Available online at www.sciencedirect.com Radiation Measurements 37 (2003) 493 – 498 www.elsevier.com/locate/radmeas Towards a prediction of long-term anomalous fading of feldspar IRSL M. Lamothe ∗ , M. Auclair, C. Hamzaoui, S. Huot D epartement des Sciences de la Terre et de l’Atmosph ere, Universit e du Qu ebec  a Montr eal, CP 8888, Succursale Centre-Ville, Montr eal, QC, Canada H3C 3P8 Received 23 August 2002; received in revised form 5 November 2002; accepted 11 November 2002 Abstract Anomalous fading of the infrared stimulated luminescence in feldspar is described by a new equation in which the unwanted loss of luminescence is given as a function of three variables, the measured fading rate, the laboratory radiation dose rate, and the environmental radiation dose rate. This equation is labelled dose rate correction (DRC) and it is tested on (a) young sediment samples for which the radiation growth curve is in the linear part of the dose–response curve, and (b) geologically old sediments in luminescence eld saturation. Most of the experimental results are consistent with predictions that can be de- duced from this equation. Given the common range of dose rates in luminescence laboratories and in natural soil environments, the time necessary for the unstable luminescence generated by articial irradiation to fade away is predicted to be ca. 10 4 –10 5 years. c  2003 Elsevier Science Ltd. All rights reserved. 1. Introduction The luminescence of feldspar, thermally or optically stim- ulated, is known to decay without stimulation after a lab- oratory irradiation, contrary to kinetic expectations. This loss of luminescence is generally thought as resulting from quantum mechanical tunnelling (Visocekas, 1979). In order to derive true ages for geological or archaeological events using feldspar IRSL, dierent approaches have been sug- gested to correct for the observed fading. Among those, two apparently dierent methods have been proposed, namely the fadia (Lamothe and Auclair, 1999) and g (Huntley and Lamothe, 2001) methods. The fadia approach uses dierential decay of single grains or single aliquots from the same sample. If the feldspar grains or aliquots have received the same dose in the nat- ural environment, one can derive the equivalent dose for a non-fading feldspar hence the true paleodose. This fadia method was the rst approach to oer potential corrections ∗ Corresponding author. Tel.: +1-514-987-3000x3361; fax: +1-514-987-7749. E-mail address: lamothe.michel@uqam.ca (M. Lamothe). for fading in sediment samples of any age up to saturation. However, diculties arise when the sample is composed of more than one grain population, as for partially bleached sediments, or when there is not enough variability in anoma- lous fading rates between dierent aliquots. The correction method of Huntley and Lamothe (2001) is based on the measurement of the percentage fading loss of luminescence per decade of time, a parameter termed g by Aitken (1985, Appendix F). The correction is obtained by integrating the loss of luminescence over geological time, the resulting luminescence intensity being I f . This is com- pared with the intensity of luminescence observed in the laboratory before fading (Io). This approach introduces a correction by assuming that I f =Io = T f =T , where T is the true age and T f is the age measured before correction. The ad- vantage of this approach is that fading can be corrected on a single aliquot. However, it can be applied only to the linear part of the growth curve, and hence to young samples. The purpose of this paper is to introduce and test a new idea for correcting fading in feldspar minerals that would cir- cumvent the limitations described above. In other words, it should be applicable to a single aliquot and over the whole range of radiation dose, up to saturation. We rst describe a new equation stating that the loss of luminescence 1350-4487/03/$ - see front matter c  2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S1350-4487(03)00016-7