PARTICLE SIZE DISTRIBUTIONS OF THE LAST INTERGLACIAL PALEOSOL S1 PROFILES IN THE CHINESE LOESS PLATEAU: THEIR PEDOGENIC AND CLIMATIC IMPLICATIONS

The last interglacial paleosol S1 gradually differentiated from the northwest to the southeast. At the Lanzhou and Dingxi sections in the northwestern part of the Loess Plateau, the three paleosols (S1S1, S1S2, S1S3) corresponding to marine isotope sub-stages 5a, 5c and 5e and the two intervening loess units (S1L1, S1L2) corresponding to marine isotope sub-stages 5b and 5d are completely preserved and the post-depositional in situ weathering and the post-weathering translocation of clay and carbonate did occur in the paleosols but only to limited extents. At the Tianshui section in the southern part of the western Loess Plateau, the three paleosols were partially welded, and the post-depositional in situ weathering and the post-weathering translocation within the S1 profile occurred more pronouncedly and the S1 intruded into the underlying older and coarser loess L2 more deeply than in the northwestern part. The three soil-forming events occurred repeatedly just in one single soil profile at the Lantian section in the southeastern part of the Loess Plateau and a major portion of the S1 paleosol developed in the underlying older and coarser loess L2. Undoubtedly, the post-depositional in situ weathering and the post-weathering translocation within the S1 profile at the Lantian section were the predominant processes probably during the entire period of the last interglacial. The Qingyang section is an expanded version of the Lantian section. The laboratory data at the Huanxian section in the northern part of the eastern Loess Plateau are quite similar to those at the Qinan section in the central part of the western Loess Plateau. At all sections investigated with an exception of the Lantian section where the particle-size difference between the L1 and S1 was blurred by a strong weathering of the L1, not only are the paleosols but also the loess units S1L1 and S1L2 within the S1 remarkably distinguishable from the overlying L1 and underlying L2 by particle size distribution, implying that the S1 parent material was considerably finer in the source areas during interglacial periods than during glacial periods. To conclude, because a number of factors might have involved in shaping the particle size distribution curve in the S1 profiles, it is thus unrealistic to reconstruct high-resolution climatic records from the S1 paleosols.