Monsoons are seasonal variations in atmospheric circulations accompanied by the reversal of dry and wet surface climate conditions, and are mainly forced by seasonal pressure system shifts driven by the annual and seasonal solar radiative cycle. On longer time scales, it can also strongly influenced by high latitude ice sheet. In East Asia, monsoon variability has profound societal and economic influences on this densely populated region. Understanding the past monsoon variability will provide insights into future monsoon climate projection.

The classic Quaternary East Asian summer monsoon records including Chinese stalagmite and loess, however, receiving much debate in the past years. An increasing number of works in recent years based on data-model approaches interpreted Chinese stalagmite δ¹⁸O as summer monsoon intensity. The loess record such as magnetic susceptibility can be strongly influenced by dust sedimentation rate. Thus relatively “pure” rainfall record in East Asia is still scarce.

The chemical composition of river sediments can be strongly influenced by chemical weathering, which mainly controlled by rainfall and temperature. The chemical weathering intensity can be evaluated for mobilized elements (e.g., K) during incongruent weathering of silicates by comparing their concentration to that of non-mobile elements (e.g., Al). As a result, we can use such element ratio to reconstruct the chemical weathering variability and further rainfall changes after the temperature and other effects being excluded.

Here we focused on the Yellow River, which is a large river in East Asia, and has an arid to semi-arid continental climate, being more arid and cold in the upper and middle reaches, and more humid and temperate in the lower reaches. We selected a set of river sediment samples distribute from upper to lower reaches, and further established a model covering variables including K/Al ratio (chemical weathering proxy), rainfall and temperature based on the multiple statistical analyses. Besides, we analyzed the K/Al ratio of IODP Site U1429 sediments in the East China Sea, which mainly supplied from the Yellow River during the last 400 ka based on sediment source tracing. Finally, we got a quantitative reconstruction of rainfall changes in northern China during the last 400 ka according to the K/Al-rainfall-temperature model.

Our rainfall record shows strong cycle of 23 thousand years and is consistent with the paleoclimate model simulation of annual rainfall in northern China. This rainfall pattern is consistent with insolation change, thus suggesting the insolation forcing on local rainfall (largely contributed from summer) variability. We further found the southern rainfall variation has a dominant cycle of 100 thousand years, based on the reconstruction-model comparison, suggesting the Northern Hemisphere ice sheet forcing. Here, an interesting finding is the simulated summer rainfall in southern China shows dominant cycle of 21 thousand years. We found such ice sheet signal in southern China annual rainfall is largely contributed from spring and autumn rainfall.

Our results highlight the seasonal contributions to orbital-scale of rainfall changes. The large proportions of spring and autumn rainfall in southern China indicate that rainfall proxies previously established in this region cannot be simply interpreted as summer rainfall indexes. Reconstruction of Quaternary summer monsoon based on the rainfall proxies in different regions of East Asia should be cautious in the future works.