loess deposition and palaeoclimate

loess deposition and palaeoclimate Since the 1970s, the analysis of oxygen isotope ratios (¹⁸O/¹⁶O) in foraminifera in deep-sea cores, particularly some from the equatorial Pacific, has provided an unprecedented proxy record of climate change throughout Quaternary times. In essence, increased concentrations of ¹⁸O in foraminifera (unicellular marine organisms) within the deep-sea sediment reflect colder times of lower sea levels and larger ice sheets. The variations in ¹⁸O/¹⁶O ratios within foraminifera in deep-sea sediments, which have been continuously deposited throughout Quaternary times, show a complex pattern of climate change. This pattern corresponds well to theoretical cycles (Milankovich cycles) of change in the distribution of incoming solar radiation across the Earth's surface. This is the result of variations in the orientation of Earth's axis and its orbit around the Sun. The analysis of oxygen isotopes therefore shows that climate change was much more complex than had previously been thought. This stimulated the need to look for detailed proxy data for climate change on land surfaces. Unfortunately, subaerial erosion has made the continental sedimentary record very incomplete and discontinuous. The most complete terrestrial sedimentary record is provided by thick loess sequences. Much attention has therefore been given to trying to reconstruct palaeoclimates from loess.

Loess is a wind-deposited sediment made up of angular quartz and feldspar grains with diameters of between 20 and 60 microns (silt size). This sediment is essentially unlaminated; structurally it is held together by the point-to-point contacts between silt grains, although clays may form bridges between the silt grains or carbonate cements may provide support. Loess is loose and friable and has a high porosity, commonly more than half its bulk volume. It is capable of supporting substantial structures, but when wet the contacts between grains break and the loess may collapse or flow. In some deposits described as loess, fine sand and clay fractions may be more dominant and careful examination often reveals laminations. A precise definition of loess is therefore difficult.

Two models of evolution exist for the origin of the loessic silt. The first suggests that silt was formed by the grinding of rock beneath ice sheets during glacial times. The silts were washed out of the glaciers and deposited on outwash plains, where they were deflated by wind action and carried away from the margins of the ice and subsequently redeposited. The loess deposits of Europe and North America are explained in this fashion. The second school of thought advocates a desert origin for loess. In this hypothesis, the silts form by aeolian abrasion and attrition of sand grains and by rock-weathering processes. Winds blow the silts from the deserts and deposit them in adjacent regions. The loess of central China is explained in this way. Substantial quantities of silt can also be produced by other processes, such as fluvial abrasion and attrition, rock-fall processes, and freeze–thaw (cryogenic) weathering. In support of the two main mechanisms loess deposits have finer grain size away from their glacial or desert sources, for example, the Chinese loess fines westwards away from its source, the Gobi Desert. Characteristic heavy-mineral assemblages also enable the source of the silt to be identified. There is much debate over the exact mechanism of silt deposition, but in the main it can be regarded as a general fallout process.

The critical point, however, is that the processes of formation, transportation, and deposition of silt are strongly controlled by climatic conditions. During cold glacial times, desert regions become more arid and the production of silt increases and its transportation is enhanced. In glaciated areas glacial grinding increases and wind systems are intensified, producing more silt and transporting the sediment further away. Adjacent to both regions the deposition of loess increases. The grain-size characteristics and mineralogy of the loess should reflect climatic conditions. It is generally considered that the deposition of loess has been an almost continuous process for about 2.5 Ma (million years), but the rate of deposition has been highly variable, reflecting past climate conditions. Subsequent erosion of loess, however, makes a continuous record of deposition rare, and the stratigraphy is often difficult to interpret. In addition, during interglacial times soil-forming processes dominate and silt deposition is greatly reduced. As a result, many loess successions have distinct horizons representing ancient soils (palaeosols). The structure and mineralogy of these provides particularly valuable information about past climatic conditions. Loess and palaeosols have been particularly helpful in reconstructing the fluctuations of the desert margins and the intensity of past monsoons.

The thickest loess occurs in central China on the Loess Plateau, where it reaches a thickness of about 330 m near Lanzhou. Very little reworked and continuous deposition appears to have been the rule on the Loess Plateau. The Chinese loess sequences therefore provide the most complete terrestrial record of climate change for the Quaternary. The loess overlies sedimentary rocks of Late Pliocene age, the ‘Red Clay Formation’, and palaeomagnetic work has established an age for the basal contact of the loess of 2.48 Ma. It may be no coincidence that the onset of loess deposition is approximately contemporaneous with the start of the Quaternary Ice Age. This may be because the build-up of ice sheets and the intense rapid global climate change may have resulted in loess-forming processes becoming more dominant. It is also argued that the uplift of the Tibetan plateau during late Tertiary may have lead to increased aridity in central Asia and the onset of loess deposition. Most geologists however, believe that the Tibetan plateau attained its height about 14 Ma ago and that the loess cannot be attributed to tectonics (see Himalayan–Tibetan uplift and global climate change).

Up to 32 identifiable palaeosols (S1 to S32, S1 being the youngest) alternating with loess units (L1 to L33, L1 being the youngest) have been identified in the Chinese loess (Fig. 1a). These are thought to indicate cold dry periods, with high rates of loess deposition, and warm wet periods, with lower rates of deposition and the formation of soils.

At the time of deposition, magnetic minerals within the loess align themselves parallel to the Earth's magnetic field. The Earth's magnetic field fluctuates and reverses periodically. The nature and timing of these changes are reasonably well known, and the measurement of the orientation of these magnetic minerals within the loess therefore enable to loess to be dated. Such palaeomagnetic dating studies have been undertaken on several key sections (see Fig. 1b). Luminescence dating techniques, which measure the remnant radiation left in mineral grains after they were exposed to light, has provided an additional calibration of the loess successions. Accurate dating is critical to apply multi-proxy measures of climate to the loess–palaeosol successions. Many techniques have been applied, including particle size analysis, magnetic susceptibility, sediment fabric, micromorphology, mineralogy, palaeontology (mainly molluscan fauna, but also vertebrates and flora), carbonate content, organic carbon, and extractable iron.

Magnetic susceptibility (MS) has been used to detect palaeoclimatic variations, to correlate palaeosols and to correlate between the oxygen isotope record. MS is generally higher in palaeosols than in loess. Although the reasons for this variability are not fully understood, it may be attributed to the enrichment of detrital magnetic minerals in soil during interglacials owing to concentration by decalcification and soil compaction processes. It may also be the result of subaerial deposition of ultrafine magnetic minerals from distant sources whose concentrations are diluted during the higher rates of silt deposition associated with cold times. Alternatively, the MS may be the result of the formation in situ of magnetic minerals by soil-forming processes. On the basis of the types of magnetic minerals present within the loess, it has been suggested that the formation in situ of magnetic minerals by soil-forming processes is the most important control of the MS. The MS may therefore be broadly considered to be a function of palaeoprecipitation. One of the most intensive MS studies on loess was undertaken at three key sections on the Loess Plateau (Xifeng, Luochuan: Fig. 1). The sediments at these locations represent a time-span of about 2.5 Ma. The combined magnetic susceptibility results from these sections showed a general agreement with the astronomically tuned oxygen isotope deep-sea chronology in the upper part of the succession, but less agreement prior to 0.5 Ma (Fig. 1). Gross correlations between other sections throughout the Loess Plateau, for example at Lui Jia Po and Baoji near Xian in the humid warm south and at Lanzhou in the semi-arid west. Minor variations are more difficult to correlate, probably because of regional variations in climate and soil-forming processes. Broad correlations have also been made between the MS results and aeolian sediment present in deep-sea cores from the Pacific Ocean. High MS values correspond well with high concentrations of aeolian sediment in deep-sea cores (Fig. 1e, f, and g). This probably indicates glacial times when stronger westerly winds carried sediment from China into the Pacific Ocean. A better understanding of the controls of MS will help to refine its use as a detailed proxy measure of climate change.

In the Chinese loess, the median grain size is essentially a measure of the vigour of the north-westerly (winter) monsoon. Coarse median values probably represent cold and dry glacial times. Some loess horizons are particularly sandy, such as the L9 and L15 sandy loess units (Fig. 1a), and these are thought to represent extensive advances of the desert margins. The median grain sizes match closely with the MS, but recent results show a much more complex pattern than that obtained from the MS analysis. This technique has great potential for detailed interpretation of past climate.

Study of the microscopic structures (the micromorphology), the clay mineralogy, organic carbon, and faunal excrements within palaeosols has helped to determine the nature of soil-forming processes across the plateau and between palaeosols of different ages. The abundant molluscan fauna within the loess is also being used to help provide information on past humidity and temperature. Early results show that the abundance of mollusca closely parallels the MS, further supporting the idea that these were times of thermal and humidity maxima.

Loess successions have also been studied in detail elsewhere in the world, but the record is less complete and more complex than in the thick loess of China. In the ‘Palouse’ region of Eastern Washington, USA, for example, the loess sequence reaches 75 m in thickness and contains 19 or more palaeosols. They possibly record 1.5 to 2 Ma of deposition, but the record is discontinuous, with the initiation of cycles of loess deposition controlled by giant glacial floods from the Laurentide (North American) ice sheet during glacial times. The loess sequences in Kashmir and Europe are discontinuous and fragmented, yet quite convincing correlations have been made, mainly on the basis of the MS, with the oxygen isotope deep-sea chronology.

A record of palaeoclimates derived from the interpretation of the loess–palaeosol sequences is becoming greatly improved through higher resolution of the data and a better understanding of the mechanics involved in controlling regional climate change and orbital forcing. Subtle changes in grain size, MS, mineralogy, micromorphology, and other characteristics of loess provide great potential for reconstructing past climate change on the continents.

Lewis A. Owen

Bibliography

Hovan, S. A.,, Rea, D. K.,, Pisias, N. G.,, and and Shackleton, N. J. (1989) A direct link between the China loess and marine records: aeolian flux to the north Pacific. Nature, 340, 296–8.
Kukla, G.,, Heller, F.,, Lui, X. M.,, Xu, T. C.,, Lui, T. S.,, and and An, Z. S. (1988) Pleistocene climates in China dated by magnetic susceptibility. Geology, 16, 811–14.
Rutter, N. (1992) Presidential address, XIII INQUA Congress 1991: Chinese loess and global change. Quaternary Science Reviews, 11, 275–81.
Xiuming Lui,, Shaw, J.,, Lui,, Tungheng,, Heller, F.,, and Baoyin,, and Yuan (1992) Magnetic mineralogy of Chinese loess and its significance. Geophysical Journal International, 108, 301–8.