Reconciling the 'westerlies' and 'monsoon' models: A new hypothesis for the Holocene moisture evolution of the Xinjiang region, NW China

The Holocene evolution of moisture conditions in the arid Xinjiang region in inland NW China is increasingly debated. Analysis of climatic reconstructions from lake sediments, loess, sand-dunes and peats reveals a long-term wetting trend ('westerlies model'), whereas other studies, mainly of lake sediments and stalagmites, have indicated a long-term drying trend ('monsoon model'). Two observations can be made about these conflicting interpretations: (i) The 'monsoon model' records are mainly from high-altitude mountain areas with higher precipitation, whereas the 'westerlies model' records are mainly from low-altitude piedmont areas and basins, with lower precipitation. (ii) An increasing number of studies of the Xinjiang region and its surroundings demonstrate a long-term warming trend during the Holocene. Based on these observations, we propose the following simple hypothesis which is a compromise between the 'westerlies' and 'monsoon' models. Against the background of long-term Holocene warming, the increased meltwater supply from snow and ice in high-altitude mountain areas fed the low-altitude lakes, which finally resulted in higher lake levels, expanded oasis areas, and enhanced local moisture recycling. This potentially explains why the 'westerlies model' evidenced in the moisture records from lakes and adjacent loess and peat deposits comprises a long-term wetting trend. This hypothesis is largely consistent with the parts of the region that are currently experiencing climatic warming. More importantly, with the likely acceleration of climatic warming in the near future, the further ablation of ice and snow in the high-altitude mountain regions will present a major and growing threat to the sustainable socioeconomic development of Xinjiang. Consequently, it is important that both the government and the public seriously consider the implications of an ice-free environment in the region.