Published online by Cambridge University Press: 20 January 2017
Magnetic susceptibility profiles χ(z) in loess sequences reflect a combination of two climatically modulated processes: dust deposition and pedogenic development. Prominent soils with high χ values, for example, likely reflect periods of slow dust deposition, and warm wet conditions favorable for rapid chemical weathering. To refine our understanding of the climate records contained in loess, we develop a numerical model as a tool for exploring quantitatively the integrated record of the temporal variation in rates of loess aggradation and soil development contained in χ profiles. In our model, the aggrading loess is pedogenically altered in a reactive zone near the ground surface. The strength of the χ signal is dictated by both the depth-dependent intensity of pedogenic processes and the rate of dust accumulation, which dictates the total time a loess parcel spends in the near-surface reactive zone. The model can accommodate both pedogenic production of magnetically susceptible minerals and arrival of magnetically susceptible grains as eolian dust. To explore the model performance and develop a sense of time and length scales implicit in χ profiles, we first examine simple synthetic cases with idealized steady and cyclic climatic forcing. Reported χ profiles in three Chinese loess sequences at varying distances from the western China dust source are then modeled in two illustrative ways: (i) by imposing a specific dust deposition rate history that is proportional to the dust accumulation rate history reported from western Pacific deep-sea cores, allowing the time variation of pedogenic rates to be calculated directly from χ profiles; and (ii) by imposing both dust accumulation and pedogenic rate histories that are independently scaled by the deep-sea δ18O history, which reflects global climate cycles to which regional climate forcing is linked. We find that the zone of pedogenic activity is roughly 0.5–1.0 m thick, both deposition rate and pedogenic intensity have varied dramatically over the last 140,000 yr, the age structure of the Luochuan loess sequence is best fit by driving the model with the δ18O record, and environmental conditions must have been anomalously favorable for pedogenesis during isotope stage 3 at all three sites. Finally, we advocate the assembly of a variety of data types at a suite of sites within any loess field that taken together will better constrain the temporal and spatial patterns of climatically modulated deposition and pedogenic processes.