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

Aeolian Research 35 (2018) 1–8

Contents lists available at ScienceDirect

Aeolian Research journal homepage: www.elsevier.com/locate/aeolia

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

T

⁎

Yue Lia,b,c, Yougui Songa,d, , Kathryn E. Fitzsimmonsb, Xiuling Chene, Qiansuo Wangf, Huanyu Sune, Zhiping Zhangg a

State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, Xi’an 710061, China Research Group for Terrestrial Palaeoclimates, Max Planck Institute for Chemistry, Mainz 55128, Germany c College of Earth Science, University of Chinese Academy of Sciences, Beijing 100049, China d Research Center for Ecology and Environment of Central Asia, Chinese Academy of Sciences, Urumqi 830011, China e Institute of Geography, Fujian Normal University, Fuzhou 350007, China f College of Resources and Environment, Linyi University, Linyi 276000, China g Environmental Monitoring Station of Ili Kazakh Autonomous Prefecture, Yining 835000, China b

A R T I C LE I N FO

A B S T R A C T

Keywords: Central Asia Loess Trace element analysis Grain-size end-member analysis Fingerprinting technique Provenance

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.

1. Introduction Loess deposits are widespread across the piedmonts of the high mountains of Arid Central Asia (ACA) (Fig. 1a) (Schaetzl et al., 2018). The thick loess deposits of ACA represent promising palaeoenvironmental archives, as has been demonstrated in other regions of the world (Muhs, 2013), but are as yet poorly explored beyond documentation of loess stratigraphy (Ding et al., 2002; Dodonov, 1991; Dodonov and Baiguzina, 1995; Dodonov et al., 1999; Frechen and Dodonov, 1998; Smalley et al., 2006b; Song et al., 2014; Zhou et al., 1995) or the investigation of relative climatic changes recorded in the loess (Dodonov et al., 2006, 1999; Feng et al., 2011b; Fitzsimmons et al., 2018; Song

et al., 2018a, 2015, 2018b; Yang et al., 2006; Youn et al., 2014; Zeng et al., 2018). The endorheic Ili River basin is located in eastern ACA and straddles southeast Kazakhstan and northwest China. The upper part of the catchment (referred to henceforth as the Ili valley) is an intermontane valley surrounded to the north, east and south by the Northern and Southern Tianshan Mountains (Fig. 1b); beyond the Kapshagay dam and north of the city of Almaty, the valley opens to the Kazakh Semirechiye and Gobi Desert steppe northwest (Fig. 1b), and the Ili River forms a large ephemeral delta draining into Lake Balkhash. The topography of the valley floor slopes westward; loess sediments are extensive along the valley slopes (Song et al., 2014).

⁎ Corresponding author at: State Key Laboratory of Loess and Quaternary Geology, Institute of Earth Environment, Chinese Academy of Sciences, No. 97 Yanxiang Road, Yanta, Xi’an 710061, China. E-mail address: [email protected] (Y. Song).

https://doi.org/10.1016/j.aeolia.2018.08.002 Received 10 December 2017; Received in revised form 16 August 2018; Accepted 16 August 2018 Available online 03 September 2018 1875-9637/ © 2018 Elsevier B.V. All rights reserved.

Aeolian Research 35 (2018) 1–8

Y. Li et al.

Fig. 1. Location of Arid Central Asia (a) and sampling sites within the Ili valley (b). (a) shows the loess areas in Central Asia. The four pale pink ellipses in (b) represent the geomorphic zones identified in this study. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

analysis, each with a mass of c. 500 g. Of these, 46 loess samples were collected at depths of 50–100 cm from the eastern part of the Ili Valley in China, and 35 desert sand samples were taken from the Kazakh deserts, 29 fluvial alluvium samples from riverbeds, 68 topsoil samples from piedmont slopes and alluvial-proluvial plains at depths of 2–5 cm (Fig. 1b). The main dust transport trajectories within the Ili valley derive from the western desert part of the Basin and from within the valley itself (Fig. S1). The complex topography of the intermontane valley can influence not only wind trajectories but also the associated transport of aeolian sediments on a local scale, as suggested by Nottebaum et al. (2015a), Qin et al. (2005) and Sprafke et al. (2018). Therefore it is important to consider the geomorphic variability within the valley and its potential influence on loess transport and accumulation. We classified the Ili Valley into four geomorphic zones (Fig. 1b; Supplementary Information): (a) valley plains; (b) higher altitude piedmont; (c) lower altitude piedmont; (d) intermontane valley. Our sampling strategy aimed to cover the full variety of sites and landform types across the Ili Valley.

Identifying the sources of loess, and the mechanisms of loess accumulation, is critical for clarifying the degree to which piedmont loess can be used as a climate archive, as well as for understanding the overall ecological impacts of the dust (Che and Li, 2013; Harrison et al., 2001). The Chinese Loess Plateau (CLP) has formed much of the focus for investigating loess provenance, based on geochemistry, mineralogy, meteorological observation and modeling (e.g. Chen et al. (2007); Hou et al. (2003); Li et al. (2007); Shi and Liu (2011); Sun (2002a); Sun et al. (2013); Xiao et al. (2012)). Although the scale and complexity of the CLP means that researchers are yet to resolve a formation model for the entire region (Chen and Li, 2011; Rao et al., 2006), it is generally agreed that the Alxa arid lands, fed by rivers in the Gobi Altay Mountains to the north and in the Qilian Mountains to the south (Li et al., 2011), provide most of the sediment for the late Pleistocene loess on the CLP (Chen and Li, 2011). A similar, theoretical model exists for the ACA piedmont loess (Dodonov and Baiguzina, 1995; Machalett et al., 2006b; Smalley et al., 2006b), whereby the vast desert steppe acts as an interim storage for fine-grained particles transported from the mountains by rivers, prior to aeolian transport back onto the lower mountain slopes. However, as yet the provenance of Central Asian loess is largely unknown and mostly inferred. The Ili Basin provides an interesting test case for a targeted study due to its more clearly defined catchment and relationship to the westerly wind trajectories relative to other locations along the ACA piedmonts. Studies so far support the desert-source-model, indicating that the dust is transported from the west of the Ili Basin into the intermontane valley (Li et al., 2012, 2015; Ye, 2000). Sun (2002b) speculated that the < 20 µm fraction of Ili loess is mainly derived from the Sary-Ishikotrau Desert in Kazakhstan, based on observations of geographic context and present-day air circulation systems rather than sedimentological data. Our study addresses this knowledge gap by providing the first dataset aimed at understanding loess provenance and formation in the eastern part of the Ili valley, within the Ili Kazakh Autonomous Prefecture of Xinjiang, China (Fig. 1). We identify distinct source areas based on grain-size characteristics and elemental geochemistry, and analyse the quantatitive contributions of different source areas to the Ili loess.

2.2. Identification of potential source areas We based our identification of potential source areas on airmass trajectories calculated from present-day data. We employed HYSPLIT (Draxler, 2011; Draxler et al., 1997) to conduct three-day airmass backward trajectories at different heights during the high-frequency dust storm months of March and April (details in Supplementary). The predominant air mass transport paths are oriented east–west, with additional localized mountain and valley breezes depending on topography (Fig. S1 in Supplementary). We assume that the regions through which air masses flow are the most likely source areas for dust entrainment, and identified areas for sediment sampling on that basis. 2.3. Grain size analysis Grain size measurements were carried out using published methods (Lu and An, 1997). Carbonates and organic matter were removed with 10 mL of 10% HCl and 10 mL of 10% H2O2, respectively. Deionized water was then added and the sample suspension was left for 12 h prior to pipetting, to remove acidic ions. After that, the samples were dispersed with 10 mL of 30% (NaPO3)6 and placed in an ultrasonic vibrator for 10 min. Finally, the treated samples were measured on a Malvern Mastersizer 2000 hosted at the State Key Laboratory of Loess

2. Material and methods 2.1 F. ield sampling We collected a total of 178 bulk sediment samples for provenance 2

Aeolian Research 35 (2018) 1–8

Y. Li et al.

and Quaternary Geology, Institute of Earth Environment, CAS. Particle size distribution was calculated for 100 grain-size classes within a measuring range of 0.02–2000 μm. The Mie theory was applied for particle size calculation based on the diffraction pattern, which predicts the light scattering behavior of all materials (Grehan and Gouesbet, 1979; Wiscombe, 1980). Replicate analyses indicated an analytical error of < 2% for the mean grain size. 2.4. Element geochemistry analysis Trace element concentrations were determined by an X-SERIES inductively-coupled plasma mass spectrometer (ICP-MS) at the Fujian Normal University. Homogenized bulk sample of ∼40 mg mass was weighed into a Teflon bomb and moistened with a few drops of ultrapure water. After adding mixed acid (MOS HCl, HNO3 and HF), the Teflon bomb was sealed and heated at 150 °C in Microwave Digestion System for 12 h. After cooling, HClO4 was added. Then the mixture was dried at 120 °C. The resultant salt was re-dissolved in HNO3 and high purity water, and then diluted to ∼40 mL for ICP-MS analysis. A standard loess sample (GBW 07454) for quality control was simultaneously analyzed and the deviation was < 10%. Replicate analyses of samples showed relative standard deviations (RSDs) of < 5%, and internal standard In and Re (5 µg L−1) values exhibited recovery rates of 90%–97%. 3. Results and discussion 3.1. Aeolian dust dynamics: loess grain-size frequency distributions The grain-size characteristics of loess from each zone are illustrated in Fig. S3. Grain size distributions have often been used to interpret wind strength on glacial/interglacial to millennial timescales (Ujvari et al. (2016) and references within). However, oversimplification based on interpretation of single statistical descriptors such as the mean, mode or median grain size, U-ratio, or grain-size index (GSI) as proxies for to wind speed should be avoided (Ujvari et al., 2016; Varga et al., 2017), since loess grain size is determined by a number of factors in addition to wind strength (Rousseau et al., 2011; Vandenberghe, 2013). In this study, we apply a recently developed approach (Paterson and Heslop, 2015) to separate grain-size curves into appropriate end members (Hateren et al., 2017; Varga et al., 2017). The new model enables more statistically and physically robust interpretations of parametric end-members than previous approaches (Sun et al., 2004), as suggested by Paterson and Heslop (2015). Three unmixed grain size end members (EM1, EM2 and EM3) were identified for each geomorphic zone (see Section 4 in Supplementary), and the relevant modal sizes are shown in Fig. 2. EM1 (0.4–200 µm) is relatively poorly sorted and positively skewed, with platykurtic kurtosis. By contrast, EM2 (1–100 µm or 2–200 µm) and EM3 (3–200 µm or 7–300 µm) are well sorted with leptokurtic kurtosis. These end members can be used to infer distinct atmospheric transportation mechanisms, models and travel distances (Ujvari et al., 2016). We interpret dominant aeolian processes and likely trajectories in the Ili Valley using the sediment groups identified in Vandenberghe (2013) and Sun et al. (2004). It is possible for the loess sediments to have multiple sources, as has been demonstrated by similar analyses from loess in other regions, such as northern and western China and the northeastern Tibetan Plateau (Nottebaum et al., 2015b; Sun et al., 2008). Furthermore, the end members identified may have different sources depending on the location of samples within the valley, implying that the relative contributions of these multiple sources may vary across the geomorphic zones identified across the Ili Valley. EM3 in Zone a yields a mode at 84.48 µm, which corresponds to a very fine sand-sized population which is a typical component of loess close to river terraces. It can only originate from proximal sources since the primary mode of transport for this grain size is saltation

Fig. 2. Grain-size partitioned components of loess deposits selected from the four zones.

(Vandenberghe, 2013). By contrast, EM3 in Zone c (a finer-grained mode at 53.30 µm) corresponds to the ‘subgroup 1.b.1’ of Vandenberghe (2013). EM3s in b and d (modes at 42.34 µm and 33.63 µm, respectively) are similar to ‘subgroup 1.b.2’. Both components 1.b.1 and 1.b.2 occur together but are distinct from one another. They are most likely transported during short-term, near-surface suspension by dust storms according to Sun et al. (2004). However, the wind strength indicated by component 1.b.2 is likely to be weaker than for 1.b.1 (Vandenberghe, 2013). In Zones b, c and d, EM2 (modes at 18.91 µm or 21.22 µm) corresponds to a fine silt-sized population, falling into ‘subgroup 1.c.1’ of Vandenberghe (2013). We propose that it derives from “non-dust storm processes” associated with surface winds, as suggested by Zhang et al. (1999). Qiang et al. (2010) found no correlation between the volume percentages of fine particles (clay/fine silt) and the wind strength, which implies that the fine silt grains are settled as floating dust (Lin et al., 2016) during low velocity wind conditions. Vegetation plays a major role in capturing dust (DiPietro et al., 2017; Qin et al., 2005; Schaetzl et al., 2018; Ujvari et al., 2016), although reduced vegetation cover coinciding with periods of loess accumulation probably facilitated the re-entrainment of this fine silt fraction (Klose and Shao, 2012; Macpherson et al., 2008; Qin et al., 2005; Sweeney and Mason, 2013), reducing the floating dust in Zone a which is the most downstream location within the Ili valley. EM1 (mode at 2.67–7.53 µm) corresponds to the clay fraction. This 3

Aeolian Research 35 (2018) 1–8

Y. Li et al.

component is also observed in loess from the Chinese Loess Plateau (CLP) (Sun et al., 2008) and the Danube River basin (Vriend et al., 2011). EM1 is mainly transported in long-term suspension by high-level westerly air flow according to Sun et al. (2004, 2008). Our results based on the new mixing model of grain-size distributions indicate predominantly local, but also distal sources for the Ili Valley loess. 3.2. Loess sediment sources: Trace element source discrimination analysis Geochemistry is a common tool for identifying loess provenance. Specific trace elements, major elements, as well as radiogenic neodymium and strontium isotopes, are typically used as indicators for loess provenance (Chen et al., 2007, 2017; Ferrat et al., 2011; Grousset and Biscaye, 2005; Guan et al., 2008; Hao et al., 2010; Hu and Yang, 2016; Li et al., 2009; Li et al., 2016; Muhs, 2018; Muhs and Budahn, 2006; Muhs et al., 2016). In particular, the elements Zr and Hf, which occur in substantial concentrations within the stable mineral zircon (Mclennan, 1993), are frequently used in sediment source discrimination. Zr and Hf concentrations can be indicative of zircon enrichment within loess, and therefore the degree of recycling within the sedimentary system. Our analyses of Zr concentration within the Ili valley loess indicate no significant sediment sorting downstream within the valley (correlation coefficient r = 0.28) and therefore limited genetic connection between loess sediments in different zones. The Ili valley loess was therefore most likely directly derived from primary alluvial deposits rather than from aeolian deposits such as dune sand, since the latter more frequently involves multiple cycles of recycling and reactivation (Xie et al., 2018). Zr/Hf at least partly reflect the compositional changes of zircon. Taylor (1965) showed that the rapid crystallization of volcanic rocks caused little difference in Zr/Hf ratios but the ratios increased from ultramafic to felsic plutonic rocks. Here, Zr/Hf ratios of different types of sediments decrease from Kazakhstan to China (Fig. 3a). This is because those sediments are derived from nearby bedrocks of Tianshan Mountains and the bedrocks contribute more felsic compositions to those sediments in the west of Tianshan (Fig. 4). This scenario provides us a good opportunity to distinguish these potential sources with Zr/Hf, as presented in Fig. 3a. In previous studies, Zr/Hf ratio allowed researchers to discriminate provenance linked to different parent rocks in different tectonic settings (Ahmad and Chandra, 2013; Hao et al., 2010; Hu and Yang, 2016; Sun, 2002a). Zr/Hf ratios of samples from potential source areas in this study suggest that local sediments within the Ili Valley could serve as nearby sources for the Ili loess (Fig. 3a). However, we cannot rule out the contributions of fine particles from Kazakhstan transported in longer-term suspension by high-level westerly air flow (Sun, 2002b), since Zr and Hf provide information on nearby sources. REE distribution patterns, and especially Heavy Rare Earth Elements (HREEs), within the Ili Valley loess differ from those found in rocks from the Upper Continental Crust (UCC) (Fig. 3b). This suggests either that REE distribution patterns are influenced by post-depositional pedogenesis, or that REE characteristics are inherited from the original source rocks. Rb is mainly concentrated in mica and K-feldspar, whereas Sr is mainly present in Ca-bearing minerals such as plagioclase and carbonate minerals. Rb generally remains immobile during weathering and Sr is characterized by high mobility during pedogenesis, because of the different weathering resistance of those minerals. Therefore, the Rb/Sr ratio can reflect intensities of weathering and pedogenesis in the loess-paleosol sequences (Buggle et al., 2011; Chen et al., 1999; Gallet et al., 1996; Tugulan et al., 2016), with higher values occurring under stronger weathering conditions (Nesbitt and Young, 1982). We found no relationship between Rb/Sr, ƩLREE/ ƩHREE, LaN/SmN, LaN/YbN and GdN/YbN (indexes indicative of the degree of fractionation between LREE (Light Rare Earth Element) and HREE) (Fig. 5), and can therefore exclude influences of post-depositional pedogenesis on the loess geochemistry. Our REE results indicate

Fig. 3. (a) Ratios of Zr and Hf in different sediments for identifying source areas of loess deposits. The grey shaded area indicates the variation range of Zr/Hf ratios in Ili loess samples. Y-axis just represents consecutive numbers, with no real significance; (b) Box plots illustrating the concentrations of determined rare earth elements (REE) normalized to UCC in the Ili loess.

minimal sedimentary recycling of the Ili loess and consequently a strong likelihood of local sourcing. Trace elements and REE have proven to be powerful tools for studying provenance of loess and dust (Muhs (2013) and references within). We also considered the relatively immobile trace elements Ba, Rb, Th, Zr, Hf, Y and Nb as more reliable fingerprinting tools in this study. However, since particle size also influences sediment source signatures (Feng et al., 2011a; Liang et al., 2013; Xie et al., 2014; Xiong et al., 2010), both grain size and loess geochemistry must be reconciled before interpretations can be made. We discriminated the trace elements independent of grain size. Table S1 shows the correlation coefficient matrix for the clay, fine silt, coarse silt and sand component contents, and selected immobile trace elements above. Rb, Eu, Th, Nb and Ba contents at the different levels, correlate positively with fine fractions and negatively with coarse fractions. Zr and Hf are mainly enriched in heavy minerals such as zircon, and are more likely to reflect nearby sources only. Therefore, we focused on Y and the REEs (except for Eu), including their ratios, to quantify source contributions, since these elements are hosted in a broad suite of minerals of varying densities, their ratio/pattern does not differentiate significantly between minerals (Bhatia, 1985; Bhatia and Crook, 1986; Chen and Li, 2011), and can reflect proximal and distal sources (Muhs and Budahn, 2006). 3.3. Spatial shifts in loess provenance across the Ili Valley: implications for loess formation Previous studies have asserted that bedrock from the Tianshan 4

Aeolian Research 35 (2018) 1–8

Y. Li et al.

Fig. 4. The geological map of the Tianshan orogen and adjacent region (modified after Institute of Geology of Chinese Academy of Geological Sciences (IGCAGS) (2006)). 9.2

4

Y = 0.61 * X + 7.52 R2=0.02

8.8

mass movement onto mountain piedmonts (Wright, 2001), which may then be deflated during cold and dry climate phases. We differentiated the sediments with respect to distal or proximal sourcing based on 1) the concentrations of trace elements selected from the two-step statistical analysis, and 2) the grain-size mixing model. The relative contributions of distally and proximally sourced material to the loess deposits at each geomorphic zone were then calculated (see Supplementary). The results are illustrated in Fig. 6. Our results suggest that local topsoil contributes a substantial component to the Ili loess (c. 54–90%) (Fig. 6). In particular, those soils developed on local piedmont slopes and alluvial-proluvial plains make the greatest contribution to loess deposits in the Ili Valley irrespective of location within the intermontane basin, reinforcing the hypothesis of mountain silt production for deflation described above. Relative contributions of local river alluvium increase downstream, but never dominate (Fig. 6). This result implies that fluvial abrasion, along with glacial grinding is also an efficient mechanism for producing silt particles (Muhs, 2013; Smith et al., 2002; Wright and Smith, 1993; Wright et al., 1998; Wright, 2001). The relative influence of fluvial processes increases downstream, including increased sediment input from tributaries of the Ili River. Rivers, therefore, also play a significant role in providing source material for loess deposits (Smalley et al., 2009), depending on geomorphic context and particularly context within a catchment. As yet we have no convincing data indicating that the fine-grained sediments generated in the Tianshan are initially transported to, and stored in, the great deserts of Central Asia (the Taukum, Muyunkum, Kysylkum, and Saryyesik Atyrau deserts) prior to transport back onto the piedmonts as dust (Forster and Heller, 1994; Machalett et al., 2006a; Smalley et al., 2006a; Youn et al., 2014). Distal desert sand and topsoil from Kazakhstan represents only a small proportion of the total loess in the eastern Ili Valley (Fig. 6). The dominance of proximally sourced material in the Ili loess (Fig. S5) appears to be in agreement with data from the CLP (Chen and Li, 2011) and challenges the assumption that desert basins act as significant interim storage reservoirs for Central Asian loess (Li et al., 2015; Sun, 2002b; Ye, 2000). The relatively enclosed Ili Valley morphology is the most likely barrier to

3.6

8.4

3.2

8

2.8

7.6

2.4

7.2

2

6.8

Y = -0.26 * X + 3.23 R2=0.01

1.6 0.1

0.2

0.3 0.4 Rb/Sr

0.5

0.6

0.1

10

4

9

3.6

8

3.2

7

2.8

6

2.4

Y = -0.36 * X + 7.84 R2=0.002

5

0.2

0.3 0.4 Rb/Sr

0.5

0.6

0.5

0.6

Y = -0.26* X + 3.23 R2=0.01

2

4

1.6 0.1

0.2

0.3 0.4 Rb/Sr

0.5

0.6

0.1

0.2

0.3 0.4 Rb/Sr

Fig. 5. Correlation analyses of Rb/Sr, ƩLREE/ƩHREE, LaN/SmN, LaN/YbN and GdN/YbN (the subscript N denotes chondrite-normalized values).

Mountains provided the primary source material for loess in this region (Machalett et al., 2006a; Smalley et al., 2006a; Youn et al., 2014), resulting from silt and sand production associated with mountain uplift (Machalett et al., 2006a). In addition, conditions in this region are particularly favorable for the production of silt particles due to frost/ mechanical weathering, land surface instability, and glacial grinding in the high-altitude mountains (Wright et al., 1998). The latter is considered one of the dominant processes for loess production (Smalley, 1995, 1966, 1990; Smalley et al., 2009; Smalley and Leach, 1978; Smalley and Vita-Finzi, 1968), leading to the hypothesis that dust flux increases throughout glacial periods (Harrison et al., 2001). Particles released by glacial grinding initially become available for transport by 5

Aeolian Research 35 (2018) 1–8

Y. Li et al.

Fig. 6. Map showing percentages of the relative contributions of proximal and distal source material to different loess depositional areas in the Ili valley.

significant input of far-sourced dust. Based on the grain-size distributions and provenancing for the Ili Valley, we can formulate a model for loess production and transport in this region. The low degree of sorting and proximal sourcing of loess in the Ili Valley implies silt production by glacial scouring and frost weathering in the Tianshan Mountains, followed by transported by spring meltwater onto glacio-fluvial plains in the piedmont (e.g. Smalley (1995); Wright et al. (1998)). Outwash deposits subsequently provide exposed, fine material for deflation, which are most likely then entrained and transported by wind storms associated with anticyclones associated with the Siberian High pressure system (Li et al., 2018). Minimal material is transported long distances from the Central Asian desert basins.

Zhiping Zhang: Investigation.

4. Summary and conclusions

Appendix A. Supplementary data

Here we identify the source of loess sediment in the Ili Valley, an intermontane basin in eastern Central Asia, and formulate a model for dust production and transport based on a mixing model of grain-size distribution and elemental geochemistry. Depending on the topographic context of localities within the valley, dominant aeolian transport processes for loess accumulation vary, from strong wind storms to continuous background, lower velocity wind transport. We found that local mountain (including glacial) and fluvial processes dominate the production of fine-grained silts which drape the Tianshan piedmonts in the valley. Likewise, most Ili Valley loess is locally derived, with minimal sedimentary recycling within the system. The relative contribution of river alluvium to the loess sediment increases downstream. Distal sediments from the Kazakh deserts are uncommon in the Ili Valley piedmont loess, indicating that the Central Asian deserts are unlikely to have formed interim storage reservoirs for the loess deposits downwind. Our findings elucidate the provenance and processes involved in accumulation of substantial loess deposits common across the Central Asian piedmont, and has implications for comparable settings worldwide.

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.aeolia.2018.08.002.

Acknowledgements This work was financially supported by the Strategic Priority Research Program of Chinese Academy of Sciences [grant number XDB26020403], National Basic Research and Development Program of China [grant number 2016YFA0601902], International partnership Program of the Chinese Academy of Science [grant number 132B61KYS20160002] and Natural Science Foundation of China [grant numbers 41572162, 41302149]. The authors would like to express the sincere appreciation to Prof. Hamid Gholami from Hormozgan University for his help in mathematical analysis.

References Ahmad, I., Chandra, R., 2013. Geochemistry of loess-paleosol sediments of Kashmir Valley, India: provenance and weathering. J. Asian Earth Sci. 66, 73–89. Bhatia, M.R., 1985. Rare earth element geochemistry of Australian Paleozoic graywackes and mudrocks: provenance and tectonic control. Sed. Geol. 45, 97–113. Bhatia, M.R., Crook, K.A.W., 1986. Trace-element characteristics of graywackes and tectonic setting discrimination of sedimentary basins. Contrib. Miner. Petrol. 92, 181–193. Buggle, B., Glaser, B., Hambach, U., Gerasimenko, N., Markovic, S., 2011. An evaluation of geochemical weathering indices in loess-paleosol studies. Quat. Int. 240, 12–21. Che, X.D., Li, G.J., 2013. Binary sources of loess on the Chinese Loess Plateau revealed by U-Pb ages of zircon. Quat. Res. 80, 545–551. Chen, J., Li, G.J., 2011. Geochemical studies on the source region of Asian dust. Sci. China-Earth Sci. 54, 1279–1301. Chen, J., An, Z.S., Wang, Y.J., Ji, J.F., Chen, Y., Lu, H.Y., 1999. Distribution of Rb and Sr in the Luochuan loess-paleosal sequence of China during the last 800 ka Implications for paleomonsoon variations. Sci. China Ser. D-Earth Sci. 42, 225–232. Chen, J., Li, G.J., Yang, J.D., Rao, W.B., Lu, H.Y., Balsam, W., Sun, Y.B., Ji, J.F., 2007. Nd and Sr isotopic characteristics of Chinese deserts: implications for the provenances of Asian dust. Geochim. Cosmochim. Acta 71, 3904–3914. Chen, X.L., Song, Y.G., Li, J.C., Fang, H., Li, Z.Z., Liu, X.M., Li, Y., Orozbaev, R., 2017. Size-differentiated REE characteristics and environmental significance of aeolian sediments in the Ili Basin of Xinjiang, NW China. J. Asian Earth Sci. 143, 30–38. Ding, Z.L., Ranov, V., Yang, S.L., Finaev, A., Han, J.M., Wang, G.A., 2002. The loess record in southern Tajikistan and correlation with Chinese loess. Earth Planet. Sci. Lett. 200, 387–400. DiPietro, L.M., Driese, S.G., Nelson, T.W., Harvill, J.L., 2017. Variations in late Quaternary wind intensity from grain-size partitioning of loess deposits in the Nenana River Valley, Alaska. Quat. Res. 87, 258–274. Dodonov, A.E., 1991. Loess of Central Asia. GeoJournal 24, 185–194. Dodonov, A.E., Baiguzina, L.L., 1995. Loess stratigraphy of Central Asia: palaeoclimatic and palaeoenvironmental aspects. Quat. Sci. Rev. 14, 707–720.

5. Credit author statement Yue Li: Writing – Original Draft, Conceptualization, Investigation, Formal Analysis, Visualization. Yougui Song: Resources, Writing – Review & Editing, Supervision. Kathryn E. Fitzsimmons: Writing – Review & Editing, Visualization, Formal Analysis. Xiuling Chen: Investigation, Data Curation, Formal Analysis. Qiansuo Wang: Investigation, Validation. Huanyu Sun: Investigation, Validation. 6

Aeolian Research 35 (2018) 1–8

Y. Li et al.

loess sequence from Remisowka (northern boundary of the Tien Shan Mountains, Kazakhstan)—Part I: Luminescence dating. Quat. Int. 152, 192–201. Macpherson, T., Nickling, W.G., Gillies, J.A., Etyemezian, V., 2008. Dust emissions from undisturbed and disturbed supply-limited desert surfaces. J. Geophys. Res.-Earth Surf. 113. Mclennan, S.M., 1993. Weathering and global denudation. J. Geol. 101, 295–303. Muhs, D.R., 2013. The geologic records of dust in the Quaternary. Aeolian Res. 9, 3–48. Muhs, D.R., 2018. The geochemistry of loess: Asian and North American deposits compared. J. Asian Earth Sci. 155, 81–115. Muhs, D.R., Budahn, J.R., 2006. Geochemical evidence for the origin of late Quaternary loess in central Alaska. Can. J. Earth Sci. 43, 323–337. Muhs, D.R., Budahn, J.R., Skipp, G.L., McGeehin, J.P., 2016. Geochemical evidence for seasonal controls on the transportation of Holocene loess, Matanuska Valley, southern Alaska, USA. Aeolian Res. 21, 61–73. Nesbitt, H.W., Young, G.M., 1982. Early proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature 299, 715–717. Nottebaum, V., Stauch, G., Hartmann, K., Zhang, J., Lehmkuhl, F., 2015b. Unmixed loess grain size populations along the northern Qilian Shan (China): relationships between geomorphologic, sedimentologic and climatic controls. Quat. Int. 372, 151–166. Nottebaum, V., Lehmkuhl, F., Stauch, G., Hartmann, K., Wünnemann, B., Schimpf, S., Lu, H., 2015a. Regional grain size variations in aeolian sediments along the transition between Tibetan highlands and north-western Chinese deserts – the influence of geomorphological settings on aeolian transport pathways. Earth Surf. Proc. Land. 39, 1960–1978. Paterson, G.A., Heslop, D., 2015. New methods for unmixing sediment grain size data. Geochem. Geophys. Geosyst. 16, 4494–4506. Qiang, M., Lang, L., Wang, Z., 2010. Do fine-grained components of loess indicate westerlies: insights from observations of dust storm deposits at Lenghu (Qaidam Basin, China). J. Arid Environ. 74, 1232–1239. Qin, X.G., Cai, B.G., Liu, T.S., 2005. Loess record of the aerodynamic environment in the East Asia monsoon area since 60,000 years before present. Journal of Geophysical Research-Solid. Earth 110. Rao, W.B., Yang, J.D., Chen, J., Li, G.J., 2006. Sr-Nd isotope geochemistry of eolian dust of the arid-semiarid areas in China: Implications for loess provenance and monsoon evolution. Chin. Sci. Bull. 51, 1401–1412. Rousseau, D.D., Antoine, P., Gerasimenko, N., Sima, A., Fuchs, M., Hatte, C., Moine, O., Zoeller, L., 2011. North Atlantic abrupt climatic events of the last glacial period recorded in Ukrainian loess deposits. Clim. Past 7, 221–234. Schaetzl, R.J., Bettis, E.A., Crouvic, O., Fitzsimmons, K.E., Grimley, D.A., Hambach, U., Lehmkuhl, F., Markovic, S.B., Mason, J.A., Owczarek, P., Roberts, H.M., Rousseau, D.D., Stevens, T., Vandenberghe, J., Zarate, M., Veres, D., Yang, S.L., Zech, M., Conroy, J.L., Dave, A.K., Faust, D., Hao, Q.Z., Obreht, I., Prud'homme, C., Smalley, I., Tripaldi, A., Zeeden, C., Zech, R., Issue, Special, 2018. Approaches and challenges to the study of loess introduction to the Loess-Fest. Quat. Res. 89, 563–618. Shi, Z.G., Liu, X.D., 2011. Distinguishing the provenance of fine-grained eolian dust over the Chinese Loess Plateau from a modelling perspective. Tellus Ser. B-Chem. Phys. Meteorol. 63, 959–970. Smalley, I., 1966. The properties of glacial loess and the formation of loess deposits. J. Sediment. Res. 36, 669–676. Smalley, I., 1990. Possible formation mechanisms for the modal coarse-silt quartz particles in loess deposits. Quat. Int. 7–8, 23–27. Smalley, I., 1995. Making the material: The formation of silt-sized primary mineral particles for loess deposits. Quat. Sci. Rev. 14, 645–651. 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. Sed. Geol. 21, 1–26. Smalley, I., Mavlyanova, N., Rakhmatullaev, K.L., Shermatov, M.S., Machalett, B., Dhand, K.H., Jefferson, I., 2006. The formation of loess deposits in the Tashkent region and parts of Central Asia; and problems with irrigation, hydrocollapse and soil erosion. Quat. Int. 152, 59–69. Smalley, I., O’Hara-Dhand, K., Wint, J., Machalett, B., Jary, Z., Jefferson, I., 2009. Rivers and loess: the significance of long river transportation in the complex event-sequence approach to loess deposit formation. Quat. Int. 198, 7–18. Smalley, I.J., Vita-Finzi, C., 1968. The Formation of fine particles in sandy deserts and the nature of desert loess. J. Sediment. Res. 38, 766–774. Smith, B.J., Wright, J.S., Whalley, W.B., 2002. Sources of non-glacial, loess-size quartz silt and the origins of “desert loess”. Earth Sci. Rev. 59, 1–26. Song, Y.G., Chen, X.L., Qian, L.B., Li, C.X., Li, Y., Li, X.X., Chang, H., An, Z.S., 2014. Distribution and composition of loess sediments in the Ili Basin, Central Asia. Quat. Int. 334, 61–73. Song, Y.G., Lai, Z.P., Li, Y., Chen, T., Wang, Y.X., 2015. Comparison between luminescence and radiocarbon dating of late Quaternary loess from the Ili Basin in Central Asia. Quat. Geochronol. 30, 405–410. Song, Y.G., Luo, D., Du, J.H., Kang, S.G., Cheng, P., Fu, C.F., Guo, X.H., 2018a. Radiometric dating of late Quaternary loess in the northern piedmont of South Tianshan Mountains: implications for reliable dating. Geol. J. 53, 417–426. Song, Y.G., Zeng, M.X., Chen, X.L., Li, Y., Chang, H., An, Z.S., Guo, X.H., 2018b. Abrupt climatic events recorded by the Ili loess during the last glaciation in Central Asia: evidence from grain-size and minerals. J. Asian Earth Sci. 155, 58–67. Sprafke, T., Fitzsimmons, K.E., Grützner, C., Elliot, A., Marquer, L., Nigmatova, S., 2018. Reevaluation of Late Pleistocene loess profiles at Remizovka (Kazakhstan) indicates the significance of topography in evaluating terrestrial paleoclimate records. Quat. Res. 89, 674–690. Sun, J.M., 2002a. Provenance of loess material and formation of loess deposits on the Chinese Loess Plateau. Earth Planet. Sci. Lett. 203, 845–859. Sun, J.M., 2002b. Source regions and formation of the Loess sediments on the high mountain regions of northwestern China. Quat. Res. 58, 341–351.

Dodonov, A.E., Shackleton, N., Zhou, L.P., Lomov, S.P., Finaev, A.F., 1999. Quaternary loess-paleosol stratigraphy of Central Asia: geochronology, correlation, and evolution of paleoenvironments. Stratigr. Geol. Correl. 7, 581–593. Dodonov, A.E., Sadchikova, T.A., Sedov, S.N., Simakova, A.N., Zhou, L.P., 2006. Multidisciplinary approach for paleoenvironmental reconstruction in loess-paleosol studies of the Darai Kalon section, Southern Tajikistan. Quat. Int. 152, 48–58. Draxler, R.R., Hess, G.D., Draxler, R.R., 1997. Description of the HYSPLIT_4 modeling system, National Oceanic & Atmospheric Administration Technical Memorandum Erl Arl, pp. 197-199. Draxler, R.R., 2011. HYSPLIT (HYbrid Single-Particle Lagrangian Integrated Trajectory) Model access via NOAA ARL READY Website. Feng, J.L., Hu, Z.G., Ju, J.T., Zhu, L.P., 2011a. Variations in trace element (including rare earth element) concentrations with grain sizes in loess and their implications for tracing the provenance of eolian deposits. Quat. Int. 236, 116–126. Feng, Z.D., Ran, M., Yang, Q.L., Zhai, X.W., Wang, W., Zhang, X.S., Huang, C.Q., 2011b. Stratigraphies and chronologies of late Quaternary loess-paleosol sequences in the core area of the central Asian arid zone. Quat. Int. 240, 156–166. Ferrat, M., Weiss, D.J., Strekopytov, S., Dong, S.F., Chen, H.Y., Najorka, J., Sun, Y.B., Gupta, S., Tada, R., Sinha, R., 2011. Improved provenance tracing of Asian dust sources using rare earth elements and selected trace elements for palaeomonsoon studies on the eastern Tibetan Plateau. Geochim. Cosmochim. Acta 75, 6374–6399. Fitzsimmons, K.E., Sprafke, T., Zielhofer, C., Gunter, C., Deom, J.M., Sala, R., Iovita, R., 2018. Loess accumulation in the Tian Shan piedmont: implications for palaeoenvironmental change in arid Central Asia. Quat. Int. 469, 30–43. Forster, T., Heller, F., 1994. Loess deposits from the Tajik depression (Central Asia): magnetic properties and paleoclimate. Earth Planet. Sci. Lett. 128, 501–512. Frechen, M., Dodonov, A.E., 1998. Loess chronology of the Middle and Upper Pleistocene in Tadjikistan. Geol. Rundsch. 87, 2–20. Gallet, S., Jahn, B.M., Torii, M., 1996. Geochemical characterization of the Luochuan loess-paleosol sequence, China, and paleoclimatic implications. Chem. Geol. 133, 67–88. Grehan, G., Gouesbet, G., 1979. Mie theory calculations – new progress, with emphasis on particle sizing. Appl. Opt. 18, 3489–3493. Grousset, F.E., Biscaye, P.E., 2005. Tracing dust sources and transport patterns using Sr, Nd and Pb isotopes. Chem. Geol. 222, 149–167. Guan, Q.Y., Pan, B.T., Gao, H.S., Li, N., Zhang, H., Wang, J.P., 2008. Geochemical evidence of the Chinese loess provenance during the Late Pleistocene. Palaeogeogr. Palaeoclimatol. Palaeoecol. 270, 53–58. Hao, Q.Z., Guo, Z.T., Qiao, Y.S., Xu, B., Oldfield, F., 2010. Geochemical evidence for the provenance of middle Pleistocene loess deposits in southern China. Quat. Sci. Rev. 29, 3317–3326. Harrison, S.P., Kohfeld, K.E., Roelandt, C., Claquin, T., 2001. The role of dust in climate changes today, at the last glacial maximum and in the future. Earth Sci. Rev. 54, 43–80. Hateren, J.A.V., Prins, M.A., Balen, R.T.V., 2017. On the genetically meaningful decomposition of grain-size distributions: a comparison of different end-member modelling algorithms. Sediment. Geol. Hou, S.S., Yang, S.L., Sun, J.M., Ding, Z.L., 2003. Oxygen isotope compositions of quartz grains (4–16 mu m) from Chinese eolian deposits and their implications for provenance. Sci. China Ser. D-Earth Sci. 46, 1003–1011. Hu, F.G., Yang, X.P., 2016. Geochemical and geomorphological evidence for the provenance of aeolian deposits in the Badain Jaran Desert, northwestern China. Quat. Sci. Rev. 131, 179–192. Institute of Geology of Chinese Academy of Geological Sciences (IGCAGS), 2006. 1:2,500,000 Geological Map of the Western China and Adjacent Regions. Geological Publishing House, Beijing (in Chinese). Klose, M., Shao, Y., 2012. Stochastic parameterization of dust emission and application to convective atmospheric conditions. Atmos. Chem. Phys. 12, 7309–7320. Li, G., Chen, J., Chen, Y., Yang, J., Ji, J., Liu, L., 2007. Dolomite as a tracer for the source regions of Asian dust. J. Geophys. Res.-Atmos. 112. Li, G.J., Chen, J., Ji, J.F., Yang, J.D., Conway, T.M., 2009. Natural and anthropogenic sources of East Asian dust. Geology 37, 727–730. Li, G.J., Pettke, T., Chen, J., 2011. Increasing Nd isotopic ratio of Asian dust indicates progressive uplift of the north Tibetan Plateau since the middle Miocene. Geology 39, 199–202. Li, C.X., Song, Y.G., Wang, Y.M., 2012. Distribution, age and dust sources of loess in the Ili basin. Earth Environment 40, 314–320 (in China). Li, Y., Song, Y.G., Yan, L.B., Chen, T., An, Z.S., 2015. Timing and spatial distribution of loess in Xinjiang, NW China. PLoS One 10, e0125492. Li, Y., Song, Y.G., Chen, X.L., Li, J.C., Mamadjanov, Y., Aminov, J., 2016. Geochemical composition of Tajikistan loess and its provenance implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 446, 186–194. Li, Y., Song, Y., Fitzsimmons, K.E., Chang, H., Orozbaev, R., Li, X., 2018. Eolian dust dispersal patterns since the last glacial period in eastern Central Asia: insights from a loess-paleosol sequence in the Ili Basin. Clim. Past 14, 271–286. Liang, L.J., Sun, Y.B., Beets, C.J., Prins, M.A., Wu, F., Vandenberghe, J., 2013. Impacts of grain size sorting and chemical weathering on the geochemistry of Jingyuan loess in the northwestern Chinese Loess Plateau. J. Asian Earth Sci. 69, 177–184. Lin, Y.C., Mu, G.J., Xu, L.S., Zhao, X., 2016. The origin of bimodal grain-size distribution for aeolian deposits. Aeolian Res. 20, 80–88. Lu, H.Y., An, Z.S., 1997. Pretreatment methods in loess–palaeosol granulometry. Chin. Sci. Bull. 42, 237–240. Machalett, B., Frechen, M., Hambach, U., Oches, E.A., Zoller, L., Markovic, S.B., 2006b. The loess sequence from Remisowka (northern boundary of the Tien Shan Mountains, Kazakhstan)-Part I: luminescence dating. Quat. Int. 152–153, 192–201. Machalett, B., Frechen, M., Hambach, U., Oches, E., Zöller, L., Marković, S., 2006a. The

7

Aeolian Research 35 (2018) 1–8

Y. Li et al.

Wright, J., Smith, B., 1993. Fluvial comminution and the production of loess-sized quartz silt: a simulation study. Geogr. Ann. 75, 25–34. Wright, J., Smith, B., Whalley, B., 1998. Mechanisms of loess-sized quartz silt production and their relative effectiveness: laboratory simulations. Geomorphology 23, 15–34. Xiao, G.Q., Zong, K.Q., Li, G.J., Hu, Z.C., Dupont-Nivet, G., Peng, S.Z., Zhang, K.X., 2012. Spatial and glacial-interglacial variations in provenance of the Chinese Loess Plateau. Geophys. Res. Lett. 39. Xie, Y.Y., Meng, J., Guo, L.F., 2014. REE geochemistry of modern eolian dust deposits in Harbin city, Heilongjiang province, China: implications for provenance. Catena 123, 70–78. Xie, Y.Y., Yuan, F., Zhan, T., Kang, C.G., Chi, Y.P., Ma, Y.F., 2018. Geochemistry of loess deposits in northeastern China: constraint on provenance and implication for disappearance of the large Songliao palaeolake. J. Geol. Soc. 175, 146–162. Xiong, S.F., Ding, Z.L., Zhu, Y.J., Zhou, R., Lu, H.J., 2010. A similar to 6 Ma chemical weathering history, the grain size dependence of chemical weathering intensity, and its implications for provenance change of the Chinese loess-red clay deposit. Quat. Sci. Rev. 29, 1911–1922. Yang, S.L., Ding, F., Ding, Z.L., 2006. Pleistocene chemical weathering history of Asian arid and semi-arid regions recorded in loess deposits of China and Tajikistan. Geochim. Cosmochim. Acta 70, 1695–1709. Ye, W., 2000. The mineral characteristics of loess and depositional environment in Yili area, Xinjinag. Arid Zone Res. 17, 1–10 (in Chinese). Youn, J.H., Seong, Y.B., Choi, J.H., Abdrakhmatov, K., Ormukov, C., 2014. Loess deposits in the northern Kyrgyz Tien Shan: implications for the paleoclimate reconstruction during the Late Quaternary. Catena 117, 81–93. Zeng, M.X., Song, Y.G., Li, Y., Fu, C.F., Qiang, X.K., Chang, H., Zhu, L.D., Zhang, Z.P., Cheng, L.Q., 2018. The relationship between environmental factors and magnetic susceptibility in the Ili loess, Tianshan Mountains, Central Asia. Geol. Journal. Zhang, X.Y., Arimoto, R., An, Z.S., 1999. Glacial and interglacial patterns for Asian dust transport. Quat. Sci. Rev. 18, 811–819. Zhou, L.P., Dodonov, A.E., Shackleton, N.J., 1995. Thermoluminescence dating of the Orkutsay loess section in Tashkent Region, Uzbekistan, Central Asia. Quat. Sci. Rev. 14, 721–730.

Sun, D.H., Bloemendal, J., Rea, D.K., An, Z.S., Vandenberghe, J., Lu, H.Y., Su, R.X., Liu, T.S., 2004. Bimodal grain-size distribution of Chinese loess, and its palaeoclimatic implications. Catena 55, 325–340. Sun, Y.B., Chen, H.Y., Tada, R., Weiss, D., Lin, M., Toyoda, S., Yan, Y., Isozaki, Y., 2013. ESR signal intensity and crystallinity of quartz from Gobi and sandy deserts in East Asia and implication for tracing Asian dust provenance. Geochem. Geophys. Geosyst. 14, 2615–2627. Sun, D.H., Su, R.X., Bloemendal, J., Lu, H.Y., 2008. Grain-size and accumulation rate records from Late Cenozoic aeolian sequences in northern China: implications for variations in the East Asian winter monsoon and westerly atmospheric circulation. Palaeogeogr. Palaeoclimatol. Palaeoecol. 264, 39–53. Sweeney, M.R., Mason, J.A., 2013. Mechanisms of dust emission from Pleistocene loess deposits, Nebraska, USA. J. Geophys. Res.-Earth Surf. 118, 1460–1471. Taylor, S.R., 1965. The application of trace element data to problems in petrology. Phys. Chem. Earth. 6, 133–213. Tugulan, L.C., Duliu, O.G., Bojar, A.V., Dumitras, D., Zinicovskaia, I., Culicov, O.A., Frontasyeva, M.V., 2016. On the geochemistry of the Late Quaternary loess deposits of Dobrogea (Romania). Quat. Int. 399, 100–110. Ujvari, G., Kok, J.F., Varga, G., Kovacs, J., 2016. The physics of wind-blown loess: implications for grain size proxy interpretations in Quaternary paleoclimate studies. Earth Sci. Rev. 154, 247–278. Vandenberghe, J., 2013. Grain size of fine-grained windblown sediment: a powerful proxy for process identification. Earth Sci. Rev. 121, 18–30. Varga, G., Újvári, G., Kovács, J., 2017. Interpretation of sedimentary (sub)populations extracted from grain size distributions of Central European loess-paleosol series. Quat. Int in press. Vriend, M., Prins, M.A., Buylaert, J.P., Vandenberghe, J., Lu, H.Y., 2011. Contrasting dust supply patterns across the north-western Chinese Loess Plateau during the last glacial-interglacial cycle. Quat. Int. 240, 167–180. Wiscombe, W.J., 1980. Improved Mie scattering algorithms. Appl. Opt. 19, 1505–1509. Wright, J.S., 2001. “Desert” loess versus “glacial” loess: quartz silt formation, source areas and sediment pathways in the formation of loess deposits. Geomorphology 36, 231–256.

8