Ice-volume-forced erosion of the Chinese Loess Plateau global Quaternary stratotype site.

The International Commission on Stratigraphy (ICS) utilises benchmark chronostratigraphies to divide geologic time. The reliability of these records is fundamental to understand past global change. Here we use the most detailed luminescence dating age model yet published to show that the ICS chronology for the Quaternary terrestrial type section at Jingbian, desert marginal Chinese Loess Plateau, is inaccurate. There are large hiatuses and depositional changes expressed across a dynamic gully landform at the site, which demonstrates rapid environmental shifts at the East Asian desert margin. We propose a new independent age model and reconstruct monsoon climate and desert expansion/contraction for the last ~250 ka. Our record demonstrates the dominant influence of ice volume on desert expansion, dust dynamics and sediment preservation, and further shows that East Asian Summer Monsoon (EASM) variation closely matches that of ice volume, but lags insolation by ~5 ka. These observations show that the EASM at the monsoon margin does not respond directly to precessional forcing.

A comparison of quartz OSL and isothermal TL measurements on Chinese loess.

The reliability of equivalent doses (De) from Chinese loess, measured using isothermal thermoluminescence (ITL) is tested. Dose calculations use the single-aliquot regenerative-dose (SAR) procedure. Despite good reproducibility of laboratory-induced signals and negligible response at zero dose, a significant overestimation of De is observed, compared with OSL measurements. Measurement of a known laboratory dose administered after optical bleaching, but before any heating, demonstrates that the first heating during measurement of the natural signal causes a significant sensitivity change, undetected by SAR. Using the single-aliquot regeneration and added (SARA) dose procedure, which allows for initial sensitivity change, good agreement with OSL is obtained after allowance is made for initial incomplete bleaching. It is concluded that SAR-ITL, in its present form, is not a suitable method for dating Chinese loess; it is very important to undertake a dose recovery test before any TL procedure is used to date sediments.