Luminescence dating of glacial sediments of penultimate glaciation in SE Tibetan Plateau using single grains of K-feldspar.

Luminescence dating is a useful chronological tool for glaciogenic sediments, especially when the targeted sediments lack organic materials or are too old for radiocarbon dating. However, problems like partial bleaching make its application on glaciogenic sediments challenging. Poor luminescence characteristics and low saturation level of quartz limit its application in glacial environments, especially dating old glacial sediments. Developments of the single-grain technique and post-infrared infrared stimulated luminescence (post-IR IRSL) protocols offer opportunities for mitigating or overcoming these problems. Here we test the use of single grains of K-feldspar for luminescence dating of glacial sediments of penultimate glaciation: Guxiang glaciation, the largest known Quaternary glacier advance in SE Tibetan Plateau and was previously constrained to marine isotope stage 6 (MIS-6) by cosmogenic nuclide 10Be exposure dating [1]. We took 6 samples (GX601 to GX606) from glaciofluvial sand lens within the lateral moraines of Guxiang glaciation and 2 samples (GX607 and GX608) from overlying loess and paleosol. Dose recovery tests show that the two-step post-IR50 IRSL225 protocol was able to recover the given dose and assumed suitable for this study. The equivalent doses (Des) of each sample show large scatter, suggesting partial bleaching of these samples. It is therefore unsurprising that Des derived from the minimum age model (MAM) are smaller than those derived from the central age model (CAM). Measurement of the fading rates show that, g-values of the IR50 signal range between 3-6%, whereas those of the pIR50 IRSL225 signal are all less than 3%. After correction of the fading rates, the final ages derived from the pIR50 IRSL225 signal fall into MIS-6 if the CAM is used, and fall into MIS-5 if the MAM is used. In addition, we compared the single-grain K-feldspar and quartz ages of the younger overlying loess and paleosol samples. For both minerals, the MAM ages are smaller than the CAM ages. Moreover, the MAM ages of feldspar (~4 ka for GX607 and ~20 ka for GX608) are similar to the CAM ages of quartz, and thus larger than the MAM ages of quartz (~2 ka for GX607 and ~10 ka for GX608). This seems indicate the residual doses of feldspar are probably larger than those of quartz. For most of these samples, there is a trend that the brighter the grains, the lower their Des. The discrepancies between those results of CAM and MAM age models, quartz and feldspar, dim and bright feldspar grains, and comparation between luminescence and 10Be exposure ages, will be investigated and discussed. [ABSTRACT FROM AUTHOR]