ORIGINAL ARTICLE

Effectiveness of rare earth elements constrain on differentmaterials: a case study in central Asia

Chaoliu Li • Shichang Kang • Qianggong Zhang •

Chhatra Mani Sharma

Received: 18 December 2010 / Accepted: 28 January 2012 / Published online: 15 February 2012

� Springer-Verlag 2012

Abstract Rare earth elements (REE) have been exten-

sively used to indicate for material provenance since they

behave conservatively and mainly transport in particulate

form during earth surface processes. Nevertheless, the

application of REE for material provenance study has to be

cautious because grain size and mineral fractionations can

alter bulk compositions of weathered sediments. Central

Asia is one of the most important dust source regions

globally and numerous studies on REE compositions of

surface materials have been conducted. In this study, REE

compositions of various materials from this area are sum-

marized to explore the existing REE-related problems.

Overall, chondrite-normalized REE patterns for many

surface materials are so uniform that they cannot serve as

reliable approaches in tracing material source regions. In

contrast, great variations of REE compositions occur

among different materials that are derived even from the

same parent rock due to influences of grain-size distribu-

tions and heavy minerals. For the same reason, small-scale

loess around the Tibetan Plateau has different upper con-

tinental crust (UCC)-normalized REE patterns compared to

those of typical loess. Therefore, great cautions should be

made when UCC-normalized REE patterns and REE ratios

are utilized to investigate material provenance. Finally,

some suggestions are proposed for such studies in future.

Keywords Rare earth elements � Material provenance �Upper continental crust � Central Asia

Introduction

Rare earth elements (REE) are a series of typical immobile

elements that do not readily fractionate during earth surface

processes (Taylor and McLennan 1985). This set of ele-

ments is widely used as robust evidence to determine

source regions of dust aerosols (Svensson et al. 2000; Yang

et al. 2007a; Zdanowicz et al. 2006; Zhang et al. 2009),

loess (Honda et al. 2004; Sun 2002a), fluvial and marine

sediments (Nakai et al. 1993; Yang et al. 2002), especially

after the publication of a classic book by Taylor and

McLennan (1985). The theories proposed in this book

provide fundamental principles for subsequent geochemi-

cal studies. However, some theories on REE mentioned in

the book sometimes are partially understood and misused.

Among them the commonly misused two theories are

‘‘there is no significant difference in the REE patterns

between the finest fractions (\1 lm) and the silt-sized

(\20 lm) fraction (P37)’’ and ‘‘the nature of the Europium

(Eu)-anomalies is similar in all of the size fractions con-

sidered (P38)’’. Based on these two theories it is sometimes

even considered that REE compositions are constant in all

size fractions of the same material. Consequently, REE

compositions and Eu anomalies were simply compared

among different grain sizes in material provenance studies

(Chang et al. 2000; Sun et al. 2007; Wu et al. 2009; Yang

et al. 2007a), ignoring those cautions noted by the authors

in the same book. Many of these cautions are related to the

C. Li � S. Kang (&) � Q. Zhang � C. M. Sharma

Key Laboratory of Tibetan Environment Changes and Land

Surface Processes, Institute of Tibetan Plateau Research,

CAS, Beijing 100085, China

e-mail: shichang.kang@itpcas.ac.cn

S. Kang

State Key Laboratory of Cryospheric Sciences,

CAS, Lanzhou 730000, China

C. M. Sharma

Human and Natural Resources Studies Centre, Kathmandu

University, P.O. Box 6250, Dhulikhel, Kathmandu, Nepal

123

Environ Earth Sci (2012) 67:1415–1421

DOI 10.1007/s12665-012-1586-2

influence of heavy minerals such as: ‘‘heavy minerals

reside most in silt fraction is generally HREE-enriched and

high in total REE, and may significantly affect REE pat-

terns of the sediments when clay fractions are minor

(P31)’’. Additionally, all of these theories are inferred

mainly from studies of fine-grained clayey sediments and

chondrite-normalized REE patterns, hence cannot be gen-

eralized to other materials directly and are not totally fit for

the upper continental crust (UCC) normalized REE pat-

terns and REE ratios. Therefore, great cautions should be

made when these theories are applied directly to prove-

nance studies of fine clay-lacked materials.

Central Asia is one of the largest dust source regions in the

world (Prospero et al. 2002), and dust derived from this area is

of particular interest in research areas such as paleoclimate

(Ding et al. 2001), atmospheric circulation (Aloys et al. 2003;

Kang et al. 2002; Prospero et al. 2002; Rea et al. 1998), and

even tectonic activities (An et al. 2001). The Loess Plateau

(LP) and the Taklimakan Desert (TAK) are the two most

important dust source areas in this region. It has been pro-

posed that loess is contributed from deserts, materials of

which are initially derived from various processes such as

glacial grinding and frost weathering in high mountain areas

(Sun 2002a, 2007). Meanwhile, the TAK and its surrounding

areas possess various materials from primary weathered

material (e.g., moraine) to highly weathered materials (e.g.,

loess and dust), serving as an ideal place to deeply discuss

variations of REE compositions among different materials.

To date, numerous studies on REE compositions have been

conducted around this area (Fig. 1) (Chang et al. 2000; Honda

et al. 2004; Liu et al. 1993; Sun 2002a; Wu et al. 2008; Yang

et al. 2007b), while lots of researchers only performed direct

comparison of REE values among materials with totally

different grain-size distributions and mineral compositions

(Chang et al. 2000; Liu et al. 1993; Wu et al. 2008, 2009;

Yang et al. 2007a), which is unrigorous and may result in false

conclusions. A sophisticated analysis is required to evaluate

interrelations among REEs and corresponding control factors

of REE compositions of these materials. With this as moti-

vation, we performed a critical analysis of variations in REE

compositions among different materials from central Asia by

integrating data from former studies. Two examples of small-

scale loess and ice core dust relating to the TAK and the

Tibetan Plateau (TP) are particularly discussed to illustrate

factors that deserve attention in future REE-related studies.

As influences of grain-size distributions and heavy minerals

on REE compositions are major topics of this study, Cerium

(Ce) anomalies are not discussed since they are generally

closely related to redox conditions of samples (Brookins

1989; Rankin and Childs 1976) and kept constant among

different grain sizes. In addition, although redox system may

change during transport and sedimentation, Ce anomalies

keep almost constant among different materials due to rela-

tively cold and dry environments of studied area (Fig. 2).

Materials

Published REE concentrations from central Asia are listed

in Table 1. Materials of bulk and fine fractions for the TAK

Fig. 1 Study location of

central Asia

1416 Environ Earth Sci (2012) 67:1415–1421

123

sand, loess accumulated around the TP, the typical loess,

sediment, and dust deposited on the Dunde glacier (Fig. 1)

are adopted to investigate variations of REE concentrations

and compositions.

Results and discussion

Chondrite-normalized REE patterns

Generally, chondrite-normalized REE patterns, UCC-nor-

malized REE patterns and REE ratios are three effective

approaches for describing and comparing REE composi-

tions among different materials. However, ratios of REE

concentrations between normal surface materials and

chondrite usually vary so widely that some detailed

characteristics cannot easily be detected (Marx et al. 2005),

especially for those materials with uniform REE compo-

sitions due to multiple recycle processes. For example,

chondrite-normalized REE patterns of all materials listed in

Table 1 are uniform with steep LREE, fairly flat HREE

profiles and moderate Eu depletion (Fig. 3). This phe-

nomenon is in accordance with those theories suggested by

Taylor and McLennan (1985). However, significant dif-

ferences appear for UCC-normalized REE patterns

(Fig. 4). Despite chondrite-normalized REE patterns have

been successfully used in distinguishing material prove-

nance in some studies (Nakai et al. 1993; Taylor and

McLennan 1985), UCC-normalized REE patterns and REE

ratios are another two prevalent approaches in source

tracing studies (Wu et al. 2009; Yang et al. 2007a).

Therefore, the effectiveness and potential inaccuracy in

practical use of these two methods are evaluated in the

following parts.

Fig. 2 Relationships between Eu anomalies among various materials

within the TAK. Data of river sediments are adopted from Liu et al.

(1993)

Fig. 3 Chondrite-normalized REE distribution patterns for materials

listed in Table 1

Table 1 REE concentrations (mg/kg) in different materials and grain sizes from the TP and central Asia

Sample sites La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu

1 (n = 9) 68.54 140.32 15.69 57.39 10.84 1.45 8.28 1.27 7.55 1.41 4.1 0.61 4 0.61

2 (n = 35) 88.84 179.13 20.24 73.55 14.19 2.16 13.27 2.25 10.35 1.86 6.11 0.79 5.47 0.82

3 (n = 8) 33.66 67.61 7.5 26.98 4.88 0.86 3.92 0.63 3.54 0.75 1.98 0.29 1.91 0.28

4 (n = 10) 44.26 85.64 10.08 37.69 7.2 1.27 5.58 0.86 5.05 1.03 2.75 0.41 2.55 0.39

5 (n = 39) 32.6 68.9 7.7 29.5 5.9 1.2 5.4 0.8 4.7 1 2.9 0.4 2.9 0.4

6 (n = 15) 31.83 68.48 7.59 27.91 5.63 0.87 4.89 0.83 4.73 0.95 2.77 0.41 2.67 0.39

7 (n = 18) 70.11 135.52 15.47 52.95 9.75 1.68 8.28 1.31 7.02 1.53 4.01 0.6 3.84 0.6

8 (n = 17) 32.91 66.31 7.52 28.01 5.57 1.1 5.44 0.76 4.56 0.92 2.68 0.38 2.47 0.37

9 (n = 8) 32.58 67.91 7.77 28.54 5.52 1.02 4.87 0.75 4.34 0.83 2.42 0.36 2.32 0.35

1:\45 lm fraction of the TAK sand (Honda et al. 2004); 2:\53 lm fraction of TAK (Yang et al. 2007b); 3: bulk sand samples of TAK (Honda

et al. 2004); 4: loess of the south TP (from Sun JM, personal communication); 5: loess of the LP (Ding et al. 2001); 6: moraine of high mountains

around TAK (Chang et al. 2000); 7: \63 lm fraction of the Yarlung Tsangbo sediment (Li et al. 2009); 8: dust trapped in Dunde ice core

(Wu et al. 2009); 9: loess of the TAK (Honda et al. 2004)

Environ Earth Sci (2012) 67:1415–1421 1417

123

UCC-normalized REE patterns from materials

of the TAK

UCC-normalized REE patterns from moraine to dust of the

TAK are plotted in Fig. 4. It has been suggested that these

materials are closely connected with each other (Chang

et al. 2000; Sun 2002a; Yang et al. 2007b) and the pro-

cesses can be described briefly as follows: first, large

amount of clastic materials such as moraine are produced

by high mountain weathering processes in the TP and the

Tianshan surround the TAK. Second, weathered clastic

materials are transported to the Tarim basin by ephemeral

glacial river to form the TAK sand. Finally, fine fractions

of TAK sands are entrained to high altitude during dust

storm periods and transported to distant areas by the

westerlies. During this process some dust materials are

accumulated on slope of mountains facing the TAK to form

small-scale loess (Sun 2002a). However, REE patterns

among some of these materials vary greatly due mainly to

grain size and mineral sorting (Fig. 4). For example, both

the\45 and the\53 lm fractions of the TAK sand exhibit

higher REE concentrations, totally different REE patterns

and strong negative Eu anomalies compared to those of

other materials. These characteristics are different from the

general opinion that the finer fraction holds the higher REE

concentrations (Taylor and McLennan 1985); and it is

attributed mainly to the influence of REE-enriched heavy

minerals that are generally resided in coarse silt fraction

(Liu et al. 1993; Taylor and McLennan 1985; Yang et al.

2007b). Meanwhile, study on grain-size distributions has

shown that the TAK sand lacks fine clay fractions (Wang

et al. 2003), supporting the argument that heavy minerals

significantly affect REE concentrations and patterns when

clay fractions are minor (Taylor and McLennan 1985).

Thus, it is the lack of fine clay fraction in the \45 and the

\53 lm fractions of the TAK sand that leads to heavy

mineral-affected REE patterns. It is proposed that some

high field elements such as Zirconium (Zr) and Hafnium

(Hf) reside especially in REE-enriched heavy minerals

(e.g., Zircon) (McLennan 1989; Taylor and McLennan

1985). Correspondingly, Zr concentration within the

\53 lm fraction for the TAK sand is abnormally high

(446.92 mg/kg) (Yang et al. 2007b), exceeding the

threshold of 300 mg/kg (Taylor and McLennan 1995),

further indicating the contribution of heavy minerals to

REE concentrations and compositions. For the same rea-

son, moraines at high altitude mountain areas generally

experience immature weathering process and almost all the

fine clay fractions are removed away by wind and water,

resulting in abnormal REE patterns.

Virtually, a distinct influence of heavy minerals on REE

compositions was widely observed for many weathered

materials in the TP. In southern TP, similar to those of the

\45 and\53 lm fractions for the TAK sand, the\63 lm

fraction of the Yarlung Tsangbo (YT) sediment also

exhibited higher REE concentrations and different REE

patterns when compared with those of the bulk sediment

(Li et al. 2009). In addition, some bulk sand samples from

the TAK and bulk river sediments from the eastern TP

exhibited heavy mineral-related REE patterns (Borges

et al. 2008; Honda et al. 2004). It has been proposed that

some critical elemental ratios (e.g., Lanthanum (La)/

Ytterbium (Yb), Eu/Eu*) are not greatly affected by sedi-

mentary process (Taylor and McLennan 1985). However,

our analysis indicated that these two ratios vary with a

great range (Fig. 4), which should inevitably exert signif-

icant influence on material provenance related studies.

The TP is the youngest and highest plateau in the

world, its dry and cold environments result in extremely

immature weathered surface soil with relatively low

concentrations of fine clay fractions. Therefore influence

of heavy minerals on the REE patterns can easily be

unveiled out. This phenomenon is clear in the TAK sand

because fine clay fractions are constantly transported out

of the TAK by dust storms, resulting in intense abnormal

REE patterns for materials (e.g., the \53 lm fraction of

the TAK sand).

Distinct REE patterns for small-scale loess around

the TP

It has been suggested that REE compositions of loess are

uniform all over the world, even for loess with small-scale

source regions based on chondrite-normalized REE pat-

terns (Gallet et al. 1996; Taylor and McLennan 1985). In

addition, heavy minerals exert minor influence on loess

REE compositions (Gallet et al. 1996; Liu 1985) despite

Fig. 4 Upper continental crust-normalized REE distribution patterns

for different materials around the Taklimakan desert. LREE: La–Eu,

HREE: Gd–Lu, Eu/Eu* = EuN/SQRT (SmN 9 GdN)

1418 Environ Earth Sci (2012) 67:1415–1421

123

concentrations of some typical heavy mineral-related ele-

ments (e.g., Zr and Hf) of loess are higher than those of

UCC (Taylor and McLennan 1985). However, this is not

the case for the loess accumulated around the TP.

Large quantities of small-scale loess sediments are dis-

tributed around the TP and other high mountains in China

besides the famous LP (Fang et al. 1999; Sun 2002b; Sun

et al. 2007) (Fig. 1). REE compositions of small-scale loess

have been reported and attributed to local provenance

(Fang 1995; Honda et al. 2004; Sun 2002b; Sun et al.

2007). Honda et al. (2004) measured eight bulk loess

samples around the TAK. These samples can be easily

divided into two groups based on REE concentrations and

compositions. The first group includes five samples (sam-

ples 49, 105, 124, 134, 142) having uniform REE patterns

(hereafter called the ‘‘TAK1’’) resemble to that of the LP

loess; whereas the other group includes three samples

(samples 128, 129, 132) clearly having higher REE con-

centrations and stronger Eu depletion (hereafter called the

‘‘TAK2’’) (Fig. 5a). Interestingly, REE patterns of the

TAK2 are significantly inherited from that of the \45 lm

fraction of the TAK sand. In addition, these three samples

were collected from the same site, indicating that it is the

same wind regime transporting the heavy mineral-enriched

coarse fraction of the TAK sand to this place. Meanwhile,

REE patterns of the total TAK loess and the YT loess

accumulated in the southern TP (Sun et al. 2007) (Fig. 5a)

are similar, resulting from their similar grain-size distri-

butions (Fig. 5b). Generally, loess in the LP are transported

from areas hundreds of kilometers away, whereas loess

accumulated around the TP are subjected to relatively

short-distance transport (Sun et al. 2007), resulting in

totally different grain-size distributions between loess of

the LP and the TP (e.g., the Kunlun Mts. and the YT)

(Fig. 5b). It is clear that grain size of the \30 lm fraction

accounts for most parts of the LP loess, whereas this

fraction accounts for small parts of the Kunlun and YT

loess. Therefore, the lack of fine clay fractions of the

Kunlun and YT loess causes a clear influence on the heavy

minerals REE patterns. In summary, unlike conclusions

proposed by other study (Gallet et al. 1996), the UCC-

normalized REE patterns of small-scale loess accumulated

around the TP are distinct, resulting mainly from their

unique grain-size distributions and special mineral

compositions.

REE ratios

Similar to fine-grained sedimentary rocks, long-range

transported dust materials generally are sufficiently mixed

because they are derived from various kinds of material

rocks, resulting in quite uniform UCC-normalized REE

patterns. In this case REE patterns are not effective for

distinguishing provenance, and REE ratios are adopted

and proved as effective approaches (Sun 2002a; Taylor

and McLennan 1985; Wu et al. 2008). However, this

method tends to achieve unilateral or wrong results

because it is produced from fewer elements compared to

REE patterns. To address this issue, a provenance study

on dust trapped within the Dunde ice core (Fig. 1) over

the northeastern TP is adopted and discussed (Wu et al.

2009). In that study REE ratios (Ce/Yb vs. Eu/Yb) of ice

core dust and ‘‘fine’’ particle (the \53 lm fraction) of the

TAK sand were compared to demonstrate the contribution

of material of the TAK to the ice core dust. From Fig. 6a

it can be found that the REE ratios of the ice core dust fall

within those of the fine particles. Thus, it seems that dust

derived from the TAK is transported onto the Dunde

glacier. However, totally opposite conclusion can be

achieved from plot of other REE ratios (La/Gadolinium

(Gd) vs. Terbium (Tb)/Yb) (Fig. 6b), implying no con-

nection between sand of the TAK and the ice core dust.

Actually, the \53 lm fraction of the sand is not fit for the

Fig. 5 a Upper continental crust-normalized REE distribution

patterns for loess of the TP and the LP, b grain-size distributions of

loess from different regions (quoted from Sun 2002b; Sun et al. 2007)

Environ Earth Sci (2012) 67:1415–1421 1419

123

comparison. Firstly, only particles finer than 20 lm can

be entrained into air and transported over long distances,

whereas particles with size range of 20–70 lm can be

suspended in air and transported only for short distances

(Pye 1987). However, the distance between the Dunde

glacier and the TAK is about 1,000 km. Thus the \53 lm

fraction of the sand is unlikely to be transported onto the

Dunde glacier. Secondly, as discussed previously, the

\53 lm fraction contained relatively large amount of

heavy minerals, which greatly influence its REE compo-

sitions. Virtually, distinct differences of REE composi-

tions between the \53 lm fraction and the ice core dust

can be discovered from their UCC-normalized REE pat-

terns (Fig. 4), making consequent comparison between

REE ratios useless. Thus, only those fine particles (e.g.,

the \20 lm fraction) potentially being transported for

long distances are fit for comparison of REE ratios with

long-range transported dust. In summary, great cautions

should be made when REE ratios are used for provenance

study if either grain size or UCC-normalized REE patterns

among various materials are clearly different.

Conclusions

Several conclusions can be drawn based on the above

discussion:

1. Some REE-related theories proposed by Taylor and

McLennan are based on chondrite-normalized REE

patterns, which cannot simply be applied to material

provenance study. Generally, chondrite-normalized

REE patterns for almost all the surface materials from

central Asia are uniform and not vigorous in tracing

material source regions.

2. Heavy minerals can affect REE compositions of

materials that lack of fine clay fractions, especially

for moraine and some ‘‘fine’’ fractions (such as the

\45 lm fraction and the\53 lm fraction) of the TAK

sand. Therefore, REE compositions of these materials

cannot simply be used as effective tracing proxies in

dust and loess provenance studies.

3. Small-scale loess around the TP derived from local

regions has different UCC-normalized REE patterns

compared with that of typical loess, resulting from

their abnormal grain-size distribution characteristics

and relatively high heavy mineral concentrations.

Thus, if grain-size distributions of loess are greatly

different, it is not effective to compare their REE

compositions when performing provenance analysis.

4. Great care must be made when REE ratios are utilized

to distinguish provenance. If the target dust and its

potential source material have different grain-size

distributions or clear different UCC-normalized REE

patterns, the utilization of REE ratios is almost

worthless.

5. Finally, as REE compositions are closely connected

with grain-size distributions and mineral compositions

of the material, and it is generally hard to find same

fraction of the target material and potential provenance

material, REE compositions may hardly provide robust

evidence in material provenance studies. Therefore,

other approaches such as atmospheric circulation

patterns and isotopic compositions should also be

utilized as supporting evidences.

Acknowledgments This study was supported by the National Nat-

ure Science Foundation of China (40901270, 41171398 and

40830743), the Global Change Research Program of China

(2010CB951401), and State Key Laboratory of Cryospheric Science

(SKLCS-ZZ-2008-01). The authors acknowledge the staff at

NAMOR for great help in the field and thank Dr Shouye Yang and

Dr Jinliang Feng for their assistance on both science and English.

Fig. 6 Relationships between Eu/Yb and Ce/Yb (a), and Tb/Yb and

La/Gd (b) between dust samples from Dunde ice core and Taklima-

kan finer fraction of eolian sediment

1420 Environ Earth Sci (2012) 67:1415–1421

123

References

Aloys JM, Biscaye PE, Grousset FE (2003) Two distinct seasonal

Asian source regions for mineral dust deposited in Greenland

(NorthGRIP). Geophys Res Lett 30:1167

An ZS, Kutzbach JE, Prell WL, Porter SC (2001) Evolution of Asian

monsoons and phased uplift of the Himalaya–Tibetan Plateau

since Late Miocene times. Nature 411(6833):62–66

Borges JB, Huh Y, Moon S, Noh H (2008) Provenance and

weathering control on river bed sediments of the eastern Tibetan

Plateau and the Russian Far East. Chem Geol 254:52–72

Brookins DG (1989) Aqueous geochemistry of rare earth elements.

Rev Mineral Geochem 21:201

Chang Q, Mishima T, Yabuki S (2000) Sr and Nd isotope ratios and

REE abundances of moraines in the mountain areas surrounding

the Taklimakan Desert, NW China. Geochem J 34:407–427

Ding ZL, Sun JM, Yang SL, Liu TS (2001) Geochemistry of the

Pliocene red clay formation in the Chinese Loess Plateau and

implications for its origin, source provenance and paleoclimate

change. Geochim Cosmochim Acta 65:901–913

Fang XM (1995) The origin and provenance of Malan Loess along the

Eastern margin of Qinghai-Xizang (Tibetan) Plateau and its

adjacent area. Sci China Ser B Chem Life Sci Earth Sci

38:876–887

Fang XM, Li JJ, Van der Voo R (1999) Rock magnetic and grain size

evidence for intensified Asian atmospheric circulation since

800,000 years BP related to Tibetan uplift. Earth Planet Sci Lett

165:129–144

Gallet S, Jahn B, Torii M (1996) Geochemical characterization of the

Luochuan loess-paleosol sequence, China, and paleoclimatic

implications. Chem Geol 133:67–88

Honda M, Yabuki S, Shimizu H (2004) Geochemical and isotopic

studies of aeolian sediments in China. Sedimentology

51:211–230

Kang SC, Mayewski PA, Qin DH, Yan YP, Hou SG, Zhang DQ, Ren

JW, Kruetz K (2002) Glaciochemical records from a Mt. Everest

ice core: relationship to atmospheric circulation over Asia.

Atmos Environ 36:3351–3361

Li CL, Kang SC, Zhang QG, Wang FY (2009) Rare earth elements in

the surface sediments of the Yarlung Tsangbo (upper Brah-

maputra river) sediments, southern Tibetan Plateau. Quat Int

208:151–157

Liu TS (1985) Loess and the environment. China Ocean Press,

Beijing

Liu CQ, Masuda A, Okada A, Yabuki S, Zhang J, Fan ZL (1993) A

geochemical study of loess and desert sand in northern China:

implications for continental crust weathering and composition.

Chem Geol 106:359–374

Marx SK, Kamber BS, McGowan HA (2005) Provenance of long-

travelled dust determined with ultra-trace-element composition:

a pilot study with samples from New Zealand glaciers. Earth

Surf Proc Land 30:699–716

McLennan SM (1989) Rare earth elements in sedimentary rocks;

influence of provenance and sedimentary processes. Rev Mineral

Geochem 21:169–200

Nakai S, Halliday AN, Rea DK (1993) Provenance of dust in the

Pacific Ocean. Earth Planet Sci Lett 119:143–157

Prospero JM, Ginoux P, Torres O, Nicholson SE, Gill TE (2002)

Environmental characterization of global sources of atmospheric

soil dust identified with the NIMBUS 7 total ozone mapping

spectrometer (TOMS) absorbing aerosol product. Rev Geophys

40:1002

Pye K (1987) Aeolian dust and dust deposits. Academic Press,

London

Rankin PC, Childs CW (1976) Rare-earth elements in iron-manga-

nese concretions from some New Zealand soils. Chem Geol

18:55–64

Rea DK, Snoeckx H, Joseph LH (1998) Late Cenozoic eolian

deposition in the North Pacific: Asian drying, Tibetan uplift, and

cooling of the northern hemisphere. Paleoceanography

13:215–224

Sun JM (2002a) Provenance of loess material and formation of loess

deposits on the Chinese Loess Plateau. Earth Planet Sci Lett

203:845–859

Sun JM (2002b) Source regions and formation of the loess sediments

on the high mountain regions of northwestern China. Quat Res

58:341–351

Sun JM, Li SH, Muhs DR, Li B (2007) Loess sedimentation in Tibet:

provenance, processes, and link with quaternary glaciations.

Quat Sci Rev 26:2265–2280

Svensson A, Biscaye PE, Grousset FE (2000) Characterization of late

glacial continental dust in the Greenland ice core project ice

core. J Geophys Res Atmospheres 105:4637–4656

Taylor SR, McLennan SM (1985) The continental crust: its compo-

sition and evolution. Blackwell, Oxford

Taylor SR, McLennan SM (1995) The geochemical evolution of the

continental crust. Rev Geophys 33:241–265

Wang XM, Dong ZB, Zhang JW, Qu JJ, Zhao AG (2003) Grain size

characteristics of dune sands in the central Taklimakan sand sea.

Sed Geol 161:1–14

Wu GJ, Xu BQ, Zhang CL, Gao SP, Yao TD (2008) Geochemistry of

dust aerosol over the Eastern Pamirs. Geochim Cosmochim Acta

73:977–989

Wu GJ, Zhang CL, Gao SP, Yao TD, Tian LD, Xia DS (2009)

Element composition of dust from a shallow Dunde ice core,

Northern China. Glob Planet Change 67:186–192

Yang SY, Jung HS, Choi MS, Li CX (2002) The rare earth element

compositions of the Changjiang (Yangtze) and Huanghe (Yel-

low) river sediments. Earth Planet Sci Lett 201:407–419

Yang XP, Liu YS, Li CZ, Song YL, Zhu HP, Jin XD (2007a) Rare

earth elements of aeolian deposits in Northern China and their

implications for determining the provenance of dust storms in

Beijing. Geomorphology 87:365–377

Yang XP, Zhu BG, White PD (2007b) Provenance of aeolian

sediment in the Taklamakan desert of Western China, inferred

from REE and major-elemental data. Quat Int 175:71–85

Zdanowicz C, Hall G, Vaive J, Amelin Y, Percival J, Girard I,

Biscaye P, Bory A (2006) Asian dustfall in the St. Elias

mountains, Yukon, Canada. Geochim Cosmochim Acta

70:3493–3507

Zhang QG, Kang SC, Kaspari S, Li CL, Qin DH, Mayewski PA, Hou

S (2009) Rare earth elements in an ice core from Mt. Everest:

seasonal variations and potential sources. Atmospheric Res

94:300–312

Environ Earth Sci (2012) 67:1415–1421 1421

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.