Introduction

Analysis of terrestrial sedimentary sequences in the Alpine region resulted in the ‘four-glaciation’ model of Penck & Bruckner, a model that was the paradigm for Quaternary science from 1909 until the 1970s. That this model was wrong was demonstrated only when cores of undisturbed and continuously accumulating sediment were retrieved from the deep-sea floor and their multiple oscillations of oxygen isotope ratio revealed. Yet this revolution in our view of the Quaternary might instead have taken place in north-central China rather than at the bottom of the ocean. For here exists the closest land-based analogue of the deep sea, with quasi-continuous accretion of sediment over the entire span of the Quaternary – the interbedded layers of windblown dust (loess) and buried soils (palaeosols) of the Chinese loess plateau (see cover photograph and Fig. 3.1). Safely distal from the erosive glacial processes of the mid-latitudes, these sediments record thirty or more alternations in climate. Transport of dust from source areas and ensuing dust deposition on the loess plateau were at a maximum during cooler, drier climate stages; weathering and soil formation were enhanced during warmer, wetter stages. These soil–loess oscillations were initially investigated as long ago as the 1930s (Thorp, 1936; Teilhard de Chardin & Young, 1930) and field mapping of the units was carried out through the 1950s (Zhu, 1958; Liu, T.S., 1959). Until the 1980s, it was thought that the timespan represented by the sediments was of the order of one million years.