Mariana Sontag-González , Bo Li, Kieran O'Gorman, Basran Burhan, Budianto Hakim, Adam Brumm, Richard G. Roberts

Volcanic sediments are challenging to date with luminescence-based methods. Both main minerals used for dating—quartz and K-rich feldspar—commonly have suboptimal luminescence properties when of volcanic origins, primarily a low signal intensity and, for K-rich feldspars, high rates of anomalous fading. The present work provides a case study of the Leang Bulu Bettue (LBB) archaeological site in Indonesia, a key site in our understanding of the early human occupation of the island of Sulawesi, showing that currently available state-of-the-art methods can provide robust chronologies for volcanic sediments. Through various reliability tests, we establish the suitability of a post-IR IR stimulated luminescence (pIRIR) standardised growth curve (SGC) procedure for De estimation of ‘micro-aliquots’ containing 5–10 grains of K-rich feldspar and apply these methods together with dose rate estimates to determine 10 new reliable finite ages for the sedimentary deposits at LBB. Further, by applying the LnTn method to circumvent truncation of the De distribution caused by the non-linearity of the dose response curve in the higher dose range, we could extend the site’s chronology to span ~20–210 ka. The new chronology is in keeping with previous radiocarbon and U-series ages for the site’s upper layers. We show that the low fading rate of micro-aliquots relative to using large aliquots composed of hundreds of grains lies in a brightness-based selection. By targeting bright grains with low fading rates, the need for fading corrections is greatly reduced and they can be made with a smaller uncertainty propagated through into the final age estimate, resulting in optical ages of improved accuracy and precision for volcanic sediments.