Quantitative relationships between magnetic enhancement of modern soils and climatic variables over the Chinese Loess Plateau

摘要

Environmental magnetism has been widely employed to reconstruct past climate changes on the Chinese Loess Plateau (CLP), and several climofunctions based on the magnetic properties of loess have been developed. However, systematic investigation of the quantitative relationship between topsoil magnetic enhancement and modern climate remains uncommon. In this study, we obtained surface soil samples from 257 sites over the CLP and adjacent regions. From this set, we used 180 samples from sites unaffected by potential contamination to investigate the relationship between the commonly measured magnetic properties of magnetic susceptibility and magnetic remanence and modern climatic variables. The spatial distribution of the results demonstrates a strong NW-SE gradient of the magnetic enhancement of surface soils. The results of more detailed magnetic parameters indicate that pedogenic viscous superparamagnetic and stable single-domain particles are mainly responsible for the magnetic enhancement of the soils in the studied region; and that the magnetic grain-size distribution of the ferrimagnetic components of pedogenic origin remains almost constant, independent of pedogenic intensity. The uniform mechanism of magnetic enhancement, mainly linked with the concentration of the pedogenic components rather than with variations in magnetic grain-size, decreases the level of ambiguity in climate reconstructions based on magnetic measurements of Chinese loess. Statistical analyses, including correlation, Principal Component Analysis (PCA) and multiple regression analysis, suggest that annual rainfall rather than temperature exerts the dominant effect on soil magnetic enhancement Finally, we used the results to develop several transfer functions to reconstruct mean annual precipitation (MAP). Transfer functions based on frequency-dependent susceptibility (X_(fd)) and anhysteretic remanent magnetization (X_(ARM)) provide the most reliable estimates of MAP. This study significantly improves the understanding of the relationship between soil magnetic properties and climate.