Loess deposits in the northern Kyrgyz Tien Shan: Implications for the paleoclimate reconstruction during the Late Quaternary

Catena 117 (2014) 81–93

Contents lists available at ScienceDirect

Catena journal homepage: www.elsevier.com/locate/catena

Loess deposits in the northern Kyrgyz Tien Shan: Implications for the paleoclimate reconstruction during the Late Quaternary J.H. Youn a,b, Y.B. Seong a,⁎, J.H. Choi b,⁎⁎, K. Abdrakhmatov c, C. Ormukov c a b c

Department of Geography, Korea University, Seoul 136-701, South Korea Division of Earth and Environmental Science, Korea Basic Science Institute, Chungbuk 363-883, South Korea Kyrgyz Institute of Seismology, Kyrgyz Academy of Sciences, Bishkek 720060, Kyrgyzstan

a r t i c l e

i n f o

Article history: Received 5 April 2013 Received in revised form 23 September 2013 Accepted 23 September 2013 Keywords: Loess Optically stimulated luminescence (OSL) dating Central Asia Siberian high pressure (SHP) Mid-latitude westerlies

a b s t r a c t Loess deposits on the northern slopes of the Kyrgyz Tien Shan were examined. Their particle size characteristics show silt size dominancy (N 80%) with minor contribution from sand (12%) and clay (7%). The loess was dated using optically stimulated luminescence (OSL) and radiocarbon methods to define the timing of deposition. The OSL ages of fine and coarse quartz fractions were consistent with each other within 2σ uncertainty level, except several samples deposited during MIS 2. Based on the OSL ages, four major loess depositional periods are recognized in the northern Kyrgyz Tien Shan during the Late Quaternary: the Holocene, MIS 2, MIS 3, and MIS 4. The rate of dust accumulation in the northern Tien Shan during MIS 2 was greater than that during MIS 3 or MIS 4. This implies that cold–dry conditions varied significantly during the Late Quaternary in the study area. The accumulation patterns of the Kyrgyz loess deposits in the northern Tien Shan are closely related to climate fluctuations during the Late Quaternary, influenced by changes in the mid-latitude westerlies, Asian summer monsoons, and Siberian High Pressure (SHP) systems, during which there was no significant cessation of deposition. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The Quaternary was characterized by remarkable large-amplitude cyclic variations associated with climatic fluctuations, sea level changes, and the advance and retreat of glaciers. Understanding of Quaternary environmental changes provides information on climate variability and predictions about future climate changes. Numerical dating of Quaternary terrestrial fine sediments, such as loess, has contributed to long-term archival records of environmental change, and to an understanding of how local environments have changed in response to climate shifts occurring on global and hemispherical scales. Continental loess deposits provide one of the most important archives of data on climate fluctuations during the Quaternary; they are especially sensitive indicators of dry–windy conditions associated with glacial activity (Burbank and Li, 1985; Derbyshire et al., 1997; Guo et al., 2002; Kapp et al., 2011; Pye, 1995; Sun et al., 2008). Thick and extensive loess deposits are present in a wide band across Asia, extending from northwestern Europe to Central Asia and China. The Chinese Loess Plateau (CLP) is known for its long succession of loess and paleosols, which provides first-order proxy records of alternating glacial and interglacial cycles related to climate change. Northern Central Asian regions, located north of the Tien Shan, also preserve

⁎ Corresponding author. Tel.: +82 2 3290 2367; fax: +82 2 3290 2360. ⁎⁎ Corresponding author. Tel.: +82 43 240 5333; fax: +82 43 240 5319. E-mail addresses: [email protected] (Y.B. Seong), [email protected] (J.H. Choi). 0341-8162/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.catena.2013.09.007

multiple levels of thick and extensive loess deposits. However, except for a handful of pioneering research studies (Bronger et al., 1995; Ding et al., 2002; Dodonov and Baiguzina, 1995; Forster and Heller, 1994; Frechen and Dodonov, 1998; Machalett et al., 2006; Shackleton et al., 1995), relatively little attention has been given to the quantitative study of these loess deposits. Despite the lack of detailed researches, loess in northern Central Asia is likely to be useful for reconstructing past climate changes, based on correlations with loess records on the adjacent CLP and nearby regions. Some attempts to compare the loess records of Central Asia with those of the CLP have suggested that both regions share similar loess–paleosol sequence cyclicities, as based on global ice volume correlations. The climatic cycles preserved in the loess of northern Central Asia, such as the Chashmanigar loess–soil sequences in southern Tajikistan and the Remisowka loess in Kazahstan, are well correlated with CLP records (Ding et al., 2002; Dodonov and Baiguzina, 1995; Machalett et al., 2006). The data from both regions show that loess deposits accumulated during the cold–dry glacial period, and that they have been altered to paleosols through pedogenic processes occurring during the warm–humid period when rates of loess accumulation were reduced. In recent years, optically stimulated luminescence (OSL) dating has been applied extensively to loess deposits throughout the world (Küster et al., 2006; Machalett et al., 2006; Olley, 2004; Roberts, 2008; Roberts et al., 1999; Stevens et al., 2006, and more references therein). In this study, we also applied OSL dating method to establish chronology of loess deposition in the northern part of the Kyrgyz Range in the Kyrgyz Tien Shan.

82

J.H. Youn et al. / Catena 117 (2014) 81–93

At present, the study area in the northern part of the Kyrgyz Range is dominated by mid-latitude westerlies and Siberian high-pressure (SHP) systems, both of which contribute to an extreme continental climate. The influence of these systems, together with the Asian monsoon, is likely to have varied throughout the Quaternary (Kukla, 1987; Pye, 1995; Shackleton et al., 1995). In this study, we examine the past climate changes of the late Quaternary in the northern part of the Kyrgyz Tien Shan using OSL dating and other sedimentary characteristics of loess deposits.

2. Study area The Tien Shan, located at approximately 40°–45°N, 67°–95°E, extends in an east–west direction across Central Asia. They are ~2000 km long and ~400 km wide, and include ranges with several high peaks (elevations of N 6000 m a.s.l). The mountain belt borders the Tarim Basin and the Taklamakan Desert to the south and plains dotted with mid-latitude continental deserts such as the Karakum, Kysylkurm, and Muyunkum deserts, to the north (Fig. 1). The belts

Fig. 1. Distribution of loess across Central Asia. Loess deposits of Central Asia are found on the piedmont of high mountains near the great deserts. The Siberian High Pressure (SHP) was strengthened in northern Central Asia during the cold–dry glacier period, whereas mid-latitude westerlies were more influential in the northern Tien Shan during the warm–humid interglacial periods. The westerlies tend to transport moisture into northern Central Asia as they pass over the Caspian and Aral seas, the North Atlantic, and the Mediterranean Sea. The diagram is modified from Dodonov and Baiguzina (1995).

J.H. Youn et al. / Catena 117 (2014) 81–93

can be divided into two parts, consisting of eastern parts in China and western parts in the Kyrgyz Republic (Glorie et al., 2010). The Kyrgyz Tien Shan can be further sub-classified into three sections, representing northern, central, and southern sections of the ranges. The southern border of the northern Kyrgyz Tien Shan is defined tectonically by a fault zone known as the “Nikolaev Line” (NL) (Burtman et al., 1996). The northern Kyrgyz Tien Shan occurs in the area 42°N and 72°–74°E, and is oriented in an east–west direction. The sampling site for this study (at 42°N, 74°E) is situated on the northern footslope of the northern Kyrgyz Tien Shan. Central Asia is an arid and semi-arid region, climatologically dominated by mid-latitude westerlies and the SHP, that play pivotal roles in the region by affecting air-current moisture availability, with a minor influence caused by penetration of the Asian summer monsoon (ASM) (Aizen et al., 1996; Cheng et al., 2012). The mean annual precipitation in the northern Krygyz Tien Shan (due to orographic effects) is 200–600 mm; the maximum annual precipitation is 1500 mm (Aizen et al., 1996). Precipitation is greatest during the spring; the peak precipitation is in May, at which time mid-latitude westerlies carry warm and humid air masses from the Caspian and Aral seas, originating in North Atlantic and Mediterranean regions to the west. Precipitation during the period of July–December is low (with slight increases in autumn), under the influence of the SHP, whose strength becomes more elevated during the winter (Aizen et al., 1996; Beer et al., 2008; Fig. 2). Precipitation levels decrease with elevation up to 60% of maximum monthly values and snowfall occurs at elevations greater than 3000 m a.s.l in the high mountain regions, which contributes to the formation of mountain glaciers in these areas. Glacial runoff contributes to the flow of most rivers during the warm season. The volume of glacial runoff begins to increase from late May to the middle of June, sometimes reaching up to 40–60% of the total flow during the early summer (Williams and Konovalov, 2008). Western cyclones and warm air masses affect the Tien Shan region during the early summer, resulting in flash flooding events on rivers (Aizen et al., 1996; Bugaev et al., 1957). The majority of the surface water on the slopes of the Tien Shan, originating from rainfall and glacier meltwater (during the warm season), flows into intermittent channels during the pluvial period and in the summer season. Some surface water reaches the playa on the foreland basin of the mountain system, forming pluvial lakes, which are temporary reservoirs of dust particles in this arid region. The annual temperature and precipitation in the northern Kyrgyz Tien Shan vary as a function of altitude and topography. The Kyrgyz loess deposits investigated in this study are located south of Bishkek, the capital city of the Kyrgyz Republic. In the region of Bishkek, the annual precipitation is ~363 mm, the average elevation is ~774 m a.s.l., and the monthly mean temperatures in January and July are −4 °C and 25 °C, respectively. The vegetation in the northern Tien Shan shows altitudinal zonation, from lowland grasses and deciduous trees in lowland areas, to piney woods at mid-elevations, to alpine grasses and meadows at elevations above 3000 m a.s.l (Fig. 3). Broad meadows of wheat are cultivated on loess terraces. In addition, forests of walnut and fruit trees (apricot, apple, and pear) cover broad areas of the valley floor. The walnut trees are deciduous and shed their leaves during the dry season. The varieties of walnut and fruit trees are extensively cultivated in the Kyrgyz Republic, covering an area of 2300 km2; the populations may serve as a source of hybrid varieties suitable for conservation, landscaping, and breeding (Beer et al., 2008). 3. Geomorphic setting and loess characteristics In general, the so-called ‘piedmont loess’ deposits of northern Central Asia are distributed along the edges of high-elevation glaciated mountain belts; they occur mostly in periglacial regions with extreme diurnal and/or seasonal climate changes. The accumulation of piedmont loess deposits in Central Asia is associated with areas geologically

83

uplifted since the Middle Tertiary; uplift of the Tien Shan has resulted in the production of a large amount of silt and sand particles in this region (Machalett et al., 2006). Most of loess deposits caps on the bedrock of Paleozoic metamorphic and Tertiary sedimentary rocks and glacial deposits (Thompson et al., 2002). The climatic and tectonic conditions in the region are particularly favorable for the production of silt particles, on account of frost/mechanical weathering, instability, and glacial grinding by high-altitude mountain glaciers. The Tien Shan is the primary source area for the loess materials in the region, and the great deserts to the north and northwest of the Krygyz Tien Shan (the Taukum, Muyunkum, Kysylkum, and Saryyesik Atyrau deserts) are secondary source areas that serve as interim storage reservoirs for the silt fraction of the loess deposits (Forster and Heller, 1994; Figs. 1 and 4). Most of the materials produced by glacial scouring and frost weathering in the high mountains are transported downward by glacial meltwater during warm seasons, and frequently occurring flash floods in the spring and summer produce mega fans and other fluvial landforms on the footslopes of the mountains (Machalett et al., 2006). The fine materials, such as the clay and silt fractions, tend to be transported further away from the mountains and deposited in deserts and lakes (Fig. 4). Frequent dust storms occur in the deserts and dried lakes due to the influx of anticyclones in the area, especially during winter. It has been reported that the mean annual frequency of dust storms in the center of the Karakum deserts over a 60-year period (1936–1995) is 62 days (Orlovsky et al., 2005). The dust storms are prompted by intrusions of western or southwestern anticyclones into northern Central Asia in the spring and summer (Orlovsky et al., 2005). The fine materials stored in deserts and dried lakes in deflation zones are lifted by wind storms and transported by high-altitude winds during dry-windy winters or glacial periods (Goudie, 2009; Kapp et al., 2011; Li et al., 2009; Lu and Sun, 2000; Pye, 1995; Fig. 4). The aeolian dusts are finally deposited and accumulated vertically in piedmont zones along the footslopes of high mountains, forming thick and extensive loess deposits. We investigated loess successions in the piedmont zone. The study section, which is situated to the south of the capital town Bishkek, located in the northern part of the Tien Shan, consists of two parts, referred to as section-1 (42°42′14.99″N, 74°46′51.00″E; elevation, 1432 m a.s.l) and section-2 (42°42′08.93″N, 74°46′47.84″E; elevation, 1417 m a.s.l). We excavated section-1 first, and then added section-2 because we were not able to dig to the bottom of the loess deposit in section-1. Section-2 is located ~20 m horizontally headward (towards the mountain belt) from section-1; because of their proximity, the two sections are likely to be stratigraphically linked to one another (Fig. 5). The loess deposit in section-1 (left in Fig. 5) has a total thickness of ~20 m and is vertically well developed. The deposit is homogeneous and massive and shows little stratification, apparently implying fluvial or hillslope processes were little contributed and in turn, aeolian depositional processes were dominant. The general color of the loess deposit is pale yellow, which is attributed to finely dispersed brown hematite, or light-yellowish; there is little color variation from top to bottom of the sections. One horizon of section-1 (1620–1780 cm below the surface) has abundant carbonate material with a dark brown color, due to strong pedogenic processes occurring during the warm–humid period. Most sub-samples consist mainly of medium and coarse silt fractions; hence, the so-called designation of the deposits as ‘silty loess’ (Pye, 1995; Fig. 6). Loess sediments of section-2 are interbedded with two imbricate gravel and boulder deposits. The fluvial deposit, consisting of upper and lower gravel layers, contains sub-angular, matrix-supported, and well imbricated pebbles and cobbles, implying that the debris or high-density flows occurred during the transition time from glacial to interglacial conditions in an arid mountainous area (Thompson et al., 2002). However, we cannot rule out the possibility that some loess deposits might have been eroded or transected by alluvial processes (sheet flows or slides) that contaminated the deposit with nonaeolian materials.

84

J.H. Youn et al. / Catena 117 (2014) 81–93

J.H. Youn et al. / Catena 117 (2014) 81–93

85

Fig. 3. Vegetation types in the northern Tien Shan. (a) Coniferous forests in the northern Kyrgyz range. (b) and (c) Wheat meadows growing on loess terraces; the sampling site is in the road cut. (d) Apricot tree, which is one of the major vegetation types on the footslopes of the Kyrgyz Tien Shan.

4. Methods 4.1. OSL dating 4.1.1. Samples and sample preparation Medium (16–32 μm) and coarse (32–63 μm) silt is the most abundant particle size fraction of the samples investigated here, and some OSL dating studies of loess or dust deposits have used only this size fraction in their analyses (Roberts et al., 2003; Yang et al., 2012, and more references therein). However, in this study, we used mainly the coarse quartz-grain size fraction (90–250 μm) to derive OSL ages. It is mainly because, within the size range of 32–250 μm, the amount of quartz grains was pretty low and thus, not enough to be divided into more tightly-spaced size ranges (e.g. 32–63 μm, 63–90 μm, 90–125 μm, and so forth); Even with the quartz grains in the size range of 90–250 μm, we could make, at the most, only ~20 large aliquots (composed of several tens of hundreds of grains) for SAR-based OSL dating. Also, it is partly because the coarse-quartz fraction (90–250 μm) may be better bleached at deposition than the fine (4–11 μm) and medium (45–63 μm) fractions, as reported by Zhang et al. (2010), although their dating results were from highly turbid fluvial sediments. Nomenclature on grain size is differently used in geomorphology and OSL dating. For example, coarse silt (32–63 μm) is usually referred to as medium fraction in OSL dating. Thus, for clarity in this paper, we kept using full designation of quartz fraction for mentioning the grain size for OSL dating. Samples for OSL dating were collected from Sections-1 and -2 (20 samples per 1 m interval) by inserting light-tight plastic tubes horizontally into the loess profile. The coarse-sized (90–250 μm) quartz fraction was prepared using a standard procedure involving 10% HCl, 10% H2O2, and 40% HF etching for 1 h; where it is possible (samples NRO-1, 3, 7, 11, 17, and 20), the medium-sized (45–63 μm) quartz

fraction was also prepared with the same procedure as that for the coarse quartz fraction. Fine quartz grains (4–11 μm) were recovered using Stoke's law of settling, followed by standard acid treatments and H2SiF6 treatment for 3 weeks. The grains separated as such were checked for the absence of feldspar contamination by IR stimulation on the natural and beta-irradiated (~20 Gy) aliquots. After the sample preparation procedure, the IRSL (InfraRed Stimulated Luminescence) intensities for all the samples were far less than 5% of BLSL (Blue Light Stimulated Luminescence), which implies that the feldspar grains for all the samples were effectively removed by the acid treatments. 4.1.2. Experimental details The OSL signals were measured using an automated luminescence measurement system (Risø TL/OSL-DA-15) installed at the Korea Basic Science Institute (KBSI). The samples were irradiated using a 90Sr/90Y beta source, delivering ~0.063 ± 0.002 Gy · s–1 to the sample position. The stimulation light source was blue LEDs (470 nm, FWHM 20 nm) and signals were detected using an EMI9235QA through a 7.5 mm Hoya U-340 filter. The OSL signals were measured for 40 s at a readout temperature of 125 °C, with a constant power density of blue-light set to ~40 mW · cm–2. The quartz equivalent dose (De) values were estimated using the single-aliquot regenerative-dose (SAR) procedure suggested by Murray and Wintle (2000, 2003). The OSL signals from the initial 0.8 s of the stimulation less a background estimated from the last 4 s of the stimulation curve were used for the SAR procedure. Depending on the sample amount, ~10–30 8 mm-disc aliquots were used for De estimation of both fine and coarse quartz grains. Dose rates of the samples were derived by converting the radionuclide concentrations of each sample using the data presented by Olley et al. (1996); the radionuclide concentrations were determined using lowlevel high resolution gamma spectrometer installed at KBSI. The dose rates were modified based on present water content. Beta attenuation

Fig. 2. Annual (a) and monthly (b, March; c, August) precipitations across northern Central Asia for 2009 (NASA, 2012). The annual precipitation is approximately 500–1,000 mm in the northern Tien Shan. Precipitation in the northern Tien Shan is low in winter, increases into the spring and early summer (peaking in May), and then decreases from summer to winter, with a small increase in autumn. The southern Tien Shan receives a small amount of precipitation from the Asian Summer Monsoon, penetrating from the south. The red circle indicates the study area.

86

J.H. Youn et al. / Catena 117 (2014) 81–93

particle size analyses were taken immediately adjacent to those for OSL dating, and were tightly sealed into plastic bags. Pre-treatment was conducted at Korea University, following the standard procedure for particle size analysis involving the use of 10% HCl and 10% H2O2 to remove any carbonates and organic matters in the sample. The particle size distributions of the loess samples were measured using a laser scattering particle size analyzer (HELOS) at the Center for Research Facilities at Chungnam National University.

4.3. Radiocarbon dating

Fig. 4. A schematic drawing showing modes of entrainment, transportation, and deposition of aeolian dust in arid regions in the study area. The fine particles (silts and clays) originating from the Tien Shan reach the deserts by fluvial transport; the fine fractions are lifted high into the upper atmosphere by dust storms, which are the principal mode of fine-fraction transport, when temporary reservoirs dry up. Finally, the fine fractions are transported in suspension by winds and are deposited along the footslopes of Tien Shan ranges. Some of the fine fraction entrained in the high-elevation mid-latitude westerlies can be transported far eastward to the northeastern Pacific. The diagram is modified from Machalett et al. (2006).

factor of 0.93 ± 0.03 was used for dose rate calculation. For fine grains, a-value was assumed to be 0.04 (Ree-Jones, 1995). Cosmic ray contributions were considered using the equations by Prescott and Hutton (1988, 1994). 4.2. Particle size analysis A total of 125 Kyrgyz loess samples were collected at 20 cm intervals to analyze grain size distributions in the sediments. The samples for the

Snail fossils were found in a carbonate-rich paleosol layer, which is ~17 m below the surface in the section-1. A snail sample was collected for conventional radiocarbon dating. A correction for the hard-water effects of the old carbonate in which the snail was encased was made by subtracting the mean value of the carbonate concentration of present-day snails from the same location. The carbonate sample was chemically processed and measured by accelerator mass spectrometry at the Korea Institute of Geoscience and Mineral Resources (KIGAM).

5. Results and discussion 5.1. Particle size distribution The Kyrgyz loess is typical silty loess composed of approximately 80% silt, 7% clay, and 12% sand (Fig. 6). The medium and coarse silt fraction is dominant (45.8%), followed by the very fine and fine silt fraction (35%). The grain size distribution pattern is consistent throughout the column. It has been suggested that the Muyunkum desert situated to the northwest of the Kyrgyz loess is the primary proximate source area for dust in the Kyrgyz loess (Orlovsky et al., 2005). Given the lack of grain size variation in the Kyrgyz loess deposits, it is likely that the source area has not significantly changed during the late Quaternary.

Fig. 5. Columnar sections for the loess deposits in the study area in the northern Kyrgyz Tien Shan. Optically stimulated luminescence (OSL) dating was performed on 20 samples, mostly collected at ~1 m intervals; grain size analyses were performed on all 120 samples, collected at 20-cm intervals. The loess deposit consists of two parts: section-1 (left) and section-2 (right). Section-2 is located 20-m headward (southward) of the Section-1. The loess deposit in section-2 is interbedded with fluvial sediments containing cross-bedded sandy and imbricate conglomerate deposits. The total thickness of section-1 and section-2 together is ~30 m.

J.H. Youn et al. / Catena 117 (2014) 81–93

87

Fig. 6. Grain size distribution patterns of the loess sediments in section-1 (a and b) and section-2 (c) in the study area. The silt fractions, composed mainly of fine–medium silt grains, dominate the Kyrgyz loess section.

5.2. Luminescence characteristics of quartz

5.3. Dose recovery test and De estimation

For all the samples investigated here, CW-OSL (continuous wave OSL) signals observed after preheating at 260 °C for 10 s showed fast decay, reaching 10% of their initial count rates in less than ~ 1–2 s of blue-light stimulation, implying that the initial parts of the OSL signals are dominated by the fast OSL component, which is regarded to be the most suitable signal for dating (Wintle and Murray, 2006). Although CW-OSL characteristics are almost identical for both coarse and fine quartz fractions (e.g. sample NRO-17; Fig. 7b), for some samples, particularly for those from the depth interval of 6–11 m (samples NRO-6–11), CW-OSL decay rates of fine quartz fractions are consistently lower than those of coarse fractions (e.g. sample NRO-11; Fig. 7a). This may indicate that, for fine quartz fractions from those samples, relatively higher proportions of slower OSL components contribute to the initial part of the CW-OSL signals, compared with their coarser counterparts. Sensitivity-corrected OSL signal growth, with respect to increasing regeneration doses, were well expressed as a single saturating exponential function for both coarse and fine quartz grains, as exemplified by two representative samples, NRO-11 and NRO-17 (Fig. 7c and d); the preheating conditions for constructing the growth curves were 260 °C (for 10 s) for main regeneration doses and 220 °C (for 0 s) for a test dose (cut-heat). For the sample NRO-11, characteristic doses (D0) for fine fractions are significantly lower than those of coarse fractions, indicating that OSL signals from fine quartz fractions reach to a saturation level at much lower doses than coarse fractions, while D0 values of both fine and coarse fractions are similar to each other for the sample NRO-17; For instance, in Fig. 7c and d, D0 values for fine and coarse quartz fractions of the sample NRO-11 are 96 Gy and 170 Gy, respectively. However, for the sample NRO-17, D0 values for both size fractions are ~120 Gy.

We then tested the ability of the SAR procedure to recover the doses artificially given in the laboratory. Although this test (so-called dose recovery test) does not guarantee the accuracy of De estimation, it is widely accepted as a minimum performance test for any luminescence measurement protocol. In this study, natural OSL signals in both fine and coarse quartz grains from a selected set of samples (NRO-6, -8, -13, -14 and -17, three aliquots per each sample) were optically bleached with blueLEDs for 1000 s at room temperature, and then the bleached grains were stored for 10,000 s at room temperature to allow charges in unstable traps (e.g. 110 °C TL traps) to be thermally evicted. After bleaching again for 1000 s, a series of known doses, similar to natural De, were administered to each sample. These laboratory-given doses are taken as surrogate natural, and then SAR protocol is applied to measure the given doses. The doses given in this experiment are as below; • • • • •

NRO-6: 70 Gy (coarse), 140 Gy (fine) NRO-8: 70 Gy (coarse), 140 Gy (fine) NRO-13: 100 Gy (coarse), 180 Gy (fine) NRO-14: 190 Gy (coarse), 250 Gy (fine) NRO-17: 240 Gy (coarse), 250 Gy (fine)

As shown in Fig. 8, using the SAR procedure with a preheat condition of 260 °C for 10 s and a cut-heat of 220 °C for 0 s, the given doses for all the samples were well recovered within ±10%. It is worth noticing here that, for samples NRO-14 and -17, the doses higher than the saturation level (dose values higher than 2D0 value, i.e. measured dose to 2D0 ratios higher than 1.0; Wintle and Murray, 2006) were also accurately measured (see the inset to Fig. 8). Based on this, the experimental conditions used in the dose recovery test were applied for De estimation of the samples. Almost all the samples analyzed here showed recycling

88

J.H. Youn et al. / Catena 117 (2014) 81–93

Fig. 7. (a) CW-OSL decay rates of fine quartz fractions of the sample NRO-11 are lower than that of coarse grains, (b) while CW-OSL characteristics of the sample NRO-17 are almost identical for both fine and coarse quartz grains. (c) For fine quartz fractions of the sample NRO-11, characteristic doses (D0) of the growth curve are significantly lower than those of coarse fractions. (d) In contrast, the D0 values of both fine and coarse fractions are similar to each other for the sample NRO-17.

ratios within 10% of unity and recuperations (response to 0 Gy) were less than 2%. 5.4. Dependence of OSL ages on quartz grain-size In Table 1, the OSL ages of coarse (90–250 μm) and fine (4–11 μm) quartz grains are summarized, and the ages obtained using different size fractions are compared with each other with depth. As shown in Fig. 9a, except four samples in a depth interval of 6–11 m, where the OSL ages of the fine fractions were clearly older than those based on coarse fractions (age ratios up to ~1.7; Table 1, Fig. 9b), the OSL ages of fine and coarse quartz grains are consistent with each other within 2σ uncertainty level. At a depth of 11 m (NRO-11), the OSL age of

Fig. 8. Using the SAR procedure, the doses given to a set of samples were well recovered within ±10%. In the inset, the measured dose and 2D0 ratios of the samples are shown.

medium-sized (45–63 μm) quartz is identical (~25 ka) to that from coarse quartz fraction, while the fine quartz fraction shows ~50% older ages (39 ka). However, the OSL ages of medium quartz grains of the sample NRO-17 are younger than any other size fractions by ~20%; for this sample, the OSL ages of fine and coarse quartz fractions are indistinguishable, with the age ratios of fine to coarse quartz being 1.00 ± 0.05. At this stage of our investigation, we could not give plausible explanations on the OSL age discrepancy between coarse and fine quartz fractions in a depth interval of 6–11 m. It may be due to the difference in luminescence characteristics between these size fractions, as exemplified by one of the samples in this depth interval (NRO-11); Note that the samples that show consistent OSL ages have very similar luminescence characteristics with each other (e.g. NRO-17; Fig. 7b and d). Otherwise, it may have resulted from incomplete bleaching of fine quartz grains at deposition. In most previous OSL dating works on loess deposits used fine (4–11 μm) or medium (28–63 μm) quartz fractions, due to the abundance of silt-sized grains in loess deposits (Roberts et al., 2003; Yang et al., 2012, and more references therein). However, it has been reported that, in unusual cases, the OSL ages of fine fractions can overestimate the depositional age by incomplete bleaching at grain deposition (Yang et al., 2012; Zhang et al., 2009, 2010). In addition, post-depositional processes might have been the cause of the age overestimation observed from fine quartz fractions. One of the essential assumptions in OSL dating of sediments is that the effects of post-depositional processes, such as pedogenesis and bioturbation, are negligible (Roberts et al., 1999; Stevens et al., 2006). It is also possible that partially bleached quartz grains have been incorporated into loess sections by rolling, soil creep, and/or shallow landslides, on account of gravitational forces and unstable slope configurations. These phenomena are common present-day processes in the study area, as Kyrgyz loess sections occur in periglacial environments characterized by substantial diurnal and seasonal changes in both temperature and precipitation (Aizen et al., 1996; Williams and Konovalov, 2008).

Table 1 Summary of radionuclide concentrations, equivalent doses, dose rates and OSL ages of the coarse and fine quartz fractions. Coarse quartz fraction (CG) (90–250 μm) U (Bq · kg−1)

238

NRO-1 NRO-2 NRO-3 NRO-4 NRO-5 NRO-6 NRO-7 NRO-8 NRO-9 NRO-11 NRO-12 NRO-13 NRO-14 NRO-15 NRO-16 NRO-17 NRO-18 NRO-19 NRO-20 NRO-22

40.5 32.1 40.3 40.8 45.6 38.9 39.9 27.0 36.1 46.4 41.3 38.3 36.8 49.2 45.6 43.9 30.9 39.8 41.0 44.7

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

9.8 8.1 10.2 10.1 11.2 9.2 9.4 7.3 8.9 11.5 10.3 9.5 9.2 11.6 11.6 10.7 8.1 9.8 10.1 10.7

Ra (Bq · kg−1)

226

38.4 43.2 40.0 40.8 37.1 39.1 34.7 36.0 35.1 34.7 36.7 38.1 39.0 42.3 38.9 35.6 36.2 32.7 32.2 48.0

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.6 0.8 0.8 0.8 0.7 0.5 0.5 0.6 0.7 0.9 0.9 0.7 0.7 0.6 0.9 0.8 0.7 0.7 0.6 0.7

Th (Bq · kg−1)

232

48.6 44.3 50.3 51.8 47.1 49.0 45.3 43.6 46.5 46.9 49.5 49.2 47.4 58.2 49.8 45.5 49.6 42.7 46.0 59.2

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

1.2 1.6 1.8 1.6 1.5 0.9 0.9 1.4 1.5 1.9 1.8 1.5 1.5 1.1 1.8 1.7 1.6 1.6 1.5 1.3

K (Bq · kg−1)

40

674 615 668 634 626 632 649 619 645 718 706 663 693 803 660 608 609 600 629 716

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

14 15 17 16 15 12 13 12 15 19 18 15 16 15 17 16 15 15 15 15

Fine quartz fraction (FG) (4–11 μm)

w.c.a (%)

Total dose rate b (Gy · ka−1)

De (Gy)

Age c (ka)

nd

Total dose rateb (Gy · ka−1)

De (Gy)

Age 3 (ka)

nd

Age ratio (fine to coarse)

16 7 6 3 4 1 2 1 1 6 6 23 22 15 16 14 12 14 17 12

3.28 3.33 3.65 3.72 3.54 3.72 3.62 3.47 3.64 3.74 3.74 3.03 3.13 3.90 3.27 3.08 3.20 2.96 3.08 3.83

13 26 87 43 81 71 87 71 74 93 108 111 189 168 253 243 212 222 203 231

4 ± 0.(3) 8±3 24 ± 2 12 ± 2 23 ± 1 19 ± 2 24 ± 1 21 ± 2 20 ± 2 25 ± 1 29 ± 2 37 ± 2 60 ± 2 43 ± 3 77 ± 4 79 ± 3 66 ± 4 75 ± 4 66 ± 4 60 ± 4

21 6 13 11 14 16 21 16 8 22 15 16 16 15 16 16 15 14 16 21

3.87 – 4.32 – 4.19 4.42 4.28 4.10 4.31 4.41 – 3.58 3.67 – 3.87 3.65 – – – 4.56

20 ± 0.(3) – 114 ± 2 – 112 ± 2 134 ± 2 126 ± 3 144 ± 2 137 ± 2 171 ± 4 – 179 ± 4 235 ± 6 – 327 ± 6 288 ± 7 – – – 411 ± 7

5 ± 0.(1) – 26 ± 1 – 27 ± 1 30 ± 1 30 ± 1 35 ± 1 32 ± 1 39 ± 1 – 50 ± 2 64 ± 2 – 84 ± 3 79 ± 3 – – – 90 ± 3

24 – 24 – 24 24 24 24 24 24 – 24 24 – 24 16 – – – 30

1.25 – 1.08 – 1.17 1.58 1.25 1.67 1.60 1.56 – 1.35 1.07 – 1.09 1.00 – – – 1.50

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

0.09 0.07 0.10 0.07 0.10 0.07 0.10 0.06 0.10 0.11 0.10 0.08 0.08 0.10 0.09 0.09 0.09 0.08 0.09 0.10

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

1 11 7 8 3 6 4 6 6 3 5 6 5 12 9 7 11 9 12 14

± 0.09 ± 0.10 ± ± ± ± ± ±

0.10 0.10 0.10 0.09 0.10 0.11

± 0.08 ± 0.09 ± 0.09 ± 0.09

± 0.10

± 0.10 ± 0.10 ± ± ± ± ± ±

0.07 0.17 0.07 0.17 0.17 0.07

± 0.09 ± 0.05

J.H. Youn et al. / Catena 117 (2014) 81–93

Sample

± 0.07 ± 0.05

± 0.11

– Not measured. a Present water content. b The dose rates were calculated based on present water content. Contributions from cosmic ray were calculated using the equation given by Prescott and Hutton (1994). Beta attenuation factor of 0.93 ± 0.03 was used for dose rate calculation of coarse quartz fractions. For fine quartz fractions, a-value was assumed to be 0.04 (Ree-Jones, 1995). c Uncertainties are given in 1σ standard error. d Number of aliquots used for age calculation.

89

90

J.H. Youn et al. / Catena 117 (2014) 81–93

Fig. 9. (a) Age vs. depth profile of the loess section. (b) Ratios of OSL ages from fine and coarse quartz grains. Except four samples in the depth interval of 6–11 m, the OSL ages of fine and coarse quartz grains are consistent with each other within 2σ uncertainty level (uncertainties of each point are in 2σ standard error).

Given the location of the study area, and that glacial environments are a potential source area for sediments in the loess (which tend to be less bleached; Fuchs and Owen, 2008), it is possible that some incompletely bleached glacially derived materials are incorporated into the loess section, particularly when the sediments in a depth interval of 6–11 m are deposited, resulting in an age overestimation for fine quartz grains. Although the interpretation of De distribution of the samples are limited by the small number of aliquots used for dating, De values of both coarse and fine quartz grains of the sample NRO-11, which shows the fine-tocoarse age ratio of 1.56 ± 0.07, seem to close to normal distribution (Fig. 10a). In addition, the scatter of De values in fine quartz fractions of the sample NRO-11 are observed to be smaller than that of NRO-17, the age ratio of which is unity (Fig. 10b). Thus, in this study, it is not possible to decide whether the fine grains were incompletely bleached at deposition or whether there has been significant post-depositional input of incompletely bleached fine quartz grains into the loess. Considering all these possibilities of erroneously older ages by fine quartz fractions, and that there observed no apparent undesirable luminescence properties in coarse quartz fractions which can cause OSL age underestimation (e.g. presence of ultrafast OSL component, significant contributions from slower OSL components; Choi et al., 2003a,b), the coarse quartz grains (90–250 μm) are considered to be more reliable

for OSL dating than fine grains. Thus, we take the OSL ages derived from the coarse quartz fraction as being representative of depositional timing of the loess in the study area; in the following discussions, the OSL ages refer to the ages obtained from coarse quartz fractions. 5.5. OSL ages of loess deposits Based on the samples that show consistent OSL ages of coarse and fine quartz fractions (samples NRO-1, -3, -14 and -17; Fig. 9), we designated four key beds at depths of 1 m, 3 m, 14 m and 17 m, and a general stratigraphy of the loess deposits was established. The OSL ages of the samples are ranging ~ 4–80 ka, covering the Holocene, MIS (Marine Isotope Stage) 2, MIS 3 and MIS 4 (Figs. 9 and 11). The age–depth profile (Fig. 11) shows relatively higher deposition rates during cold periods (61 cm · ka−1 and 25 cm · ka−1 for MIS 2 and MIS 4, respectively), compared with those during the Holocene (18 cm · ka−1) and MIS 3 (11 cm · ka−1). Multiple layers of fluvial gravel sheets were observed in section-2, which is located ~20 m from section-1. The age of the loess deposits between the fluvial gravel sheets can be constrained by the OSL age of the samples NRO-22 (60 ± 4 ka), thus this loess deposits are correlated to MIS 3. In terms of morphostratigraphy, the two gravel layers may have

Fig. 10. De distribution of fine and coarse quartz fractions of the samples (a) NRO-11 and (b) NRO-17.

J.H. Youn et al. / Catena 117 (2014) 81–93

91

Fig. 11. Stratigraphic sections of the loess deposits in the study area. The stratigraphy (with reference to marine isotope stages, MIS) was constructed on the basis of key beds, which show similar ages deduced from the coarse and fine quartz fractions. Two conglomerate layers are found in section-2. It is likely that the frequency of flash floods increased during the transition period from glacial to interglacial conditions, when temperatures increased in arid Central Asia.

formed during the transition from MIS 4 to MIS 3, and MIS 2 to the Holocene, respectively, when climate amelioration possibly caused more frequent flooding in mountain and foothill regions of Central Asia (Hetzel et al., 2002; Thompson et al., 2002). At a depth of 17 m, a single carbonate-rich layer is identified with a thickness of ~1 m (see Fig. 5a), from which the OSL age was estimated to be 79 ± 3 ka. From this layer, snail samples were collected for radiocarbon dating. The snail sample yielded a radiocarbon age of ~45 ka, close to the upper limit of radiocarbon dating method. Thus, this radiocarbon age should be taken as a minimum and we put more credence on the OSL age of ~80 ka. 5.6. Mass accumulation rate of loess deposits and implications for paleoclimate reconstructions Based on our OSL ages (Fig. 11), we calculated an MAR (Mass Accumulation Rate) over the Late Quaternary (since the beginning of MIS 4, ~80 ka), which can be used as a good proxy for aridity in the area (Pye, 1995). It is likely that the Kyrgyz loess deposits were formed

over the late Quaternary and that dust accumulation has continued until the present (Küster et al., 2006; Narama et al., 2009; Svendsen et al., 2004). The apparent total MAR over the past 80 ka is greater than 25 cm · ka−1, considering interrupted deposition and erosion of dust particles. The minimum MARs are variable: ~25 cm · ka−1 during MIS 4, ~11 cm · ka−1 during MIS 3, ~61 cm · ka−1 during MIS 2, and ~18 cm · ka−1 during the Holocene. The greatest MAR was found during MIS 2 (~61 cm · ka−1), possibly indicating that the area was much more arid at this time than during other periods. This finding is consistent with the fact that the area was influenced by stronger SHP systems and global cooling, which resulted in reduced vegetation cover and greater exposure of the area to dry-wind systems. It is also worth noticing that the OSL age discrepancy between fine and coarse quartz fractions is particularly large in the MIS 2 deposits (Fig. 9b). In this area, loess was deposited even during the warm and humid Holocene period, which is consistent with the previous suggestion that loess can be deposited without any significant cessation during warm–humid periods, even though the amount of dust in the source area is diminished due to increasing vegetation cover; however, the

Fig. 12. The timing of glacial advances throughout Central Asia. The red dots indicate the major timings of glacier advances in the high mountains of Central Asia. Sanhueza-Pino et al., 2011.

92

J.H. Youn et al. / Catena 117 (2014) 81–93

increased vegetation cover can also act as an effective dust trap in depositional areas (Küster et al., 2006). Glacier and related landforms are one of the most reliable proxies for understanding climate changes during the Quaternary in the northern Central Asia, as well as in other regions. In recent years, paleoclimate of this region has been reconstructed using various direct numerical dating techniques on glacial landforms, including terrestrial cosmogenic nuclide surface exposure dating, OSL dating, and conventional radiocarbon dating (Abramowski et al., 2006; Koppes et al., 2008; Narama and Okuno, 2006; Narama et al., 2007, 2009; Rupper et al., 2009; Zech, 2012; Zech et al., 2000, 2005; Fig. 12). Most glaciers in the Northern Hemisphere advanced during cold glacial times. However, glacier advances in Central Asia were variable and depended on the location. Specifically, over the northern Tien Shan, most glaciers advanced to their most distal positions during MIS 4 or MIS 5, rather than during MIS 2, implying that moisture availability is the dominant factor controlling glacier growth in the region (Koppes et al., 2008; Zech, 2012; Fig. 12). Similarly, there is no clear evidence for extensive glacier advances during MIS 2, at the time when SHP systems were likely to block advection of moist air into the region via mid-latitude westerlies. It has been reported that the deposition of loess (wind-blown sediment) during cold–dry glacial periods is due to the transportation of fine sediments by strengthened winds (Porter and An, 1995). As such, it seems that periods of major loess deposition should be synchronous with major glacier advances (Bettis et al., 2003; Lu and Sun, 2000; Vandenberghe and Nugteren, 2001). When glaciers expanded during glacial times, loess was deposited in surrounding regions. In this regard, periods of loess deposition appear to be well correlated with global climate fluctuations during the last glacial cycle. Both MIS 2 and MIS 4 are the periods of major glacier expansion in the Northern Hemisphere. The greater dust accumulation rate during MIS 2, in comparison with that for MIS 3 and MIS 4, is in good agreement with the aridity (or lack of moisture availability) over the northern Tien Shan during this time, which resulted in restricted glacier expansion during MIS 2; this pattern developed because the strengthened SHP system hindered moisture availability and the consequent growth and advance of glaciers in the region (Fig. 12). Thus, the cold–dry SHP systems can be considered as a major driving force and cause of aridity in dust source areas in the northern Central Asia. In general, there is a positive relationship between loess deposition and glacial expansion, as loess tends to be deposited in foreland basins adjacent to glaciated mountain regions during dry-windy glacial conditions. However, in the northern part of the Tien Shan, which is climatologically affected by both mid-latitude westerlies and the SHP, a seesaw relationship between two dominant landforms in this high mountain region is observed over the period of the last glacial cycle. When the dry–cold SHP extended broadly southward, the moisture-bearing midlatitude westerlies were blocked, thus restricting glacier growth and expansion on account of the lack of moisture. Instead, aridity increased under the influence of dry SHP systems. This aridity caused the high loess accumulation rates during MIS 2, and the restrictions on glacier growth and advance. The inverse relationship between glacier advance and loess accumulation supports previous views on glacier advance in northern Central Asia, which indicate that glacier advance is strictly constrained by moisture availability rather than temperature (Zech, 2012). 6. Conclusions In this paper, we applied OSL dating method to establish chronology of the Kyrgyz loess, and to reconstruct past climate changes in northern Central Asia. The OSL ages of fine and coarse quartz fractions are consistent with each other within 2σ uncertainty level. However, in a depth interval of 6–11 m, which showed greatest accumulation rate during MIS 2, there observed a substantial OSL age discrepancy between fine and coarse quartz fractions with fine-to-coarse age ratios ranging

~1.3–1.7. Based on the OSL ages of coarse quartz fractions, the MAR during MIS 2 appears to be noticeably greater than those during MIS 4 or MIS 3, suggesting that the aridity during MIS 2 was considerably higher in the northern Tien Shan than during other periods of the last glacial cycle. This finding is consistent with glacial records. During MIS 2, strengthening of the cold–dry SHP blocked the penetration of midlatitude westerlies carrying moisture from the Caspian, Aral, North Atlantic, and Mediterranean regions west of the study area, which resulted in a restriction of glacier advances in northern Central Asia. Generally, a positive relationship exists between the rate of loess deposition and the activity of glaciers, as loess tends to be deposited in periglacial regions mainly during glacial periods. However, our study suggests an inverse relationship between loess deposition and glaciation in the northern Tien Shan during the last glacial cycle. It should be noted that the SHP system is a strong driving mechanisms for the processes causing landform evolution in northern Central Asia, as the strength of the system results in variable responses of landforms to essentially the same climate. Acknowledgments This work was supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2011-32AB00268) for Y.B. Seong and also partially supported by the KBSI grant (Project No. C3371A) for J.H. Youn and J.-H. Choi. We greatly thank Dr. L.A. Owen and two anonymous reviewers for their constructive reviews on our early version of the manuscript. References Abramowski, U., Seebach, D., Zech, R., Glaser, B., Sosin, P., Kubik, P.W., Zech, W., 2006. Pleistocene glaciations of Central Asia: results from 10Be surface exposure ages of erratic boulders from Pamir (Tajikistan), and the Alay-Turkestan range (Kyrgyzstan). Quat. Sci. Rev. 25, 1080–1096. Aizen, V.B., Aizen, E.M., Melack, J.M., 1996. Precipitation, melt and runoff in the northern Tien Shan. J. Hydrol. 186, 229–251. Beer, R., Kaiser, F., Schmidt, K., Ammann, B., Carraro, G., Grisa, E., Tinner, Willy, 2008. Vegetation history of the walnut forests in Kyrgyzstan (Central Asia): natural or anthropogenic origin. Quat. Sci. Rev. 27, 621–632. Bettis, E.A., Muhs, D.R., Roberts, H.M., Wintle, A.G., 2003. Last glacial loess in the conterminous U.S.A. Quat. Sci. Rev. 22, 1907–1946. Bronger, A., Winter, R., Derevjanko, O., Aldag, S., 1995. Loess–palaeosol sequences in Tadjikistan as a palaeoclimatic record of the Quaternary in Central Asia. Quat. Proc. 4, 69–81. Bugaev, V.A., Djordjio, V.A., Kozik, E.M., Petrosyanz, M.A., Pshenichnii, A.Y., Romanov, A.A., Chernishova, O.N., 1957. Synoptic Processes in Middle Asia, Tashkent (in Russian with English abstract). Hydrometeo, Leningrad (114 pp.). Burbank, D.W., Li, J., 1985. Age and palaeoclimatic significance of the loess of Lanzhou, north China. Nature 316, 429–431. Burtman, V.S., Skobelev, S.F., Molnar, P., 1996. Late Cenozoic slip on the Talas-Ferghana fault, the Tien Shan, central Asia. Geol. Soc. Am. Bull. 108, 1004–1021. Cheng, H., Zhang, P.Z., Spötl, C., Edwards, R.S., Cai, Y.J., Zhang, D.Z., Sang, W.C., Tan, M., An, Z.S., 2012. The climatic cyclicity in semiarid–arid central Asia over the past 500,000 years. Geophys. Res. Lett. 39, L01705. http://dx.doi.org/10.1029/2011GL050202. Choi, J.H., Murray, A.S., Jain, M., Cheong, C.S., Chang, H.W., 2003a. Luminescence dating of well-sorted marine terrace sediments on the southeastern coast of Korea. Quat. Sci. Rev. 22, 407–421. Choi, J.H., Murray, A.S., Cheong, C.S., Hong, D.G., Chang, H.W., 2003b. The resolution of stratigraphic inconsistency in the luminescence ages of marine terrace sediments from Korea. Quat. Sci. Rev. 22, 1201–1206. Derbyshire, E., Kemp, R.A., Meng, X., 1997. Climate change, loess and palaeosols: proxy measures and resolution in North China. J. Geol. Soc. Lond. 154, 793–805. 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. Dodonov, A.E., Baiguzina, L.L., 1995. Loess stratigraphy of central Asia: paleoclimatic and palaeoenvironmental aspects. Quat. Sci. Rev. 95, 707–720. 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., 1998. Loess chronology of the Middle and Upper Pleistocene in Tadjikistan. Geol. Rundsch. 87, 2–20. Fuchs, M., Owen, L.A., 2008. Luminescence dating of glacial and associated sediments: review, recommendations and future directions. Boreas 37, 636–659. Glorie, S., Grave, J.D., Buslov, M.M., Elburg, M.A., Stockli, D.F., Gerdes, A., Van den haute, P., 2010. Multi-method chronometric constraints on the evolution of the Northern Kyrgyz Tien Shan granitoids (Central Asian Orogenic Belt): from emplacement to exhumation. J. Asian Eart Sci. 38, 131–146.

J.H. Youn et al. / Catena 117 (2014) 81–93 Goudie, A.S., 2009. Dust storms: Recent developments. J. Environ. Manag. 90, 89–94. Guo, Z.T., Ruddiman, W.R., Hao, Q.Z., Wu, H.B., Qiao, Y.S., Zhu, R.X., Peng, S.Z., Wei, J.J., Yuan, B.Y., Liu, T.S., 2002. Onset of Asian desertification by 22 Myr ago inferred from loess deposits in China. Nature 416, 159–162. Hetzel, R., Niedermann, S., Tao, M., Kubik, P.W., Ivy-Ochs, S., Gao, B., Strecker, M.R., 2002. Low slip rates and long-term preservation of geomorphic features in Central Asia. Nature 417, 428–432. Kapp, P., Pelletier, J.D., Rohrmann, A., 2011. Wind erosion in the Qaidam basin, central Asia: implications for tectonics, paleoclimate, and the source of the Loess Plateau. GSA Today 21. http://dx.doi.org/10.1130/GSATG99A.1. Koppes, M., Gillespie, A.R., Burke, R.M., Thompson, S.C., Stone, J., 2008. Late Quaternary glaciation in the Kyrgyz Tien Shan. Quat. Sci. Rev. 27, 846–866. Kukla, G., 1987. Loess stratigraphy in Central China. Quat. Sci. Rev. 6, 191–219. Küster, Y., Hetzel, R., Krbetscheka, M., Taoc, M., 2006. Holocene loess sedimentation along the Qilian Shan (China): significance for understanding the processes and timing of loess deposition. Quat. Sci. Rev. 25, 114–125. Li, G., Chen, J., Ji, J., Yang, J., Conway, T.M., 2009. Natural and anthropogenic sources of East Asian dust. Geology 37, 727–730. Lu, H.Y., Sun, D.H., 2000. Pathways of dust input to the Chinese Loess Plateau during the last glacial and interglacial periods. Catena 40, 251–261. Machalett, B., Frechen, M., Hambach, U., Oches, E.A., Zoller, L., Markovic, S.B., 2006. The loess sequence from Remisowka (northern boundary of the Tien Shan Mountains, Kazakhstan)-Part I: Luminescence dating. Quat. Int. 152–153, 192–201. Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single aliquot regenerative-dose protocol. Radiat. Meas. 32, 57–73. Murray, A.S., Wintle, A.G., 2003. The single aliquot regenerative dose protocol: potential for improvements in reliability. Radiat. Meas. 37, 377–381. Narama, C., Okuno, M., 2006. Record of glacier variations during the last glacial in the Turkestan range of the Pamir-Alay, Kyrgyz Republic. Ann. Glaciol. 43, 397–404. Narama, C., Kondo, R., Tsukamoto, S., Kajiura, T., Ormukov, C., Abdrakhmatov, K., 2007. OSL dating of glacial deposits during the Last Glacial in the Terskey-Alatoo Range, Kyrgyz Republic. Quat. Geochronol. 2, 249–254. Narama, C., Kondo, R., Tsukamoto, S., Kajiura, T., Duishonakunov, M., Abdrakhmatov, K., 2009. Timing of glacier expansion during the Last Glacial in the inner Tien Shan, Kyrgyz Republic by OSL dating. Quat. Int. 199, 147–156. National Aeronautics and Space Administration (NASA), 2012. Goddard Earth Sciences Data and Information Services Center, Giovanni TOVAS Program. http://disc.gsfc.nasa.gov/. Olley, J.M., 2004. Optical dating of deep-sea sediments using single grains of quartz: a comparison with radiocarbon. Sediment. Geol. 169, 175–189. Olley, J.M., Murray, A.S., Roberts, R.G., 1996. The effect of disequilibria in the uranium and thorium decay chains on burial dose rates in fluvial sediments. Quat. Sci. Rev. 15, 751–760. Orlovsky, L., Orlovsky, N., Durdyev, A., 2005. Dust storms in Turkmenistan. J. Arid Environ. 60, 83–97. Porter, S.C., An, Z., 1995. Correlation between climate events in the North Atlantic and China during the last glaciation. Nature 375, 305–308. Prescott, J.R., Hutton, J.T., 1988. Cosmic ray and gamma ray dosimetry for TL and ESR. Nucl. Tracks Radiat. Meas. 14, 223–227. Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contribution to dose rates for luminescence and ESR dating: large depths and long-term time variation. Radiat. Meas. 23, 497–500. Pye, K., 1995. The nature, origin and accumulation of loess. Quat. Sci. Rev. 14, 653–667. Ree-Jones, J., 1995. Optical dating of young sediments using fine-grain quartz. Ancient TL 13, 9–13. Roberts, H.M., 2008. The development and application of luminescence dating to loess deposits: a perspective on the past, present and future. Boreas 37, 483–507.

93

Roberts, R.G., Galbraith, F., Olley, J.M., Yoshida, H., Laslett, G.M., 1999. Optical dating of single and multiple grains of quartz from Jinmium rock shelter, northern Australia: Part II, results and implications. Archaeometry 41, 365–395. Roberts, H.M., Muhs, D.R., Wintle, A.G., Duller, G.A.T., Bettis III, E.A., 2003. Unprecedented last-glacial mass accumulation rates determined by luminescence dating of loess from western Nebraska. Quat. Res. 59, 411–419. Rupper, S., Roe, G., Gillespie, A., 2009. Spatial patterns of Holocene glacier advance and retreat in Central Asia. Quat. Res. 72, 337–346. Sanhueza-Pino, K., Korup, O., Hetzel, R., Munack, H., Weidinger, J.T., Dunning, S., Ormukov, C., Kubik, P.W., 2011. Glacial advances constrained by 10Be exposure dating of bedrock landslides, Kyrgyz Tien Shan. Quat. Res. 76, 295–304. Shackleton, N.J., An, Z., Dodonov, A.E., 1995. Accumulation rate of loess in Tadjikistan and China: relationship with global ice volume cycles. Quat. Proc. 4, 1–6. Stevens, T., Armitage, S.J., Lu, H., Thomas, D.S.G., 2006. Sedimentation and diagenesis of Chinese loess: implications for the preservation of continuous, high-resolution climate records. Geology 34, 849–852. Sun, D., Su, R., Bloemendal, J., Lu, H., 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. Svendsen, J., Alexanderson, H., Astakhov, V., Demidov, I., Dowdeswell, J., Funder, S., Gataullin, V., Henriksen, M., Hjort, C., Houmark-Nielsen, M., Hubberten, H., Ingólfson, Ó, Jakobsson, M., Kjær, K., Larsen, E., Lokrantz, H., Lunkka, J., Lysa˚, A., Mangerud, J., Matiouchkov, A., Murray, A., Mo¨ ller, P., Niessen, F., Nikolskaya, O., Polyak, P., Saarnisto, M., Siegert, C., Siegert, M., Spielhagen, R., Stein, R., 2004. Late Quaternary ice sheet history of Northern Eurasia. Quat. Sci. Rev. 3, 1229–1271. Thompson, S.C., Weldon, R., Rubin, C.M., Abdrakhmatov, K., Molnar, P., Berger, G.W., 2002. Late Quaternary slip rates across the central Tien Shan, Kyrgyzstan, central Asia. J. Geophys. Res. 107 (B9), 2203. Vandenberghe, J., Nugteren, G., 2001. Rapid climatic changes recorded in loess successions. Glob. Planet. Chang. 28, 1–9. Williams, M.W., Konovalov, V.G., 2008. Central Asia Temperature and Precipitation Data, 1879–2003. USA National Snow and Ice Data Center. Digital media, Boulder, Colorado. Wintle, A.G., Murray, A.S., 2006. A review of quartz optically stimulated luminescence characteristics and their relevance in single-aliquot regeneration dating protocols. Radiat. Meas. 41, 369–391. Yang, H., Chen, J., Thompson, J.A., Liu, J., 2012. Optical dating of the 12 May 2008, Ms 8.0 Wenchuan earthquake-related sediments: tests of zeroing assumptions. Quat. Geochronol. 10, 273–279. Zech, R., 2012. A late Pleistocene glacial chronology from the Kitschi-Kurumdu Valley, Tien Shan (Kyrgyzstan), based on 10Be surface exposure dating. Quat. Res. 77, 281–288. Zech, W., Glaser, B., Ni, A., Petrov, M., Lemzin, I., 2000. Soil as indicators of the Pleistocene and Holocene landscape history: Alay Range (Khyrgstan). Quat. Int. 65–66, 161–170. Zech, R., Abramowski, U., Glaser, B., Sosin, P., Kubik, P.W., Zech, W., 2005. Late Quaternary glacial and climate history of the Pamir Mountains derived from cosmogenic 10Be exposure ages. Quat. Res. 64, 212–220. Zhang, J., Qiu, W., Li, R., Zhou, L., 2009. The evolution of a terrace sequence along the Yellow River (HuangHe) in Hequ, Shanxi, China, as inferred from optical dating. Geomorphology 109, 54–65. Zhang, J.-F., Qiu, W.-L., Wang, X.-Q., Hu, G., Li, R.-Q., Zhou, L.-P., 2010. Opical dating of a hyperconcentrated flow deposit on a Yellow River terrace in Hukou, Shaanxi, China. Quat. Geochronol. 5, 194–199.