Constraining Holocene lake-level highstands on the Tibetan Plateau by 10Be exposure dating: a case study at Tangra Yumco, southern Tibet

Abstract Many lakes on the Tibetan Plateau are surrounded by palaeo-shorelines that document former lake-level highstands and hence past changes in climate. Establishing accurate chronologies for these shorelines is crucial to correlate lake-level fluctuations with palaeo-environmental histories that are commonly inferred from lake sedimentary records. Here we apply surface exposure dating to constrain the age of two lake-level highstands at one of the largest Tibetan lakes – Tangra Yumco – which has well developed palaeo-shorelines located up to ∼185 m above the current lake. For sampling we focussed on wave-cut bedrock terraces that vary considerably in width, because they were carved into rocks of different erodibility. Samples from the highest strath terrace (∼180–185 m above the lake) at three different sites yield tightly clustered 10 Be ages between 7.39 ± 0.19 ka and 7.87 ± 0.27 ka (internal errors), which constrain the end of an early Holocene lake-level highstand at 7.6 ± 0.6 ka (external error). The consistency of the 10 Be ages indicates that bedrock erosion by wave attack reduced the inherited 10 Be component to negligible amounts, because otherwise samples would show considerable age scatter. Two bedrock samples from a lower terrace (∼140–145 m above the lake) and one amalgamated clast sample from a beach ridge on the same terrace yield 10 Be ages between 4.04 ± 0.14 ka and 4.50 ± 0.15 ka. These ages document that the lake level dropped by ∼40 m between ∼7.6 ka and ∼4.3 ka. Granite boulders from the two investigated terraces yield apparent 10 Be ages that are ∼6 ka and ∼27 ka older than the ages of the respective terraces, indicating that this sample type is not reliable in this setting. Overall, our results demonstrate that exposure dating of bedrock terraces provides a valuable tool to reconstruct lake-level histories on the Tibetan Plateau.