Quantifying proportions of different material sources to loess based on a grid search and Monte Carlo model: A case study of the Ili Valley, Central Asia

Abstract

The Ili Valley is among the main distribution areas of loess deposits in Xinjiang Province, Central Asia, while the provenance of Ili loess remains under debate. In this study, samples from near-surface loess, two types of topsoil and modern riverbed sediment were analyzed for their concentrations of major and trace geochemical elements to determine the relative proportions of different provenances of loess deposits in different zones of the Ili Valley. The results obtained by the grid search technology and Monte Carlo model indicated that the proximal material is dominated in the Ili loess. Alluvial-diluvial sediments as the main local material source have significantly influenced loess in the western region of Ili Valley. Moreover, this influence gradually decreases in the eastern region and the Zhaosu Basin. The proportion of modern riverbed sediment in the eastern Ili Valley is significantly lower than that in the Zhaosu Basin and is lowest in the western Ili Valley. However, the proportion of dust and topsoil type-II with the mean value of 11.8% and 7.2%, respectively, is highest in the western Ili Valley and lowest in the Zhaosu Basin. The complex natural background of the Ili Valley can be used to interpret the quantitative results and the geochemical characteristics of Ili loess from different regions. The reliability of the proposed method can be assessed by environmental indicators such as grain size, and geomorphic-hydrological background and other published records.