Central Asia is the homeland of the traditional Silk Road, a trade and cultural corridor across the Eurasian continent, which played a key role in cultural exchanges between the East and the West, and advanced the development of civilization in its vicinity. Many famous ancient civilizations, such as the Loulan and Kuzi, once flourished in Central Asia, but eventually disappeared, likely affected by past climate change. In light of this, it is of great interest to study the past evolution of aridity in Central Asia, which could help us to understand the climate during the Silk Road development, as well as future changes in the region of the Silk Road economic belt.

Unlike the East Asian monsoon climate, the Central Asia is mainly controlled by the mid-latitude westerly winds. This region is the largest arid area in the mid-latitudes of the Northern Hemisphere, one of the major sources of atmospheric dust in the world (see figure). The prevailing westerly winds in the mid-latitudes allow for the long-distance transport of the eolian dust generated in Central Asia with much of it is finally deposited in the northwest Pacific Ocean. Low sedimentation rates in the deep open ocean mean that eolian dust can play an important role in determining the composition of ocean sediments. Hence, geological archives, such as those of ferromanganese (Fe-Mn) crust, from the Pacific Ocean could provide a valuable information of past climate change in Central Asia over thousands to millions of years.

In 2018, we set out to study a particular Fe-Mn crust sample (CM3D18, see figure), which had been collected from the Marcus-Wake Seamount Group in the western Pacific Ocean. Based on high-resolution multi-proxy geochemical and magnetic analyses, we were able to reconstruct past variations in dust sources and their links to Central Asia’s varied aridity over the past seven million years.

Our reconstructed potassium/aluminium (K/Al) ratio, representing a chemical weathering index, reveals that the dust source areas were drier during the warm Pliocene (5.3-2.6 million years ago). Following the onset of Northern Hemisphere glaciation and gradual glacial intensification during the Pleistocene (last 2.6 million years), the dust fluxes increased, which could be explained by the effects of both glaciation and increased moisture availability providing a greater supply of fine-grained material inthe source areas. Furthermore, we also identify a gradual shift in the lead (Pb) isotope compositions in the Fe-Mn crust, suggesting a gradual provenance change in the dust source regions, from a dominant Gobi Desert source to a mixed Gobi-Taklimakan Desert source following the global cooling in the past 5 million years.

We were able to explain the above observations by comparing them to model simulations using the Alfred Wegener Institute earth system model (AWI-ESM). Under an elevated atmospheric CO₂ forcing_, representing the Pliocene, we found increased potential evaporation in the Central Asia, resulting in a hotter and drier climate. This scenario could explain the signature of reduced chemical weathering of the sediments, while the arid conditions also restricted the creation and supply of fine-grained dust to be transported to the Pacific Ocean. Meanwhile, the warmer global temperatures led to the displacement of the westerly jet towards higher latitudes, which could explain the dominance of dust supply from the Gobi Desert. Following the Pleistocene cooling, those Pliocene changes were reversed, which led to a southward shift of the westerly winds and enhanced moisture availability in the Central Asia continental interior, driving an increased dust flux.

Overall, our study provides a paleoclimate analogue, suggesting that the on-going anthropogenic warming could systematically increase the aridity over Central Asia, a region already suffering water shortages now.