Spatiotemporal variations of Zr/Rb ratio in three last interglacial paleosol profiles across the Chinese Loess Plateau and its implications for climatic interpretation

The ratio of Zirconium to Rubidium (Zr/Rb) is suggested to be a better proxy for the East Asian winter monsoon strength than the widely-used grain size distribution. The rationale for the Zr/Rb proxy relies on the following assumptions: (1) Grain size fractionating characteristics during eolian dust transport should be archived in the Zr/Rb ratio records and this assumption is based on the premise that Zr is preferentially enriched in coarser grain size fraction while Rb tend to be enriched in finer grain size fraction; and (2) post-depositional weathering does not change the Zr/Rb ratio due to the immobility of these two elements. To examine these two assumptions, three last interglacial paleosols (S1) from Dingxi, Tianshui and Lantian, along a NW-SE transect across the Chinese Loess Plateau, were geo- chemically investigated. Our results show that the Rb concentration exhibits an increasing trend along the NW-SE transect both in the paleosol (S1) and the measured portions of the loess units (L1 and L2), being supportive to the assumption that Rb is enriched in the fine particles. But we also found that Rb loss did occur to some extent in the three profiles, contradicting to the presumption of Rb immobility during pedogenic processes. The Zr concentration exhibits an expected decreasing trend in the measured portions of the loess units and an unexpected increasing trend in the paleosol along the NW-SE transect. Moreover, the ratios of Zirconium to Hafnium (Zr/Hf) show different variation patterns between interglacial and glacial, implying that Zr-bearing minerals and their resident grain size frac- tions are probably not identical during interglacial and glacial. Thus, the assumption that Zr is enriched in coarse grain size fraction can no longer hold. We conclude that the final Zr/Rb value is not only de- pendent on grain size sorting processes but also on post-depositional alteration and source prove- nance. Under enhanced chemical weathering, especially when chemical index of alteration (CIA) is greater than 65, the Zr/Rb value loses its validity to be a reliable proxy for the winter monsoon strength. In addition, all the Zr/Rb value is constrained above 1.7, a minimum which is definitely modulated by all the three aforementioned factors.