Formation and evolution of sand deserts in Xinjiang ... - Springer Link

J. Geogr. Sci. 2014, 24(1): 177-190 DOI: 10.1007/s11442-014-1080-8 © 2014

Science Press

Springer-Verlag

Formation and evolution of sand deserts in Xinjiang, Northwest China: I. Provenances of desert sands ZHU Bingqi1, YU Jingjie1, QIN Xiaoguang2, Patrick RIOUAL2, LIU Ziting 2, XIONG Heigang3 1. Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, CAS, Beijing 100101, China; 2. Key Laboratory of Cenozoic Geology and Environment, Institute of Geology and Geophysics, CAS, Beijing 100029, China; 3. Key Laboratory of Xinjiang Oasis Ecology, Ministry of Education, Urumqi 830046, China

Abstract: Sandy desert landscape is a geological product of arid climate and abundant sand materials supply in natural conditions, therefore the formation of sandy desert is an ideal studying object for understanding the interaction between various stratigraphic/epigenetic spheres of the earth system. However, until now, the knowledge about the provenance and formation of these deserts in Xinjiang, Central Asia is diverse and inconsistent, a systematic review is necessary. Ancient aeolian sand sediments in the internal areas of these deserts and their twin dust deposits in the Cenozoic strata surrounding the Xinjiang deserts are direct and indirect clues reflecting the provenance of sand materials and the formation and evolution of deserts. Based on the geochemical, mineralogical and isotopic evidences of desert sands and relevant deposits in the peripheral regions, this paper reviews the research progress on the development of the sandy deserts in Xinjiang, Northwest China. Many proofs proposed that desert sands in Xinjiang were mainly sourced from the ancient pluvial, alluvial and fluvial sediments and were eventually achieved because of the local sand supply. It pointed out that the settings of tectonic structure in Xinjiang had made sediments both in the Tarim Basin and the Junggar Basin being influenced greatly by regional hydrological system and aeolian processes originated from the planetary wind system of westerly, the East Asian winter monsoon and the topographical mountain-valley winds. However, the directions of transportation paths are different between the hydrological dynamics and the aeolian dynamics, which are decided by the slantwise structures of the basins tectonics and the regional atmospheric circulation routes, respectively, and have determined anisotropies of the transport pathway of these sediments delivered by different agents. This results in a geological cycle of detrital sediment transportation in the two large basins and thus largely increases the degree of sediment mixing, alternation and recycle between younger and older sediment sources, as Received: 2013-05-27 Accepted: 2013-07-05 Foundation: National Basic Research Program of China (973 Program), No.2009CB421305; National Natural Science Foundation of China, No.91025023; No.41371060; No.41271049 Author: Zhu Bingqi (1976–), PhD, specialized in land surface processes in arid zones and the Quaternary environmental change. E-mail: [email protected]

www.geogsci.com

www.springerlink.com/content/1009-637x

178

Journal of Geographical Sciences

well as the complexity and diversity of sand provenances. Keywords: sandy desert; provenance; hydrodynamic and aeoliandynamic agents; sediment transportation; Xinjiang

1

Introduction

Sandy deserts in China are mainly distributed in its northern part where the annual mean precipitation is less than 400 mm (Figure 1); they have three distinguishing features: deep in the hinterlands of the middle-latitude intermontane basins and plateaus, spanning multiple bio-climate zones, and having the largest areas in Xinjiang (Wang et al., 2007). Geographically, there are eight major sandy seas and four sandy lands, among them, three of the sandy deserts are located in Xinjiang, namely, the Taklamakan, the Guerbantonggute and the Kumutage sandy deserts (Figure 1). Previous studies on Quaternary geography, palaeoclimotology, morphodynamics of dunes and aeolian physics of sandy deserts have proofed that sandy deserts are formed and developed under natural conditions with long-term arid climate and abundant sand materials (Zhu et al., 1980; Yang et al., 2011). The formations and evolutions of these deserts are directly controlled by three major factors, i.e., sand sources, climate (such as wind regime, aridity, depths of the atmospheric boundaries and the stability of atmospheric strata), and the underlying surface conditions (such as ground surface vegetation cover ratio and topography) (Yang et al., 2008; Yang et al., 2012). So understanding the formation and evolution of desert landscapes are significant for recognizing the interaction rule between various geospheres of the Earth surface.

Figure 1

Distribution of sandy deserts and desert lands in northern China (revised after Wang et al., 2008)

Sandy deserts in Xinjiang are located at the northern edge of the Qinghai-Tibet Plateau. The uplift of the Qinghai-Tibet Plateau and the formation of arid climate in the hinterland of the Asian Continent since the late Cenozoic are regarded as an important token of the deterioration of global climate in Cenozoic period (Guo et al., 2002). The special geographical positions of Xinjiang deserts make themselves a good record of the both events about the uplift of the Tibetan Plateau and the inland aridification of the Asian Continent (Sun and Liu,

ZHU Bingqi et al.: Formation and evolution of sand deserts in Xinjiang, Northwest China (I)

179

2006; Sun et al., 2008, 2009). Hence, the studies on the formation and evolution of Xinjiang deserts are involved in the questions about the rises of high plateau and its environmental effects and the drought mechanism of the west land in China, therefore, arousing wide attention and concern. Until now, long-term research works have been carried out and consequently many western and oriental literatures have been obtained that recorded the knowledge about the formation of the desert landscapes in Xinjiang. However, these works and data are scattered with various perspectives inconsistent with each other on this issues. It is necessary to make a systematic summary on these materials. In this paper, based on previous literatures, we try to sum up and induce the previous opinions about the formation and evolution of sandy deserts in Xinjiang from palaeoenvironmental perspective.

2

Distribution of sandy deserts in Xinjiang and their natural settings

Xinjiang is located in the hinterland of Asia and of Northwest China with distances to sea on all sides exceeding 2000 km, namely, a great sea-land distance. The climate is controlled all the year round by the westerlies but is strongly influenced by the Mongolia-Siberian high pressure belt in winter and is evidently characterized by a continental arid climate. The southern and eastern parts of Xinjiang are warm temperate in climate and the most part of northern Xinjiang is temperate. Some 22% of the area is occupied by sandy deserts, gobi deserts and mountain deserts with perennial snow and glaciers. Sandy deserts are mainly distributed in the center of the two big basins, the Taklamakan Desert in the Tarim Basin (southern Xinjiang) and the Guerbantonggute Desert in the Junggar Basin (northern Xinjiang), known as the largest and the second largest sandy deserts in China respectively (Figure 1). The third sandy desert, the Kumutage Desert, is situated in eastern Xinjiang (Figure 1). Besides, there are some sporadic and scattered dune fields distributed in the Tulufan Basin, Yanji Basin, Yili River Valley, and so on. The Taklamakan Sand Desert in the central Tarim Basin is the second largest (33.7×104 km2 in area) active sandy desert in the world and can be termed as the “drought pole” of the world (Sun et al., 2009). The Tarim Basin is surrounded from its southern, western and northern parts by high mountains more than 3000 m, e.g., the Kunlun Mountains, the Pamirs and the Tianshan Mountains respectively (Figure 2). The basin terrain is tilted from southwest to northeast and some inland rivers fed by snow-melt water, such as the Tarim, Akesu, Hetian and Keriya rivers, are developed in the basin. Sediments in the Taklamakan Desert are thought mainly derived from basement rocks in the surrounding mountains. The basement rocks are composed of sedimentary rocks (chiefly sandstone, shale and limestone) from Precambrian to Quaternary ages and granitic rocks from Proterozoic to Cretaceous and most of the Precambrian rocks have been matamorphased (Honda and Shimizu, 1998). Due to the great distance eastward from the Pacific Ocean, the East Asian summer monsoon can not enter the Tarim Basin. The Tianshan Mountains in the north hinder the moist air from the Arctic Ocean into the basin, and the huge Qinghai-Tibet Plateau, the Himalayas and the Kunlun Mountains in the south block the warm-moist air flow from the Indian Ocean. The annual precipitation at most part of the basin ranges between 25 and 40 mm, but less than 15 mm at the eastern part. The annual volume of potential evapotranspiration is between 2100 and 3400 mm. The annual mean temperature is 11℃. The landforms in the Tarim Basin are

180


characterized by an annular distribution, as mountainous landscape at the edges and piedmont, diluvial and alluvial fans and gobi deserts inwards, and sand deserts (the Taklamakan) in the center. The dune types in the desert are diverse and complex (Figure 2). Oases are usually distributed in the area between sand deserts and gobi deserts, or located along river banks, forming the important agricultural districts. The active dunes are widely distributed in the Taklamakan Desert, occupying about 85% of the desert area (Zhu et al., 1981), and the shifting directions of most dunes in the desert are mainly from northeast to southwest (Li et al., 1999). Although active dunes are a major landscape type in the desert, green vegetation belts are also distributed in the areas along river banks in the hinterland and at the edge of the desert and on the diluvial- and alluvial-fans. Due to water recharge from ephemeral floods, there is abundant fresh water resources in the alluvial strata, forming unique places for human activities in the Taklamakan Desert.

Figure 2 1980)

Geomorphological diagram of the Taklamakan Desert in Northwest China (revised after Zhu et al.,

The Guerbantonggute Desert (4.88×104 km2 in area) lies in the central part of the Junggar Basin (Figure 3). About 97% of the desert is occupied by fixed and semi-fixed dunes, being the biggest fixed and semi-fixed sand desert in China. Some desert researchers have recently defined the Guerbantonggute Desert as a sandy land but not a sand desert (Yang et al., 2012). The Junggar Basin is a large intermontane basin located between the Tianshan, the Junggar boundary mountain and the Altai Mountains. In tectonics, the basin terrain is higher in the east and lower in the west and is generally tilted from southeast to northwest. Three big drainage systems have developed in the basin, i.e., the Yili and the Junggar endorheic watersheds and the Erlqis exterior watershed (Ma, 2002). The surrounding mountains encircling the Junggar Basin are not rigidly closed, leading to a direct entry of moisture from the


181

westerlies into the basin through the west and northwest mountain passes. Compared to the warm and temperate climate in the Tarim Basin, the Junggar Basin is under the temperate climate. Because of the occurrence of the moist westerlies, the landscape of fixed and semi-fixed dunes are predominant in the Guerbantonggute Desert (Zhu et al., 1980; Wu et al., 1997). Unlike the scarce precipitation in the Tarim Basin, the annual mean precipitation in the Junggar Basin reaches to 70–150 mm, mainly the winter-snow precipitation. Vegetation in the desert grows well, with Haloxylon ammodendron, Chionese tamarisk and Euphratica as the mostly distributed plants. The mophodynamics influencing the buildup of sand dune landscapes are mainly the westerly circulation regime and the Mongolia High Pressure (Easterward wind system). While on a large scale to the whole basin, the latter is secondly important, owing to its seasonal and regional limitations that are only significant to the Guerbantonggute Desert during the periods of winter and spring and only at the regions of the central and northeastern parts of the basin. Only the north branch of the westerly circulation regimes prevailed the basin throughout the year, resulted in a movement of dunes from northwest to southeast in the Guerbantonggute Desert.

Figure 3 Geomorphological diagram of the Guerbantonggute Desert in Northwest China (revised after Zhu et al., 1980)

The Kumutage Desert (2.29×104 km2 in area) is situated at the eastern part of Xinjiang and the southern part of the Luobupo Lake and the northern foot of the Arljin Mountains (Figure 4) and is ranked as the sixth largest sand desert in China (Zhu et al., 1980). The strong uplifts of the Qinghai-Tibet Plateau resulted from the neotectonic movements in Cenozoic have formed a basin-mountain coupled style in this area, with a topography that is higher in the south and lower in the north. The regional climate is hyper-arid with the annual mean precipitation less than 30 mm and a low vegetation cover. The sand dunes in the desert are almost all the active dunes, which are partly covered on the slanted slopes of lithoid mountains and partly distributed on the palaeo-diluvial and alluvial or fluvial and lacustrine

182


plains (Dong et al., 2008). The Kumutage Desert is also situated at the junction between a northeast wind system and a northwest wind system, resulting in complex dune types. The feather-shaped sand dunes account for about one fifth of the whole desert area (Figure 4), being famous in the world due to its uniqueness (Dong et al., 2008). The landscapes in the desert are mainly displayed by aeolian landforms and deflation landforms that both have clear individual action places in spatial (Zhu et al., 1980).

Figure 4 1980)

3

Geomorphological diagram of the Kumutage Desert in Northwest China (revised after Zhu et al.,

Provenances of dune sands in Xinjiang deserts

Recent studies have proofed that the availability of onsite loose sediments is more important than any other factors for occurrence or formation of a sand desert on a regional scale (Yang et al., 2012), so identification of provenances of aeolian sediments at a wide scale is of prime importance for understanding the formation and evolution of dune fields and sand deserts. Determining the sources of aeolian sediment in Central Asia, especially in the extensive dune field in Xinjiang, is crucial not only to the understanding of the formation of these sand deserts but also help to promote a better understanding of the past and present global climate system, because studies have shown that the Xinjiang deserts (especially the Taklamakan Desert) in the hinterland of Central Asia is a large source for global dust production (Zhang et al., 2003; Goudie and Middleton, 2006) and is therefore a basis for understanding the global dust-aerosol processes and its climatic and environmental effects. The mineral dust that is being deposited in Greenland at recent times as well as during the Last Glacial period was recognized to be mainly from the Taklamakan Desert and from the deserts of Inner Mongolia (Svensson et al., 2000; Bory et al., 2003). The aeolian dust from the


183

Asian Continent is a major source of pelagic sediment in the western Pacific(Rea et al., 1998) and an important part of the sediment in the central North Pacific (kyte et al., 1993). In addition, the dust significantly affects the chemical composition of seawater in the western Pacific Ocean. It is recognized that the rare earth elements (REE) patterns of western Pacific Ocean seawater in the areas near Asia show close affinities to Chinese desert dust and loess (Greaves, 1999). In this context knowledge about the sources of aeolian sediments in the Xinjiang deserts has global significance. In general, opinions about the sources of the Xinjiang desert sands are diverse in the history. For instance, there are views thinking that the aeolian sand in Xinjiang deserts were blown from Central Asia, but some one thought they come from the Luobupo playas at the east (Zhu et al., 1981). However, several studies on dune geomorphology and sedimentology have the viewpoints that the deserts in Xinjiang were finally formed due to a naturally equipped feature of “in situ sand rising” (Zhu et al., 1981), because they found that the mineral and grain sizes compositions of the desert sands were only different between underlain dunes with different geological and geomorphological types, but without marked relationships with wind directions and transportation distances. 3.1

Provenances of dune sands in the Taklamakan Desert

The basin topography with its general terrain declined from southwest to northeast due to the neotectonic movement during the Tertiary and the Quaternary periods has limited the distributions of water drainage system and the nature of sediments in the Tarim Basin. Drilling data display that the sedimentary strata of 100–200 m in depth are all fine silty sand layers in the southwestern part of the Taklamakan Desert, and are alluvial sandy layers and silt layers below 300 m in depth. At the Tarim river plains in the northern part of the desert, the depth of Quaternary loose sandy sediment strata is beyond 400–500 m. The mineral and grain size compositions of dune sands are closely related to the underlying sediments (Wu, 1982). At the southwestern part of the desert, because the Kunlun Mountain mass is mainly composed of the age-old metamorphic rock series such as the Archaean and Proterozoic gneiss, schist and phyllite, the mineral compositions of aeolian sediments sourced from various river loads that originated from the Kunlun Mountains are abundant in a certain number of metamorphic minerals like garnets (Zhu et al., 1981). At the north Taklamakan, because the Tianshan Mountain mass is are mainly composed of the relatively younger igneous rocks and minerals in the Paleozoic and Mesozoic ages, so the garnet contents in the mineral compositions of the aeolian sand sediments from the Tarim alluvial plain are obviously more abundant than those in the sands from the south Taklamakan Desert, while mica becomes the major mineral and the second is epidote, which are generally consistent with the Tarim river alluviums in mineral composition (Zhu et al., 1981; Wu, 1982). After analyzing the mechanical and mineralogical compositions of the Quaternary alluvial and diluvial sediments derived from areas outside and inside the Tarim Basin, Zhu et al. (1981) proposed that the fluvial sediment was the major source of aeolian sands in the Taklamakan Desert, and based on the compositions of heavy minerals and grain size they thought that the dune sands at different areas in the Taklamakan Desert were locally originated. However, early scientists getting foot into these areas of the Taklamakan Desert have obviously different opinions, much of these opinions can be sorted as a kind of single-sourced view. For example, Siwen Hotin et al. thought that

184


the Taklamakan sands were blown from the Luobupo Depressions (Zhu et al., 1981), and Sumof B B et al. suggested that the Taklamakan sands were sourced from marine sediments (Zhu et al., 1981), while Narin E and Senicun B (Zhu et al., 1981) and Zhoufangjianwei et al. (1990) thought that the desert sands in Taklamakan were derived from ancient lacustrine sediments. Opinions of recent studies from China and abroad are diverse. For instance, based on the mineralogical and isotopic analysis of aeolian sands Qian et al. (1993) argued that dune sediments at the edges of the Taklamakan were originated from the in-situ rising sands and dune sediments in the hinterland of the Taklamakan had received exogenous materials and become homogenized. Mu et al. (1990) speculated that the provenances of the Taklamakan desert sands should be closely related to the surrounding river systems and these sediments were almost stable in texture and property. Chen (1993) suggested that dune sands in central Taklamakan had higher maturity and therefore the desert sands must have expanded from the central part to the peripheral areas. Based on the major elements and minerals compositions and grain-size distributions of dune sands from southwestern Taklamakan, Honda and Shimizu (1998) argued that the dune sands in Taklamakan, compared to dune sands from most of the world deserts such as the Libya Desert, have higher Quartz/Feldspar ratio values and Calcite/Quartz ratio values, finer grain sizes and smaller abrasion roundness. This is consistent with earlier studies (Zhu et al., 1981; Besler, 1991) on Taklamakan. However, these opinions about the sources and transportation histories of the Taklamakan desert sands have been divided into two viewpoints: one suggests that the dune sands in different areas of the Taklamakan Desert are diversely-sourced because their heavy mineral compositions are dissimilar (Zhu et al., 1981), while the other proposes that the dune sands in the entire Taklamakan Desert are homogenous, because the bulk sand samples in this desert are consistent in their major elements and isotopic compositions (Honda and Shimizu, 1998; Hattori et al., 2003). Recently, along with the further systematization of research objects and the wide applications of isotopic and geochemical methods in Taklamakan, many scholars agree that the provenances of the Taklamakan desert sands are heterogeneous and not single-sourced. According to the fact that the Sr and Nd isotopic compositions of glacial tills in the mountains around the Taklamakan Desert are obviously different, Chang et al. (2000) argued that the Taklamakan desert sands, which have taken the glacial tills as their important primary sources, should be unlikely homologous in origin; besides, each glacier in the northern flanks of the Kunlun Mountains is constrained in a single valley and therefore the sediments in these source areas can not be mixed along a direction from east to west; on the other side, not all of sand materials are derived from glacial tills, because strong river erosion processes and mechanical weathering processes in these regions can also generate mass diluvial-alluvial and fluvial-lacustrine loose sediments, thus increasing the sediment heterogeneity in the source region. Sedimentological studies have shown that dune sands in the Taklamakan Desert are still in an early stage of dune development (Besler, 1991), because there are enough fine particles in the bulk samples having not been blown away, and the sand sea of Taklamakan will become rejuvenated by the input of fluvial sediments (Yang et al., 2007). The abundant existence of large fine grain fraction in dune sands supports the idea of heterogeneity of sand sources in the Taklamakan, because the large sediment volume reduces the potentials of mixing and the effects of aeolian abrasion (Yang et al., 2007). In contrast


185

with other desert sands and aeolian sediments in northern China, both of the coarse and fine fractions of desert sands in Taklamakan have lower values of chemical weathering index (CIA = 50–55) (Zhu and Yang, 2009) and higher content of soluble salts (Zhu and Yang, 2010; Zhu et al., 2012), being close to the mean CIA values of the unweathered Upper Continental Compositions (UCC) (CIA = 50) and in the early stage of continental weathering processes, indicating that the source areas of aeolian sand in the Taklamakan Desert have only experienced a relatively lower chemical weathering process (Zhu and Yang, 2009). This is at least partially related to the lasting arid climate and the limited exposure times and ages of sand sediments (Yang et al., 2006, 2007). Based on REE and major elements evidences, Yang et al. (2007) have found that the trace elements’ abundances and the REE characteristics are all notably distinguished between the coarse and the fine fractions of dune sands in Taklamakan, indicating that the coarse fraction of aeolian sands from different sites in Taklamakan have different sources rather than a homogenization of the entire basin; the relative homogeneity of the fine fraction of desert sands in different sites of the desert indicates that the silt fraction could have experienced a higher degree of mixing processes and thus tends to be homogenous. The regional difference of coarse fractions in the Taklamakan is consistent with the fluvial and wind systems in the basin (Yang et al., 2007). This confirms that the sands are often mixed between the northern and southern parts but there is much less mixing along the east–west direction. 3.2

Provenances of sands in the Guerbantonggute Desert

As to the provenances of sand sediments in the Guerbantonggute Desert, the modern sediments may be far less important than the ancient sediments in the Junggar Basin. Because the wide and thick sedimentary deposits in the Junggar Basin are not adapt to the distribution of the present-day basin drainage systems, as the lengths of river channels and water discharges of modern rivers in the Junggar Basin are not big and strong enough to carry massive sediments into the basin on a large scale over a short time. It should be resulted from the effects of fluvial processes developed during the Quaternary wetter periods (Wu, 1997). Therefore, sources of aeolian sediments in the Guerbantonggute Desert should mainly come from the ancient fluvial sediments. Based on mineralogical analysis, Qian et al. (2001) suggest that the sources of the desert sand in Guerbantonggute are diverse. Generally speaking, in the overall processes of transportation and sedimentation in the Junggar Basin, various kinds of detrital debris derived from surrounding mountains should be the major sources of the sediments in the central basin. However, weathering and denudation products of bedrocks from the tablelands and highlands in the Junggar Basin have also important contributions to the sources of desert sands. The characteristics of mineral compositions of desert sands from different sites in the desert also differ from each other, but in a local area between different dune types and between dunes and inter-dunes the compositions of dune sands are relatively similar, proofed the opinion of Zhu et al. (1980) that “in situ sand rising” is the principal cause of dune formation, namely the dune sands are mainly sourced from the underlying sand. But the compositions of heavy minerals and the major mineral dispersions and the quartz/feldspar ratios between dune sands and the underlying sands also vary to a certain extent, indicating that there is a kind of mixing and transformation role of wind forces on sediments during the evolution of underlying sand toward dune sands. Besides,

186


climate and vegetation and other environmental conditions could have also influenced this role and thus make dune sands have wider sources than the underlying sediments. 3.3

Provenances of sands in the Kumutage Desert

The Kumutage Desert is 350 km in length from east to west and 120 km in width from south to north, with a total area about 2000 km2. For such a large desert, due to the differences between various underlying sediments in the local area and vicinity of the desert, the sources of aeolian sands must have been influence by the underlying sediments and should be diverse in origin. The grain sizes of most of sand particles in the Kumutage Desert is less than 0.5 mm, especially between 0.25–0.1 mm and below 0.1 mm, with the fine particles as the majority (Xia, 1987). Looking from the abrasion roundness of sand particle, the roundness of sand in the higher dunes of the southern part of the desert reaches about 0.7, and that of sand in the northern part is usually about 0.3, and the overall roundness of the whole desert is relatively lower, indicating that sand particles in this desert have only experienced a short time of wind processing and the transportation distance is relatively close. Surface morphology and ultrastructure of quartz particles in the desert are mainly shown as pitting surface, butterfly dent, upend lamina, slot and faulting, etc, indicating that the particles have experienced distinctly wind processing. The mineral compositions are mainly quarts (70%–80%) and feldspar. In heavy minerals, the sum of four kinds of minerals such as magnetite, epidote, garnet and hornblende accounts for more than 50% of the total and the heavy mineral assemblage of dune sands is similar to that of the underlying sediments (Xia, 1987). Thus, the compositional characteristics of aeolian sediments indicate that most of dune sands in the Kumutage Desert are sources from the underlying sediments that have been blown up by wind and accumulated, and finally formed by ‘in situ uprising of local sand’. According to the compositions of dune sands at different sites in the Kumutage Desert, Xia (1987) speculated that the wide lacustrine and fluvial sedimentary plains in the Luobupo area and the large-scale outcropped Yadan formations composed of the Quaternary lacustrine sediments in the north edge of the desert are all made up of fine sand and clay, which are the major sources of the Kumutage Desert sands under the effect of north wind. Fluvial sediments are widely distributed in the areas in the eastern part of the desert that the Sulehe River flows through, and they could be the major sources of dune sand in the eastern part of the desert. At the south foot of the Keziletage Mountains and the northern part of Sanlongsha, the widely distributed diluvial sediments could be provided as a major source for dunes sands in the northeastern part of the desert. The low hills in the Aqike valleys, which are composed of the Tertiary and older-aged rocks, are chiefly mudstones and sandstones and are strongly denudated and mechanically weathered so that having yielded large numbers of bedrock weathering products such as eluvial sediments and slope-slide sediments, which are also the important sources of the Kumutage Desert sands. 3.4

The heterogeneity and complexity in the provenances of desert sands in Xinjiang

Generally speaking, the point of view about the provenances of aeolian sands in deserts of Xinjiang has already changed in our minds from a perspective of single-sourced (a plain opinion) to multi-sourced (heterogeneity) and has experienced a variation from qualitatively description to quantitatively identification. At present, the opinion on the heterogeneity in


187

sediment sources has been widely accepted due to it being more detailed and restrictive. The occurrences and applications of some new technological methods and research approaches in the radioactive and stable isotopes, geochronology, geochemistry and spatial information system and modeling have helped to a systematic and logical understanding on the provenances of aeolian sands in deserts of Xinjiang. However, there are still many questions having no clear answers until now. For instance, (1) what aeolian-transportation processes have the Xinjiang sand deserts experienced during the Quaternary? What the real transportation approaches and modes are at that time? What the degree and strength of aeolian processes are and how long the duration time is? What the recycle degrees of aeolian sediments are in these deserts? (2) How did the hydromorphic dynamics (such as rivers, lakes and glaciers) influence the aeolian processes and the sources of aeolian sediments? What are the relationships in sedimentology and provenances between aqueous deposits and aeolian deposits? (3) The geochronology of the provenances of aeolian sediments in the Taklamakan Desert and its sedimentological history. (4) What is the role of the local geomorphological factors and the large-scale environmental and geological factors in controlling the provenances of aeolian sediments? (5) What the tectonic environments (such as the passive continental margin, the positive continental margin or the island arc environment) and the climatic conditions (precipitation rate, temperature gradient or atmospheric circulations) are in the source area of aeolian sediments from the desert? How to identify them? (6) What kinds of alteration are between the detrital sediments and their source materials? What is the alteration degree? How much is the source material information that detrital sediments have inherited and which kinds of factors are they have been influenced? How to build up an effective indicator? (7) Whether the changing degree of maternal sediments after deposited have responded to and recorded the changes of hydrological and climatic and geological conditions at regional or other scales? Apparently, to recognize and understand the heterogeneity in the provenances of desert sands in Xinjiang and to answer the questions above mentioned, further research works are still needed to carry out in the future. Looking from the aspects of tectonics and wind circulation settings, both of the south and the north basins in Xinjiang are characterized by aeolian-and-fluvial cross system with opposite direction between the tilted direction of basin topography and the wind direction of synthesis. For example, the general terrain in the Tarim Basin is tilted from southwest to northeast but the prevailed wind regime in the basin is north wind or northeast wind. While in the Junggar Basin, the land terrain is tilted from southeast to northwest and the full-year prevailed wind regime is northwest wind. As the tilted basin terrain determined by the tectonic settings controls the flow direction of rivers in the basin, namely the transportation direction of fluvial sediments, while the synthesized direction of winds determined the moving directions of shifting dunes, namely, the transportation direction of aeolian sediments. So, the specific tectonic origin of the two desert basins in Xinjiang and their atmospheric circulation patterns determine that there should be a large geological cycles between the detrital sediments transported by aeolian processes and fluvial processes respectively. This should to a large degree have caused the cycling and mixing and alternation of sediments between young and old sources in the basins and increased the heterogeneity and complexity in the provenances of desert sands in Xinjiang. Although we can generalize the classification of provenances of aeolian sands in Xinjiang deserts into several types, such as

188


(1) palaeo-fluvial and -alluvial sediments, (2) modern fluvial and alluvial sediments, (3) diluvial and alluvial sediments, (4) alluvial and lacustrine sediments, (5) eluvial and slope-slide deposits from bedrock-weathered products, (6) modern lacustrine and paludal sediments, etc. (Zhu et al., 1980), at present, these judgments are almost qualitative or semi-quantitative and need to be constrained further by detailed geochemical and isotopic evidences.

4

Conclusions

Landscape of sandy desert is a geological product of arid climate and abundant sand materials supply in natural conditions. Responding to variations in climate and source supply, the development of sand desert will change. Formations and evolutions of sand deserts are usually controlled by factors between sand sources, climate (wind regime, aridity/precipitation, hydrology and vegetation) and the underlying surface conditions (tectonic setting, topography), therefore the formation of sandy desert is an ideal research object for understanding the interaction between various stratigraphic/epigenetic spheres of the earth system. However, until now, the knowledge about the provenance and formation of deserts in Xinjiang, Central Asia is very diverse and inconsistent, a systematic review is necessary. Ancient aeolian sand sediments in the internal areas of these deserts and their twin dust deposits in the Cenozoic strata surrounding the deserts are direct and indirect clues reflecting the provenance of sand materials and the formation and evolution of deserts. This paper reviews the research progress in the development of the sandy deserts in Xinjiang, Northwest China based on the geochemical, mineralogical and isotopic evidences of desert sands and relevant deposits in the peripheral regions. Generally speaking, the point of view about the provenances of aeolian sands in Xinjiang deserts has already changed in academic communities from a perspective of single-sourced (a plain opinion) to multi-sourced (heterogeneity), and also has experienced a variation from qualitatively description to quantitatively identification. At present, the opinion of the heterogeneity on sediment sources has been widely accepted. Many proofs proposed that desert sands in Xinjiang were mainly sourced from the ancient pluvial, alluvial and fluvial sediments and were eventually achieved because of the local sand supply. Looking from the aspects of tectonics and wind circulation settings, both of the south and the north basins in Xinjiang are characterized by aeolian-and-fluvial cross system with opposite direction between the tilted direction of basin topography and the wind direction of synthesis. It pointed out that the settings of tectonic structure in Xinjiang had made sediments both in the Tarim Basin and the Junggar Basin being influenced greatly by regional hydrological system and aeolian processes originated from the planetary wind system of westerly, the East Asian winter monsoon and the topographical mountain-valley winds. The different directions of transportation paths between the hydrological dynamics and the aeolian dynamics have determined anisotropies of the transport pathway of these sediments delivered by different agents. This results in a geological cycle of detrital sediment transportation in the two large basins and thus largely increases the degree of sediment mixing, alternation and recycle between younger and older sediment sources, as well as the complexity and diversity of sand provenances.


189

Acknowledgements We are grateful to the managing editor Dr. Lufeng YAO and Xin ZHAO for their constructive comments and suggestions. Sincere thanks are also extended to Prof. Xiaoping YANG for his generous help and support in the study work.

References Besler H, 1991. The Keriya Dunes: First results of sedimentological analysis. Die Erde Erg.-H, 6: 73–88. Besler H, 1995. The Keriya dunes in the Taklimakan Sand Sea: Sedimentological evidence for a polygenetic evolution. Die Erde, 126: 205–222. Bory A, Biscaye P, Grousset F, 2003. Two distinct seasonal Asian source regions for mineral dust deposited in Greenland (NorthGRIP). Geophysical Research Letters, 30(4): 1167, doi: 10.1029/2002GL016446. Chang Q, Mishima T, Yabuki S et al., 2000. Sr and Nd isotope ratios and REE abundances of moraines in the mountain areas surrounding the Taklimakan Desert, NW China. Geochemical Journal, 34: 407–427. Chen W, 1993. Grain size parameters of arolian sediments in the vicinity of the longitude 84°E, Taklamakan Desert. Acta Geographica Sinica, 48(1): 33–46. (in Chinese) Dong Z, Qu J, Wang X et al., 2008. Pseudo-feathery dunes in the Kumtagh Desert. Geomorphology, 100: 328–334. Greaves M, Elderfield H, Sholkovitz E, 1999. Aeolian sources of rare earth elements to the Western Pacific Ocean. Marine Chemistry, 68: 31–37. Goudie A S, Middleton N J, 2006. Desert Dust in the Global System. Berlin: Springer. Guo Z, Ruddiman W, Hao Q et al., 2002. Onset of Asian desertification by 22 Myr ago inferred from loess deposits in China. Nature, 416: 159–163. Hattori Y, Suzuki K, Honda M et al., 2003. Re-Os systematics of the Taklimakan Desert sands, moraines and river sediments around the Taklimakan Desert, and of Tibetan soils. Geochimica et Cosmochimica Acta, 67: 1195–1205. Honda M, Shimizu H, 1998. Geochemical, mineralogical and sedimentological studies on the Taklimakan Desert sands. Sedimentology, 45: 1125–1143. Kyte F, Leinen M, Heath G et al., 1993. Cenozoic sedimentation history of the central North Pacific: Inferences from the elemental geochemistry of core LL44-GPC3. Geochimica et Cosmochimica Acta, 57: 1719–1740. Li Z, Chen G, 1999. Effective wind regime in the Taklimakan Desert. Journal of Desert Research, 19(1): 43–45. (in Chinese) Ma L, 2002. Atlas of Chinese Geology. Beijing: Geological Publishing House. (in Chinese) Mu G, Ji Q, 1990. Characteristics and significance of the mechanical composition of the Quaternary deposits of the Taklimakan Region, China. Arid Land Geography, 13(2): 31–36. (in Chinese) Qian Y, Wu Z, Ishii T et al., 1993. The constituent characteristics of sand materials and sand sources of the Taklamakan Desert. Journal of Desert Research, 13(4): 32–38. (in Chinese) Qian Y, Zhou X, Li C et al., 2001. Multi-sources of sand minerals for the deserts in the Jungger Basin. Journal of Desert Research, 21(2): 182–187. (in Chinese) Qu J, Zuo G, Zhang K et al., 2005. Relationship between the formation and evolution of the Kumtag Desert and the regional Neotectonic Movement. Arid Land Geography, 28(4): 424–428. (in Chinese) Rea D, Snoeckx H, Joseph L, 1998. Late Cenozoic eolian deposition in the North Pacific: Asian drying, Tibetan uplift, and cooling of the northern hemisphere. Paleoceanography, 13: 215–224. Svensson A, Biscaye P, Grousset F, 2000. Characterization of late glacial continental dust in the Greenland Ice Core Project ico core. Journal of Geophysical Research, 105(D4): 46374656. Sun J, Liu T, 2006. The age of the Taklimakan Desert. Science, 312: 1621–1621. Sun J, Zhang L, Deng C et al., 2008. Evidence for enhanced aridity in the Tarim Basin of China since 5.3 Ma. Quaternary Science Reviews, 27: 1012–1023.

190


Sun J, Zhang Z, Zhang L, 2009. New evidence on the age of the Taklimakan Desert. Geology, 37(2): 159–162. Wang X, Eerdun H, Zhou Z et al., 2007. Significance of variations in the wind energy environment over the past 50 years with respect to dune activity and desertification in arid and semiarid northern China. Geomorphology, 86: 252–266. Wu Z, 1982. Sandy Deserts in China. Beijing: The Commercial Press. Wu Z, 1997. The basic characteristic of landform development in the Jungar Basin. In: Wu Z (ed.). Studies on the Sandy Deserts and the Coastal Dunes in China. Beijing: Science Press, 29–43. (in Chinese) Xia X, 1987. The Basic characteristics of the Kumtag Desert. In: XDCAS (Xinjiang Division of Chinese Academy of Sciences) (eds.). The Scientific Survey and Study on the Luobupo. Beijing: Science Press, 78–94. (in Chinese) Yang X, Li H, Conacher A, 2012. Large-scale controls on the development of sand seas in northern China. Quaternary International, 250: 74–83. Yang X, Preusser F, Radtke U, 2006. Late Quaternary environmental changes in the Taklamakan Desert, western China, inferred from OSL-dated lacustrine and aeolian deposits. Quaternary Science Reviews, 25: 923–932. Yang X, Scuderi L, Paillou P, 2011. Quaternary environmental changes in the drylands of China: A critical review. Quaternary Science Reviews, 30: 3219–3233. Yang X, Shi C, Li B et al., 2008. Some aspects about Chinese Geomorphology: Recent progresses from an earth system science perspective. Quaternary Sciences, 28(4): 521–534. (in Chinese) Yang X, Zhu B, Whiter P D, 2007. Provenance of Aeolian sediment in the Taklamakan Desert of western China, inferred from REE and major element data. Quaternary International, 175: 71–85. Zhang X, Gong S, Zhou T et al., 2003. Sources of Asian dust and role of climate change versus desertification in Asian dust emission. Geophysical Research Letters, 30(24): 2272. Zhoufang J W, Xingye GX, Ronglin Z Z et al., 1990. Sands in the Taklamakan Desert of China. Arid Zone Research, 4: 66–70. (in Chinese) Zhu B, Yang X, 2009. Chemical weathering of detrital sediments in the Taklimakan Desert, northwestern China. Geographical Research, 47(1): 57–70. Zhu B, Yang X, 2010. The origin and distribution of soluble salts in the sand seas of northern China. Geomorphology, 123: 232–242. Zhu B, Yang X, Liu Z et al., 2012. Geochemical compositions of soluble salts in aeolian sands from the Taklamakan and Badanjilin deserts in northern China, and their influencing factors and environmental implications. Environmental Earth Sciences, 66: 337–353. Zhu Z, Chen Y, Wu Z, 1981. Study on the Aeolian Geomorphology of the Taklamakan Desert. Beijing: Science Press. (in Chinese) Zhu Z, Wu Z, Liu S, 1980. Outline of Chinese Sandy Deserts. rev. ed. Beijing: Science Press. (in Chinese)