Loess-palaeosol sequences in diverse environments: Aeolian accumulation identification and magnetic susceptibility models

The thick loess-palaeosol sequences of the Chinese Loess Plateau (CLP) are valuable continental archives of paleoclimatic and paleoenvironmental changes since at least the mid-Miocene. The magnetic susceptibility record of loess sequences is frequently used for palaeoclimatic reconstruction. However, the two magnetoclimatological models for the magnetic enhancement of loess-palaeosol sequences—the wind vigour model and the pedogenic model—are still debated, and neither model alone is sufficient to explain the magnetic susceptibility stratigraphy of loess deposits in diverse environments. Loess can be substantially altered by its environment of deposition and the macroscopic characteristics of loess may differ substantially from the well-documented sequences of the CLP, which complicates loess identification and its palaeoclimatic interpretation. In the present study, we summarize the attributes of loess deposits in different environments and review possible approaches for identifying aeolian accumulation. Loess-palaeosol sequences in various climatic contexts, from arid to subhumid, are summarized and categorized and the relationship between the magnetic susceptibility and effective precipitation is evaluated. A model comprising three cases is proposed to interpret the magnetic susceptibility of loess-palaeosol sequences in a variety of different environments. In case A (e.g., the Xinjiang loess, NW China), where physical weathering dominates and pedogenesis is weak, there is no relationship between the magnetic susceptibility and pedogenesis, and here the wind vigour model is applicable. In case B (e.g., the CLP), where, under predominantly oxidizing conditions, the magnetic susceptibility increases with pedogenic intensity, and an oxidizing-pedogenic model is applicable. In case C (e.g., Alaskan and Siberian loess), there is a negative relationship between magnetic susceptibility and pedogenesis due to the dissolution of detrital iron oxides under reducing conditions, and a reduction-pedogenesis model is applicable. In Case C, pedogenesis (or chemical weathering/dissolution) acts to largely obliterate the detrital magnetic susceptibility signal. Hence, contrary to the original interpretation of the magnetic susceptibility of Alaskan and Siberian loess, the wind vigour model is not applicable. Two thresholds of mean annual precipitation between the three cases are proposed, which enable the paleoclimatic interpretation of loess magnetic susceptibility records in different environments.