Exploring the use of a low temperature preheat in IRSL dating of feldspar in Beringian archaeological contexts

Abstract A new luminescence dating protocol was developed to circumvent the problem of thermal transfer and poor thermal stability of Infrared Stimulated Luminescence (IRSL) of samples from late glacial loess sediments in Central Alaska, USA. This new protocol is proposed as an alternative to routine procedures for which scattered and erratic results were obtained using standard IRSL and post-IR IRSL protocols on loess samples from the McDonald Creek archaeological site. The difficulty to date Alaskan loess has been acknowledged in the past, but the reasons for this phenomenon have not been clearly identified. This difficulty is even greater for young samples, as shown in this study. It is suggested here that the particular mineralogical composition of the Central Alaskan loess can be related to problematic luminescence behaviour. To overcome problematic results, we developed a Low Preheat IRSL (LPH-IRSL) protocol. This protocol is minimally affected by the presence of other minerals than K-feldspars in the samples, such as albite, muscovite and clinochlorite, which can be very difficult to eliminate during the preparation steps prior to IRSL measurements. Dose recovery experiments and multiple other tests indicate that a protocol with low preheat temperature yields the best results in this specific context. Thermally transferred luminescence induced during the measurement is kept at a minimal level. Consequently, considering independent age controls available from radiocarbon, equivalent doses are successfully determined, their distribution is narrow, and the g-value has low scatter. This protocol has great potential for dating polymineral samples, whether it is silt-sized sediments like loess, fine-grained like clay and micro-samples like art pieces, common in archaeological contexts. The chronological results obtained for this project suggest a human occupation of the area during the Late Pleistocene to mid-Holocene period, which is in excellent agreement with preliminary radiocarbon chronology obtained during site testing and also with the regional archaeological framework.