness of typically applied checks during paleointensity determination like alteration checks, pTRM*-tail checks and additivity checks is scrutinized. It is shown that pTRM*-tails significantly affect the shape of the Arai diagram. Concave curvature of the Arai plot for samples containing multidomain particles is related to remanences with unblocking temperatures below their respective blocking temperatures. pTRM*-tails, however, which affect unblocking temperatures above the respective blocking temperatures, lead to a convex shape. An overlap of both convex and concave characteristics can produce s-shaped Arai diagrams [Phys. Earth Planet. Int.]. The effect of pTRM*-tails is strongly dependent on the angle and intensity difference between applied field and NRM of the sample. The commonly used in-field alteration check is biased by pTRM*-tails and shows deviations from the expected value even if alteration is absent. pTRM*-tail checks show the presence of tails and thus give evidence for multidomain behavior, only if the laboratory field is applied in a different direction than the NRM. The larger the angle between NRM and laboratory field, the larger the value of the pTRM*-tail check. A new method for calculation of the pTRM*-tail check is proposed which accounts for such angular difference, as well as intensity differences. Conducting the Thellier experiment with alteration checks, pTRM*-tail checks, and additivity checks at an angle between NRM and laboratory field > 0, provides the possibility to unequivocally identify multidomain bias.