Development and application of new transfer functions between climate and soil weathering indices for paleoclimatic reconstructions from Chinese loess-paleosol sections

The degree of weathering and soil formation under similar conditions with respect to parent material, relief position, vegetation, and age mainly depends on the climatic conditions. Therefore, various weathering indices have been used successfully to reconstruct past mean annual temperature (MAT) and mean annual precipitation (MAP) based on the expression of these weathering indices in paleosols. Yet, no specific soil-climate transfer function has been established to reconstruct paleoclimates based on paleosols in Chinese loess profiles. In this study, we analyzed the relationships between several weathering indices and present climatic conditions for 49 soils throughout different climate zones in China (21°42′–47°54′ N), which were all developed in loess or similar Quaternary silty sediments. We aimed to establish soil-climate transfer functions that can be widely applied to paleosol-based paleoenvironmental reconstructions in Chinese loess archives and to apply them to a pedosedimentary record in northeastern China, where paleoclimatic information is still scarce. Overall, MAT showed the strongest relationship with the weathering index ba1 ((K2O + Na2O)/Al2O3), with MAT = –29.887 ba1 + 22.361 (R2 = 0.788), and MAP showed the closest relationship with the Chemical Proxy of Alteration (CPA: Al2O3 × 100/(Al2O3 + Na2O)), with MAP = 0.0843 CPA3 – 20.143 CPA2 + 1630.7 CPA – 44,244 (R2 = 0.898). We applied these two transfer functions to a Late-Pleistocene pedosedimentary record in Shenyang, Liaoning Province, to reconstruct MAT and MAP from the onset of the Last Glacial Period until the Last Glacial Maximum. Our reconstruction yielded a warm-humid climate until the Marine Isotope Stage (MIS) 5/4 transition and a trend towards a cooler and drier climate thereafter, interrupted by a somewhat warmer and more humid period during MIS 3. Despite several limitations, the newly established transfer functions can be applied to paleosols in loess-paleosol sections across the Chinese monsoon region and other regions around the world whose climates and parent materials are within the range covered by this study, because of the large supra-regional distribution of the soils included in this study. Thus, the new functions provide a new valuable tool for obtaining a substantial set of new paleoclimatic proxy data for various loess regions.