Reconstructing aeolian activities and borders shifts of the Gonghe Sandy Lands since the last Glacial Maximum

Abstract

Desert dune systems are vital components of Earth's surface landscapes. Understanding historical activities and delineating desert dune field boundaries under different climatic conditions are important for identifying both paleoenvironmental drivers and the likelihood of future reactivation. The Gonghe Sandy Lands (GSLs), situated in the northeastern Qinghai-Tibet Plateau, are at the convergence zone between the Westerlies and Asian monsoon, which are sensitive to climatic change. Consequently, GSLs are ideal regions for studying the response of the Earth's surface to environmental factors. A regional stratigraphic statistical model was developed using 30 aeolian profiles with 194 ages (including seven new profiles with 38 new optically stimulated luminescence ages) in the Gonghe Basin. Aeolian activities, climatic changes, and borders shifts in GSLs since the last Glacial Maximum (LGM) were reconstructed using the model together with other records. The results demonstrate that (1) Since the LGM, the Gonghe Basin has experienced a climatic transition from cold and dry to warm and humid, and then to mild and subhumid. Aeolian activity weakened then strengthened over the changing climatic conditions since 22 ka. (2) The spatial configurations of GSLs changed markedly throughout different periods. The area reached 3490 km² during the LGM, contracted to 903 km² during the Middle Holocene, and expanded again at 5.5 ka. (3) Climatic changes, human activities, and topographical constraints shaped the spatial patterns of the GSLs. Before the Middle Holocene, the dynamics were influenced by the monsoon system. In the late Holocene, human activity became an important factor affecting the spatial patterns of GSLs. Nevertheless, the influence of topographical constraints on the sand dune distribition in GSLs cannot be overlooked. The model outputs reflect the frequency of aeolian activities, independent of age errors, and are particularly suitable for sandy land regions that lack LGM ages.