Abstract

The widespread aeolian deposits on the Chinese Loess Plateau (CLP) have been investigated intensively as an outstanding archive of information on past variations of the East Asian Monsoon (EAM). However, differentiating the impacts of precipitation and temperature on loess-based proxies is difficult and complex, which hampers our ability to understand the variability and dynamics of the EAM. In this study, we investigated the Mg isotope composition (δ26Mg) of the secondary carbonates in fine-grained loess samples from ten Holocene profiles and a loess-paleosol sequence, as well as in the primary carbonates of nine loess-source samples from the Asian inland deserts. Our aim was to explore the potential of the δ26Mg values of secondary carbonates in tracing precipitation changes. The results demonstrate that δ26Mg values are homogeneous (−2.07‰) in primary carbonates from different loess-sources, but display large variations in secondary carbonates of loess, increasing from −3.33‰ in the northwest CLP to −1.80‰ in the southeast CLP, and from −3.58‰ during the last glacial to −1.65‰ during the last interglacial. The variations were mainly controlled by EAM precipitation via the migration of isotopically light Mg, because Mg isotope fractionation during the formation of secondary carbonates is temperature insensitive, and the primary carbonate in different loess-sources has relatively constant δ26Mg values. This conclusion is further supported by high correlation (r2 = 0.84) between modern mean annual precipitation (MAP) and δ26Mg in the secondary carbonates of spatial loess samples. Based on this δ26Mg-MAP relationship, temporal variations in MAP were quantitatively estimated, ranging from 670 mm during the last interglacial to 270 mm during the last glacial at the southern CLP. Compared with previous estimates, our MAP exhibits a larger amplitude of glacial-interglacial fluctuations (∼400 mm) with a less uncertainty (53–76 mm), implying that δ26Mg values in secondary carbonates of loess can serve as a novel precipitation proxy to infer EAM variability and dynamics.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call