Testing the application of post IR-IRSL dating to fine grain waterlain sediments

Abstract

The polymineral fine grain fraction of a set of ten known age samples were used to test the suitability of elevated temperature post-IR infrared stimulated luminescence (pIRIR) to date waterlain sediments in Switzerland. Equivalent Dose (De) values were obtained using elevated temperatures of 225 °C (pIRIR225) and 290 °C (pIRIR290) and IRSL50/225 (preceding measurement of pIRIR225), and these displayed a systematic increase with increasing stimulation temperature, which brought those derived from the pIRIR protocols into saturation for the older half of the samples (>100 ka). Significant residuals were recorded for the pIRIR signals, and these were found to further increase following attenuation of the bleaching source confirming that these signals are harder to bleach than the IRSL measured at 50 °C. Fading tests recorded g-values of between 1 and 4%, and displayed no obvious decrease with increasing stimulation temperature. For the younger half of the samples, fading corrected ages for all protocols resulted in overestimation compared to robust independent dating for all samples, and draw into question the relationship of fading measured in a laboratory generated signal to that which is naturally generated. Uncorrected IRSL50/225 ages for all samples were in good agreement with independent dating, but still displayed a general overestimation using pIRIR signals. These overestimations may largely be due to the significant residual doses resulting from the harder to bleach nature of the elevated temperature signal, and present a considerable problem when dating waterlain sediments. A considerable challenge when using the pIRIR signal is to be able to separate the effects of both residual doses and fading and confirm that one is not masking the other. Despite this, the pIRIR protocols were successful at one site, although they offered no benefit compared to dating with more conventional luminescence signals. A comparison of De values derived from a standard IRSL signal to that from IRSL50/225, suggest that the latter overestimates by ∼10% due to a decrease in sensitivity of the sample prior to measurement of the test dose.