Luminescence dating is an important tool for constraining sediment formation ages and post-depositional processes in many Quaternary environments. Ideally bright, stable, homogeneous, well-zeroed luminescence signals from materials with constant and readily measured dose rates would be used. However when dating material which crosses glacial cycle boundaries, in settings which have changed markedly during the periods under study, more challenging conditions may be encountered. Here we discuss marine cores from the Red Sea continental shelf near Farasan, collected as part of the DISPERSE project which concerns the environments associated with human dispersal along the Arabian coast. Luminescence profiling of two cores (FA12 and FA13) from the inner shelf suggested preservation of late Pleistocene material, currently at depths of 80-100m, overlain by finer grained holocene sediments. IRSL SARA dating was initially used to confirm this for core FA13. Here we present an extension of this work looking at management of signal stability, examining pre-dose of test dose response and its effect on SARA procedures, and evidence for uranium mobility particularly in the upper layers.
Thermally stabilised SARA approaches combine prolonged external preheating prior to first measurement, with later automated readout stages using shorter preheats in an attempt to minimise short-term fading of dating signals. Fading tests over 10^5 - 10^7 s storage periods produced mean decadal fading rates (1.2-3.6%) within overall uncertainties suggesting that this may have succeeded. However, test dose responses were seen to encode pre-dose dependence, with potential to bias dose estimates and response curves. These results are discussed here together with solutions based on additional cycles of external and within-reader irradiation and heating to equalise pre-dose conditions. Both test-dose normalisation and regenerated dose normalisation then give equivalent results, and dose estimates progress stratigraphically. Dosimetry using a combination of thick source beta counting (TSBC), and high-resolution gamma spectrometry (HRGS) revealed signifiance excess uranium (up to 10-15 ppm, from sediments containing 1-2 ppm Th), on the basis of 234Th analysis and excess beta dose rates (attributed to 234Pa) when comparing TSBC and HRGS. Further characterisation using alpha spectroscopy and ICPMS has been undertaken, and used to produce time dependent dose rate models for age estimation. The outcomes provide a chronology for the marine sediments and confirm the presence of late Pleistocene material on the inner shelf.