Formation Of Loess Deposits Research Articles

Mineralogical and magnetic variations of periglacial loess in SE Tibet reveal mid-Pleistocene expansion of Tibetan glacial activity

The formation and evolution of the cryosphere on the Tibetan Plateau is of great significance in understanding the Earth's carbon and climatic system. Periglacial loess deposits in southeastern Tibet offer a means to constrain this history as they contain critical information on glacial grinding and frost shattering processes in high-altitude mountain regions through time, which yield lithogenic fractions of largely loess silts and sands. Based on combined analyses of lithogenic magnetic properties and mineralogical composition, here we find that increasing high mountain production and supply of fresh detrital components since the mid-Pleistocene climate transition (MPT, 1.2–0.7 Ma) led to a substantial increase of the lithogenic susceptibility and decrease in chemical weathering intensity of periglacial loess in southeastern Tibet. The agreement of these findings with similar results from eolian loess on the northern margins of the Tibetan Plateau suggests a plateau-wide glacier expansion during the MPT. Enhanced glacial erosion and freeze-thaw activities occurred in the high-altitude mountain regions of the plateau during the MPT, thereby providing vast amounts of fresh detritus for the formation of loess deposits. High and constant values of lithogenic magnetic parameters in the loess deposits after 0.7–0.6 Ma further indicate that the Middle Pleistocene is a critical period for the establishment of modern-like glacial and periglacial landforms on the Tibetan Plateau. Our results further suggest that enhanced Quaternary glaciation in SE Tibet occurred earlier than in the north, which we attribute to climate cooling combined with regional seasonal snowline lowering. As such, a close relationship exists between global climate changes, development of pan-Tibetan glaciations, and large-scale dust emission.

Assessing Weathering, Pedogenesis, and Silt Generation in Granitoid‐Hosted Soils From Contrasting Hydroclimates

AbstractThe origin and production of silt are key factors in the formation of loess deposits. Although many processes can potentially lead to silt generation, few are known to produce silt in the volumes and particle‐size modes required to form geologically significant loess deposits. Here, we investigate the hypothesis that soil‐associated weathering in tropical and Mediterranean climates can generate abundant in situ silt, and therefore contribute significantly to loess formation throughout geologic time. We utilize granulometric and geochemical data from two granitoid‐hosted soils formed in Puerto Rico (hot‐humid climate) and Southern California (hot‐arid climate) to discern whether the silt + clay fraction (<62.5 μm) is generated from bedrock weathering (autochthonous) or sourced from eolian contributions (allochthonous). Our study demonstrates that the PR soil contains abundant (up to 72%) silt‐ and clay‐sized particles compared to the Southern California soil (<6%). However, the silt fraction of the PR soil is at least partially derived from eolian inputs, and the silt fraction of the Southern California soil is geochemically indistinguishable from allochthonous dust sources. Furthermore, while intense chemical weathering in PR can produce abundant fines, the majority are significantly finer (average mode ∼15 μm) than the modes of most “typical” loess deposits (modes more than 20–30 μm). In contrast, weathering in the Southern California soil produces volumetrically sparse silt. Hence, weathering processes within these two profiles, representing hot‐humid and hot‐dry climates, appear to be ineffectual for producing the volume and size distributions of silt needed to generate significant loess deposits.

Work on The Principles of Geology is “interrupted,” and Charles Lyell investigates the nature and formation of loess deposits

Charles Lyell's life work, the influential textbook and philosophical treatise The Principles of Geology, helped to shape the Victorian age. There were few diversions from the central task of the Principles of Geology in the course of the twelve editions. Perhaps the most significant and notable was the discovery and investigation of loess; a silty sediment, often of considerable thickness, and widely distributed in the Rhine valley. He made a significant contribution to loess studies and, with his father-in-law, Leonard Horner, explored the loess regions of the Rhine Valley. The loess period lasted roughly from 1830 to 1836, and he established the first paradigm theory for the formation of loess deposits from deposition in a lake or perhaps from slow-moving water. This held sway until displaced by the aeolian theory of F. von Richthofen.

Origins and genesis of loess deposits in central Spain, as indicated by heavy mineral compositions and grain‐size variability

AbstractThe formation of loess deposits strongly depends on the availability of deflatable dust‐sized material. Identification of source areas and mechanisms of dust production is essential for an appropriate characterization of related palaeoenvironmental conditions. So far, little research has been done on loess deposits in the Mediterranean region and information about mechanisms of dust production is very rare. In this context, it is not clear from where the loess deposits from the upper Tagus Basin in central Spain originated. The main objective of this study was to find out whether these loess deposits were formed primarily by mountain processes and fluvial comminution, or whether they have their origin in the weathering of Tertiary marls in the centre of the Madrid Basin. A further concern was to link the determined source areas with loess formation mechanisms and concomitant environmental contexts. Following a comprehensive approach using heavy minerals and grain‐size data, together with information on stratigraphic features and geomorphic positions, different local loess sources could be identified. This study shows that during the last 35 kyr a major proportion of the Tagus loess deposits was deflated from river floodplains, while the mountain region of the Iberian Range was identified as a significant source of respective floodplain sediments. Based on heavy mineral compositions and dating results, it was found that during Heinrich Stadial 3, sediment supply from the Iberian Range strongly increased, suggesting an environmental shift from initially warmer towards colder temperatures in late Marine Isotope Stage 3. Furthermore, it was found that grain‐size patterns clearly indicate maximum wind strengths during Heinrich Stadial 3, followed by Heinrich Stadial 2 and Heinrich Stadial 1. These results demonstrate that the formation of dust‐sized particles in this part of the Mediterranean operated similar to many other places in temperate and continental regions, and that weathering processes linked to hot and dry environments were only of minor importance.

New evidence for the provenance and formation of loess deposits in the Ili River Basin, Arid Central Asia

Abstract Loess deposits are thick and widespread along the piedmonts of Arid Central Asia (ACA), however the source (provenance) and processes of formation of these fine-grained aeolian deposits are poorly understood. Here we investigate the provenance and possible distribution mechanisms for loess along the slopes of the Ili River basin, located in northwest China and southeastern Kazakhstan, using a grain-size mixture model and an elemental geochemistry-based source fingerprinting technique. Our results indicate that the Ili loess experiences low rates of sedimentary recycling downstream within the basin piedmont, and are strongly dependent on local geomorphic context. Loess deposits are dominated by proximal sources, indicating short-distance aeolian transport from the Ili River alluvial plains and local proluvial fans. Local sourcing dominated regardless of location within the catchment, although the proportion of fluvial input increases proportionally with increasing distance downstream. Our results suggest that the Central Asian deserts did not act as significant interim storage reservoirs for the loess deposits in the Ili River basin, which contrasts with the popular model for piedmont loess formation across Central Asia. Most likely the relatively enclosed and highly variable basin topography precluded transport from the open desert steppe into the upper Ili River valley. Our study provides the first clear evidence for a genetic link between the Asian high mountains and the loess of the adjacent piedmonts, based on geochemical and grain-size data, with the caveat that the high degree of topographic variability along the Tianshan piedmont likely results in a strongly localized influence on loess formation and accumulation.

The formation of loess ground by the process of loessification: a history of the concept

Abstract Loessification is a process by which a body of non-loess ground is transformed into a body of loess ground. The history of loessification is one of controversy and confrontation, largely because of mutual misunderstandings between geologists and pedologists. Lev S. Berg is the ‘only begetter’ of the theory, first proposed in 1916, and propagated throughout his life. R.J. Russell proposed the same approach to the loess in the Lower Mississippi valley in his famous 1944 paper, which contributed enormously to the study of loess in North America. As understanding of the various processes involved in the formation of loess deposits has developed, a compromise position on loess formation has become possible. The major intrinsic features of loess deposits are the open structure and the collapsibility. It appears that the open structure is caused by aeolian depositional processes and the collapsibility is caused by loessification processes. The compromise was initiated by Marton Pécsi in 1990, He endeavoured to retain a loessification aspect in the study of loess deposits, as the subject appeared to be overwhelmed by the aeolian idea system promoted by geologists; it has been mostly a Central European endeavour. The history of the concept of loessification largely involves (1) its development in Russia, (2) its dissemination and discussion – and attempts at refutation and modification – in the wider world.

Controls on the nature of loess particles and the formation of loess deposits

The story (history) of a single loess particle can be told from the formation of the planet to the making of bricks. The modal loess particle chosen for study is a 30 μm quartz particle. Many factors bear on its complex history, starting with the formation of quartz units in the original granitic crustal rocks. Quartz units appear in the old granites as the product of a eutectic-like reaction. This controls particle size and ensures that sand is much the same wherever (and whenever) it is encountered. The eutectic reaction controls sand size; the silt size is controlled by crystalline defects which are introduced into the quartz by the high-low displacive transformation which the quartz undergoes as the system cools. Thus all quartz silt is much the same, as is the case for sand. The quartz clastic particles are surprisingly bi-modal, a more even size distribution might have been expected. The silt particles are largely formed by glacial grinding, by continental and mountain glaciers. These silt particles are distributed across the landscape by the action of large rivers, and deposited on floodplains. A critical stage in the story is the aeolian transportation of the particles which then form an airfall deposit. The glacially provided silt particles are quite close in size to the optimum for aeolian pick-up (∼80 μm). At this stage, on wind-fall deposition, the characteristic metastability is introduced into the loess deposit. The chief post-depositional process is the development of collapsibility, which is largely due to the movement of clay minerals to the main particle contacts. This low clay content of most loess deposits makes loess an ideal material for brick production.

Four loess pioneers: Charles Lyell, F. von Richthofen, V.A. Obruchev, L.S. Berg

The four loess pioneers were Charles Lyell, Ferdinand von Richthofen, Vladimir Obruchev and Lev Berg. Their life spans covered the period 1797–1956. They each contributed significant thoughts and ideas to the development of the study of loess deposits. Their loessic ideas can be related to their life experiences and to their own particular approaches to their chosen branches of science. Their ideas can also be set in a framework of scientific development at a time when scientific knowledge was increasing rapidly and new fields and areas of study were being demarcated. The directions that they indicated and the ideas that they implanted still influence loess research today, and some problems that they touched on remain unresolved. The basic Lyell idea of lacustrine or fluvial deposition followed by uplift held sway from about 1830 till around 1880. Richthofen developed his ideas on aeolian deposition from 1870 and Obruchev developed his ideas after the Potanin expedition, in 1895. Berg offered his concept of loess formation by weathering and soil formation in 1916, and was a keen proponent until his death in 1950. Many scholars offered ideas and proposals on the problem of loess deposit formation but the four chosen individuals appear to predominate, and have delivered the paradigms that have shaped loess research since the 1830s. Richthofen has probably acquired a little too much credit for the aeolian theory, and Obruchev not enough. Obruchev delivered a fully worked out aeolian approach which encompassed post-depositional changes to the loess material. Berg, with his total concentration on post depositional changes was only considering half a picture. Some of his ideas were important but his one eyed view meant that many scholars neglected them.

Geochemical composition of Tajikistan loess and its provenance implications

The Tajikistan (TJ) loess has great significance for the study of Quaternary climate changes and dust source in Central Asia. In this paper, we investigate the geochemical composition and provenance of loess. Our results show that the major elements, in order from high to low average content, are SiO2 > Al2O3 > CaO > Fe2O3 > MgO > K2O > Na2O > TiO2 > P2O5 > MnO. Fe, Ti, P and Mn contents in loess are similar to those in the UCC, as is the case with Si, Al and Mg contents in paleosols. Ca is distinctly enriched in loess, and Si, Al and K are depleted. Na is depleted in both loess and paleosols, and P is depleted only in paleosols. In contrast to the UCC, paleosols show enrichments in all trace elements except for Ba, Sr and Zr, but in loess, some trace elements (e.g., Ba, Co, Ni and Zr) show depletion. The CIA value of TJ loess is lower than that of Kashmir loess, the latter being influenced by differences in the sorting and tectonic environment. There are no great changes about provenance during the formation of TJ loess, and the deposit areas have a shorter distance to the provenance. Geochemical data and atmospheric circulations demonstrate that the TJ loess was mainly transported by the near-surface northerly or northwesterly winds, but there may be an additional and local source making a certain contribution to the formation of loess deposits in the south of Tianshan.

Landscape development and soils around the Bronze Age settlement Kamennyi Ambar, southern Trans-Urals, Russia

Landscape and soils around the prehistoric Bronze Age site of Kamennyi Ambar, southern Trans-Urals, have been studied. The topography of the Trans-Ural peneplain was formed by denudation during Mesozoic and Tertiary. The resulting gently undulating plain was subsequently dissected during Pleistocene under periglacial climatic conditions, and wide valleys with gentle slopes were formed. The saprolized pre-quaternary rocks are covered by thin layers of quaternary sediments, which are sometimes loess-containing. Deposits of in-situ loess are almost absent. Most probably, deluviation under periglacial conditions played the dominant role and prevented the formation of loess deposits. Thin loess deposits are preserved only in some flat watershed positions. The Quaternary layers sometimes contain strongly weathered coarse bedrock components indicating a solifluidal genesis of the cover beds on the slopes. The degree of weathering of the autochthonous components increases towards the water divide where the cover beds consist almost completely of pre-Quaternary saprolite. The slopes of the Karagaily Ajat valley are structured by numerous smaller tributary valleys which usually have a yellow-brown loam fill and show in their lower parts gully-like linear erosion features. Decalcified Haplic Chernozems are prevalent soils in the fills of the tributary valleys, at the lower slopes of the Karagaily Ajat-valley, and in the loams of the Lower Terrace. When clay cutans and coatings were observed in the subsoil the soils were identified as Luvic Chernozems. On the gentle slopes, Mollic Leptosols or Leptic Regosols (Mollic) are developed in thin cover beds. Up the slope, the depth of the humus horizons decreases. When examining the layer succession of the near surface underground and the associated soils along transects across the tributary valleys, it was found that the thickness of the cover beds on the slopes increases towards the valley floor. In the soils of the Lower Terrace varying salt contents can be found. Mollic Solonetz were identified therein. Within the excavation area in Kamennyi Ambar, Anthrosols and Solonetz were found in the compacted layers of the older Sintashta culture, while Haplic Chernozems (colluvic) with similar features like the virgin soils have developed in the fillings of younger pit houses.

Loess and Bee-eaters IV: Distribution of the rainbowbird (Merops ornatus Latham 1801) in Australia

The first three papers in the Loess and Bee-eaters series sought to establish links between the nesting of bee-eater birds (family Meropidae) and the occurrence of loess deposits. For the European bee-eater there is a close and fairly obvious relationship; for the Carmine bee-eater nesting in the 15 N band of Africa more assumptions and adjustments have to be made, and for the Blue-cheeked bee-eater nesting in the Indus region the relationship is becoming vague and more speculative. The trend continues with the Australian bee-eater- the rainbowbird (Merops ornatus). We lack precise maps of loess distribution in Australia and there is an equal lack of detailed and accurate maps of the distribution of rainbowbird nesting.Loess in Australia is elusive; the maps of Kriger and Scheidig show little detail and are essentially contradictory. If the deterministic theory of loess deposit formation were applied, it would suggest that loess should be found in the region where Pleistocene cold-phase mountain glaciers occurred. In the last glacial phase there was very little glacial occurrence in Australia; probably a limited region near Mount Kosciuszko. This places some glacial activity in the extreme southeast of the country. This agrees, more or less, with the data on the Kriger map.Fry has mapped the breeding zones of the rainbowbird and they are concentrated in the southeast and southwest. The maps are not precise but there does seem to be indications of co-existence of bee-eater nests and loessic regions. This is not the clearcut relationship which has been observed with the European bee-eater but it is suggestive. The ill-defined loess/dust regions and the poorly demarcated nesting zones of the rainbowbird do seem to be coincident. Better mapping is urgently required for loess in Australia, and nesting zones of M. ornatus. It may be that dust rather than loess is the ground material to be observed in Australia. There are still problems of integrating the dust and loess concepts, but mapping of dust deposits does produce some coincidence with rainbowbird nesting. The region of Vertisol occurrence appears to be a barrier to the spread of rainbowbird nesting.

Loess and bee-eaters III: Birds and ground in the Punjab and the Indus region

Part 3 of the study of loess ground and bee-eater birds concerns birds and ground in the north-western part of the Indian Sub-continent. Three species of bee-eaters are considered, relative to the Indus region: the Green bee-eater, the Blue-cheeked bee-eater and the European bee-eater. Loess in the Indus region is considered via the deterministic approach to loess deposit formation; the P events occur in the mountains of the western part of High Asia, particles are formed by the action of mountain glaciers. Major T actions involve the five rivers of the region which carry loess material into the Punjab plains. Subsequent T actions deliver loess material to local deposits, which provide nesting grounds for bee-eaters. To some extent the bee-eaters define the loess. The Green bee-eater is small (16–18 cm long) and lives over all of India and Pakistan-nesting ground is available everywhere. The larger bee-eaters (European and Blue-cheeked: (25 cm)) are more constrained by ground properties-they require the ideal properties of loess ground (an ideal positioning relative to the Heneberg compromise) for nesting, and thus tend to define loess extent. Soil mechanics requires that more exact stress conditions are achieved when a relatively large nest tunnel is being constructed.

John Hardcastle looks at the Timaru loess: Climatic signals are observed, and fragipans

John Hardcastle (1847–1927) made several important loess-related observations in New Zealand in 1889 and 1890. He is best known for his ‘invention’ of loess stratigraphy, his observations of loess as a ‘climate register’, but he made significant contributions in other areas. He outlined the event-sequence approach to loess deposit formation in 1889, well before the better known exposition in 1966. The basic sequence for loess deposit formation P1T1D1T2D2 could well be named the Hardcastle sequence. He pointed out the problem of silt particle production as an important question, emphasizing the role of glacial action, and he drew attention to the occurrence of bird crop stones in loess deposits, as possible markers for deposition stages. Moa stones in particular may give useful information about New Zealand deposits. The Timaru loess, investigated by Hardcastle, has been proposed by J.D.Raeside as the type section for New Zealand loess. JH was probably the first person to describe fragipans (characteristic hard layers) in loess. The fragipan was not properly defined and named until the Guy Smith proposals c.1948, but Hardcastle offered detailed descriptions in 1890.As the site of important loess investigation, and possessing impressive stratigraphy, the Timaru loess could be proposed as an international loess section for New Zealand. Hardcastle's stratigraphic observations were probably among the first relatively sophisticated palaeoclimatological observations made. The INQUA Loess Focus Group could recognize the Timaru Loess (the Dashing Rocks section) as the representative loess section for New Zealand- a reference section for studies of loess stratigraphy and chronology (hopefully preserved and curated by the local government authorities and other interested parties).

Magnesium isotopic variations in loess: Origins and implications

Loess deposits serve as important continental archives for studying Quaternary climatic variations and for estimating the average chemical composition of the upper continental crust. Here, we report high-precision Mg isotopic data for 19 loess samples from China, Argentina and Europe, which were previously used to estimate the composition of the upper continental crust. The results show that global loess samples have heterogeneous Mg isotopic compositions, with δ26Mg (per mil deviation of the 26Mg/24Mg ratio from the DSM3 standard) values ranging from −1.64‰ to +0.25‰ and a weighted average of −0.89‰, which is lighter than both crust and mantle silicates. MgO content and δ26Mg of loess positively correlate with CaO/Al2O3 ratio, suggesting a two-component mixing between carbonates and secondary silicate minerals. The large variation in Mg isotopic composition of loess results from a combination of factors, including source heterogeneity, eolian sorting during transport of the loess and chemical weathering during formation of the loess deposit. Our results suggest that Mg isotopic composition of loess has potential to be a proxy indicator to characterize the paleoclimatic change, but may not represent the average Mg isotopic composition of the upper continental crust due to mixing, sorting of isotopically distinct components and isotope fractionation during loess deposit formation.

Sand martins favour loess: How the properties of loess ground facilitate the nesting of sand martins/bank swallows/uferschwalben (Riparia riparia Linnaeus 1758)

Sand martins/bank swallows (Riparia riparia Linnaeus 1758) nest in tunnels excavated in loess deposits. Loess is particularly suitable because of its cohesive properties, its ease of excavation, its ability to form vertical faces, its ready availability in small scale brick works. A low clay content is a requirement for suitable ground (palaeosols appear to have too much clay), and in sand deposits a low level of cementation. The selective nature of nesting behaviour may have some stratigraphic value, and it may be that loess availability affects the distribution of sand martins (they move away from river banks to brick pits). Soil crust formation may affect the inhabitability of nest tunnels (the same phenomenon may be important in loess deposit formation). Penetration resistance of the ground is required to be in a middle-range, not too hard, not too soft. This is the Heneberg compromise: a compromise between tunnel stability and ease of excavation and construction.

Loess is [almost totally formed by] the accumulation of dust

In 1990 Marton Pecsi published a paper entitled ‘Loess is not just the accumulation of dust’. This 1990 paper presented the case for considering aspects of loess formation other than the aeolian deposition of silty material. Pecsi argued that there is more to loess formation than simply the fact of aeolian deposition. While not disputing the validity of much of Pecsi’s arguments a case can be made for considering that aeolian deposition is at the very heart of loess formation, that the factor of aeolian deposition is so critical that other aspects are relatively unimportant. Pecsi appeared to suggest that aeolian deposition is just one of many factors relating to the formation of loess deposits; this is not so aeolian deposition and the consequent formation of metastable structures is absolutely central. But, it is probably useful to consider the ‘aeolian deposition is absolutely critical’ approach and the ‘aeolian deposition is one of the factors’ approach in some more detail, and in relation to each other. Some insights may accrue; interesting things do happen after the deposition of the dust and this needs to be acknowledged.

Rivers and loess: The significance of long river transportation in the complex event-sequence approach to loess deposit formation

Rivers are essential for loess deposit formation. River systems which relate directly to loess deposit formation include the Danube and the Rhine; the Mississippi and the Missouri; the Thames and the Medway; the Indus, Ganges and Brahmaputra; the Dnepr, Don and Volga; the Clutha, Mataura and Rakaia—New Zealand rivers; the Yellow River; rivers in South America; Siberian rivers (flowing north); the Odra and the Wisla (between the mountains and the ice); the Syr-Darya and Amu-Darya (into the desert). The contention is that rivers are not just important in major loess deposit formation, they are mandatory (necessary, imperative, obligatory, essential, indispensable, requisite). For a complete and satisfactory study of a loess deposit, we need to know how the material is produced; how it is transported and distributed across the landscape; how intermediate deposits are formed; how transportation provides the defining properties (i.e., open structure, collapsibility, draping across the landscape); and what may happen post-deposition (e.g., chernozemisation, fragipan formation, increased collapsibility, etc.). If rivers are essential then it would appear that the British loess is probably Alpine material delivered by the proto-Rhine and the Polish loess is not derived from the ice-sheet to the north but from the mountains to the south. The lack of loess in Canada is explained by the absence of suitable rivers. The loess in Russia and neighbouring countries can be classified (as Jefferson et al. [Jefferson, I.F., Evstatiev, D., Karastenev, D., Mavlyanova, N.G., Smalley, I.J., 2003b. The engineering geology of loess and loess-like deposits: a commentary on the Russian literature. Engineering Geology 68, 333–351] suggested) on the basis of associated rivers. A ‘new’ deposit of glacial loess is recognised—associated with the Dnepr, Don and Volga rivers (the USWR loess). The Smalley–Leach [Smalley, I.J., Leach, J.A., 1978. The origin and distribution of the loess in the Danube basin and associated regions of East-Central Europe: a review. Sedimentary Geology 21, 1–26. Available from: 〈 www.geo.edu.ro/sgr/mod/downloads/PDF/Smalley-SedGeo-1978.pdf〉] vision of the loess in the Danube basin and East-Central Europe is probably misconceived. They assigned a great significance to glacial action and put loess in a ‘northern band’. It appears that this proposed glacial influence is minimal and that the Danube loess is essentially mountain loess—derived from the Alps, the Carpathians, the Sudeten mountains and other high regions. The Danube system is a classic example of a river controlling loess distribution.

The formation of loess deposits in the Tashkent region and parts of Central Asia; and problems with irrigation, hydrocollapse and soil erosion

There are large problems of irrigation, hydrocollapse and soil erosion associated with the loess deposits of Central Asia. To get to grips with these we need a clear view of the nature and formation of the loess, and in this region there are still disputes and disagreements and controversies. Important events in the loess deposit formation process can be identified and studied, and relative significances assigned. There are large loess deposits in the vicinity of Tashkent, in the north-east of Uzbekistan. These were mapped and studied for many years by G.A. Mavlyanov and associates; they used a terminology which appears to be based on Pavlov's nineteenth century usage and developed a ‘polygenetic’ formation theory for loess deposits. This was a formation theory very much associated with Central Asia and essentially favouring a ‘proluvial’ mechanism. A speculative look at the deposits near Tashkent allows alternative terminologies and formation mechanisms to be considered. In particular the widespread Soviet antipathy to aeolian deposition can be examined. The deposits associated with the Chirchik river, upstream of Tashkent, might be affected by the Pavlov mechanisms, but could be topped by aeolian deposits. The collapse of loess when wetted is controlled by the nature of the ground and is a very significant interaction of soil and water, in particular in Central Asia. Hydrocollapse and water erosion are closely related and are controlled by the complex nature of loess ground. Dust clouds in eastern Uzbekistan are largely loess related, but in the west dust material is raised from the drying Aral Sea bed; this is often clay mineral agglomerate material and can carry dangerous pollutants.

An approach to the problem of loess deposit formation: Some comments on the ‘in situ’ or ‘soil-eluvial’ hypothesis

The problem of loess formation has been widely discussed for a long time and in many languages. At the beginning of the 21st century some parts of this debate are still going on. A case is still being made for the ‘in situ’ approach, and there are problems of terminology and mechanism for the Central Asian loess. Consideration of the origin of loess material has now replaced the discussion on deposit formation as the major ‘intrinsic’ problem, and in the world of ground engineering there is still much investigation on the origins of collapsibility, which is, in effect, a discussion of loess origins. There remain some interesting and useful observations to be made on the North African loess and the loess-like materials in Australia, and the ‘desert’ loess controversy is still running. Connections of loess to chernozems, fragipans and rendzinas are becoming clearer, and it seems that it may be possible to appreciate a loess landscape in which the loess contribution is not immediately apparent. Recent SEM observations suggest that collapsibility develops after metastability—as the ‘in situ’ theory appears to imply.

Provenance of loess material and formation of loess deposits on the Chinese Loess Plateau

Results of a multiple, isotopic, chemical and mineralogical analysis of loess from the three northwestern inland basins (the Junggar Basin, the Tarim Basin and the Qaidam Basin) and the Loess Plateau region of China are summarized. They suggest a qualification of the conventional views that the three northwestern basins were the important source areas of the Loess Plateau and that the sand deserts were the primary single source of the Chinese loess. It is argued that the gobi (stony desert) in southern Mongolia and the adjoining gobi and sand deserts (the Badain Jaran Desert, Tengger Desert, Ulan Buh Desert, Hobq Desert and Mu Us Desert) in China, rather than the three inland basins, are the dominant source areas of the Loess Plateau. However, although these gobi and sand deserts are regarded as the main source regions, they serve as dust and silt holding areas rather than dominant producers. The mountain processes (including glacial grinding, frost weathering, salt weathering, tectonic processes, and some fluvial comminution) in the Gobi Altay Mts., Hangayn Mts. and the Qilian Mts. have played an important role in producing the vast amounts of loess-sized material for forming the Loess Plateau.