Timing of mega-lakes in the central-southern Tibetan Plateau constrained by K-feldspar single-grain pIRIR dating of paleo-shorelines

Abstract

Lake evolution in the Tibetan Plateau (TP) during the late Quaternary exhibits pronounced spatiotemporal complexity, manifested by mega-lake formations during MIS 5 in the northeastern TP (NETP) versus MIS 3 and early Holocene lake high-stands in the central-southern TP (CSTP). While quartz optically stimulated luminescence (OSL) remains the most widely used dating method for reconstructing paleolake levels across the TP, its reliability at specific lakes has recently been questioned due to the high contribution of medium component and low sensitivity and stability, prompting a critical re-evaluation of paleolake chronologies. Here, we determined the ages of thirteen mega-lakes in the CSTP by dating the highest (or near-highest) paleo-shorelines using both quartz single-aliquot OSL and K-feldspar single-grain (SG) post-infrared infrared stimulated luminescence (pIRIR) techniques. Dimensions and water storage changes of these mega-lakes were reconstructed. Our results indicate that quartz OSL signals exhibit thermal instability, resulting in age underestimations, particularly for samples older than ∼40 ka, potentially explaining previous MIS 3-restricted paleolake chronologies. By contrast, K-feldspar SG-pIRIR dating provided robust age constraints, revealing that the mega-lakes primarily formed during the last deglaciation-early Holocene (17–15 ka and 12–7 ka), while a subset of lakes record MIS 5 mega-lake phases (106.7–97.6 ka and ∼ 70 ka). We hypothesize that enhanced Indian summer monsoon intensity coupled with increased glacial meltwater drove mega-lake formations during MIS 5 and the early Holocene, whereas the last deglacial mega-lakes may reflect meltwater flux variations.