Precipitation evolution of Central Asia during the last 5000 years

Abstract

Central Asia is located at the confluence of large-scale atmospheric circulation systems. However, the number of Holocene climate records is still low in most parts of this region and insufficient to allow detailed discussion and comparisons to disentangle the complex climate history and interplays between the different climatic systems. Here, we present the first stalagmite record from arid Central Asia (south-western Kyrgyzstan) by using δ¹⁸O, δ¹³C, and micro x-ray fluorescence (µXRF)-sulfur data spanning the last 5000 years. The cave hosting stalagmite Uluu-2 is ideally suited to identify past shifts in seasonal variations in precipitation in this part of the world. Comparison of instrumental and paleo-isotopic studies demonstrates that the Uluu-2 speleothem isotope composition faithfully records climate changes and responds to shifts in the proportion of moisture derived from mid-latitude Westerlies during the winter/spring season. The reconstructions suggest that the area was characterized by a dry climate from 4700 to 3900 yr BP, interrupted by a wet episode around 4200 yr BP. Further drier conditions also occurred between 4000 and 3500 yr BP. Wetter conditions were re-established at ca. 2500 yr BP, after another dry episode between 3000 and 2500 yr BP. With the exception of two short dry events (1150 and 1300 yr BP), the period after 1700 yr BP shows moderate to wet conditions. Regional comparisons suggest that the strength and position of the Westerly winds control climatic shifts in arid Central Asia, leading to complex local responses.

Keywords

Central Asia North Atlantic Oscillation precipitation reconstruction Siberian High speleothem stable isotope Westerlies

Introduction

The impact of global warming on current and future water resources is attracting increasing attention. Virtually no other region on Earth is thought to be affected as strongly by future changes in water supply than arid Central Asia (CA; Chase et al., 2000; De Martino et al., 2005). A projected increase in temperature and changes in the interannual variability of rainfall may bring further environmental and economic stress to a region, which is characterized by population growth and a high vulnerability to natural disasters (De Martino et al., 2005; Held and Soden, 2006). In the light of anthropogenic forcing of climate and associated surface processes (IPCC, 2013), it is necessary to understand changes in the response of the natural climate system to long-term forcing mechanisms, such as insolation and the interaction between atmospheric and oceanic circulation systems and their relationship with short-term variations on decadal to centennial or millennial timescales. Furthermore, understanding teleconnections is crucial to decipher climate relationships over long distances that reflect organized patterns of atmospheric and oceanic circulation. Because of their effect on large-scale synoptic patterns, mountain systems in CA modulate global-scale climate variability, but at the same time are highly sensitive to such variations (Sorg et al., 2012). Different climate archives have been used for understanding the history and mechanisms of climate change in CA during the Holocene (e.g. lakes (Boomer et al., 2009; Mathis et al., 2014; Mischke et al., 2010; Ricketts et al., 2001; Sorrel et al., 2007), tree-rings (Esper et al., 2007), ice cores (Aizen et al., 2006; Thompson et al., 1997)) especially in the analyses of long-term changes in zonal and meridonal atmospheric circulation patterns and its teleconnections, such as the North Atlantic Oscillation (NAO; Aizen et al., 2006; Lauterbach et al., 2014) or the El Niño Southern Oscillation (Barlow et al., 2002). The increasing number of climate archives from CA has revealed significant spatial differences in the Holocene climate history between specific regions throughout CA. However, the number of Holocene climate records is still low in many parts of CA and insufficient to allow detailed discussion and comparisons to disentangle the complex climate history and interplays between the different climatic systems.

Here, we present evidence from a stalagmite from Uluu-Too cave located in the core region of arid CA between the foothills of the Tien Shan and the intramontane Fergana Valley in south-western Kyrgyzstan. Our results will help to understand hydrological changes in the Central Asian mountains and to unravel the forcing mechanisms behind precipitation variations that have occurred during the last 5000 years.

Study site

Cave settings

The studied Uluu-Too cave (or Onyxsovaya, named after the mined Onyx mineral) lies within folded Upper Devonian to Lower Carboniferous limestones of the Uluu-Too Mountains on the southern rim of the Fergana Valley in Kyrgyzstan (Figure 1). This small mountain range with monolithic character is situated to the west of the city of Osh and close to the Uzbek border (40°23′N, 72°21′E; Figure 1).

Figure 1.

(a) Map of the Asian continent, showing topographic conditions and principal atmospheric trajectories. Strengths and compositions of these atmospheric regimes have a direct effect on the amount of rainfall as well as influencing where the moisture comes from and the pathway that it follows. Brown dashed line indicates current monsoon limits, adapted from Holmes et al. (2009); (b), (c) position of the Uluu-Too cave (yellow and black star), on the southern rim of the Fergana Basin. The cave is located at 40°23′N, 72°21′E and at about 1490 m above sea level, in Paleozoic limestones of the Uluu-Too Mountains west of the city Osh close to the Uzbek border.

The narrow cave entrance lies on the southern mountain slope at about 1490 m a.s.l. with a horizontal length of approximately 100 m in a northern direction and ends in two poorly ventilated cave chambers (Dudashvili and Dudashvilli, 2012). These chambers are overlain by 30–50 m of karstified limestone bedrock. The area around the cave is characterized by typical steppe vegetation, with montane grasslands and shrublands, which change into bare rocks. The thin karstified bedrock above the cave promotes fast infiltration of seasonal rainfall and the marginal vegetation cover reacts in a sensitive fashion to changing moisture conditions.

A meteorological monitoring system for cave temperature and humidity, in service since October 2011, indicates a constant cave air temperature of 15.4°C (Figure 2). This is close to the mean annual air temperatures measured outside the cave (15.1°C) and the mean annual air temperature of 15.2°C recorded between 1959 and 1997 at the Osh meteorological station (875 m) in the Fergana Valley (http://www.ncdc.noaa.gov). Cave humidity during the recent observation period has been constant by 100% (Figure 2).

Figure 2.

Monitoring data of temperature, relative humidity, and atmospheric pressure inside and outside the Uluu-Too cave. Monitoring data of temperature and relative humidity from October 2011 to September 2014 indicate a constant cave air temperature of 15.4°C. This is close to the mean annual air temperatures measured outside the cave (15.1°C) and the mean annual air temperature of 15.2°C recorded between 1959 and 1997 at the Osh meteorological station (875 m a.s.l.). In-cave humidity during the observation period has been constant at 100%. Humidity outside the cave is characterized with dry conditions during the summer/autumn months in comparison with wetter conditions during the winter/spring period. Atmospheric pressure outside the cave indicates an annual cycle of low pressure during the summer in contrast to high pressure during the winter period.

Climate

Rainfall in this region is highly seasonal because of the interaction between the Siberian High and the Westerlies system (and in some more southerly regions the limits of the Indian monsoon) (Figure 1; Aizen et al., 2001; Chen et al., 2008). During winter and early spring, the southwest branch of the Siberian High extends southward, accompanied by a southward shift of the Westerlies, which carry moisture-laden air from the Atlantic, Mediterranean, and Caspian Sea regions (maximum precipitation in March, nearest station Osh and Tashkent, Figure 3). The reverse applies in late spring and summer, when the Siberian High retreats northward and interrupts the influx of moisture into this area, resulting in a very pronounced seasonality in precipitation (see Figure 3). Furthermore, annual reconstruction of monthly air mass trajectories illustrates the remote provenance of moisture between November and June mostly driven by the Westerlies system. In contrast, air mass movements during summer and early autumn (July to October) are limited to the regional scale (Figure 3; Draxler and Hess, 1998).

Figure 3.

Map with climate diagrams from meteorological stations (number 6: precipitation data from Osh via GHCNv2 and temperature data from Osh (http://www.ncdc.noaa.gov)) under the influence of the Westerlies atmospheric circulation system (IAEA, 1992; Morris et al., 2005; Peterson and Vose, 1997). The yellow star indicates Uluu-Too cave. Monthly mean of air temperature (red line), precipitation (blue bars), and oxygen isotope values of precipitation (green line) from meteorological stations in the Westerlies region (Eastern Mediterranean and Central Asia) show a robust positive correlation between air temperature and δ¹⁸O of precipitation, independent from altitude, geographic location, moisture source, and shifts in precipitation seasonality. The numbers behind each site show typical minima (winter) and maxima (summer) δ¹⁸O precipitation values. Under the same seasonal air temperature conditions, it is apparent that the δ¹⁸O of winter/spring precipitation became progressively lower eastward from the Mediterranean and Caspian Sea region to Central Asia, illustrating the ‘continental effect’ (progressive depletion of heavy isotopes with increasing distance from the coast as a consequence of the gradual rainout of the air masses entering the continent; Rozanski et al., 1982). (Map) Monthly reconstruction of air mass trajectories for the year 2014 (NOAA HYSPLIT MODEL using GDAS Meteorological data (Draxler and Hess, 1998; http://ready.arl.noaa.gov/HYSPLIT.php); starting date was set to every 15th of each month for the duration of 300 h backward) illustrates the remote provenance of moisture between November and June. In contrast, air mass movements during summer and early autumn (July to October) are limited to regional scale within few kilometers.

The long-term average annual air temperature (precipitation) at the meteorological station in Osh is 15.2°C (358 mm), with January and July means of about −3°C (35 mm) and 26°C (12 mm), respectively (http://www.ncdc.noaa.gov). Monitoring data of temperature, humidity, and atmospheric pressure outside the cave clearly reflect this seasonality, with low pressure during the summer months (~840 hPa) and high-pressure influence during the winter period (~860 hPa; Figure 2). Temperature and humidity show the same pattern, but with reverse scale. High temperatures correspond with low humidity during the summer period and vice versa during winter (Figure 2). Furthermore, the significantly reduced precipitation and associated heating of the desert lowlands cause local to regional advection, which is responsible for numerous violent dust storms (Schettler et al., 2014; Sorrel et al., 2007).

Materials and methods

The 176-mm-long, broken (cave conservation regulations) stalagmite Uluu-2 was retrieved in October 2011 as a whole sample at the end of the cave around 100 m from the cave entrance. Data for cave temperature and humidity were collected using a HumiLog ‘rugged’ logger (Driesen and Kern, Germany). Outside meteorological parameters (temperature, humidity, and atmospheric pressure) were measured with a HumiBaroLog ‘rugged’ logger (Driesen and Kern, Germany). A total of 24 samples were dated at the University of Minnesota using ²³⁰Th, in order to establish a precise chronology. For analytical details and standards used, we refer to Cheng et al. (2013) (Table 1).

Table 1.

Age dating results from Uluu-2 speleothem.

²³⁰Th dating results from Uluu-2 speleothem (the errors based on 2-sigma calculations)
Sample number	Distance from top (mm)	Distance error (mm)	²³⁸U (ppb)	²³²Th (ppt)	²³⁰Th/²³²Th (atomic ×10^–6)	d²³⁴U* (measured)	²³⁰Th/²³⁸U (activity)	²³⁰Th age (years) (uncorrected)	²³⁰Th age (years) (corrected)	d²³⁴U_Initial** (corrected)	²³⁰Th age (yr BP)*** (corrected)
Uluu2-1	3.8	0.3	522 ± 1	286	53	335.5 ± 2.4	0.0018 ± 0.0001	144 ± 6	132 ± 11	336 ± 2	70 ± 11
Uluu2-2	9.4	0.3	490 ± 1	461	51	323.8 ± 2.3	0.0029 ± 0.0001	241 ± 6	220 ± 16	324 ± 2	158 ± 16
Uluu2-3	17.4	0.3	674 ± 1	847	42	309.6 ± 2.5	0.0032 ± 0.0001	268 ± 4	240 ± 20	310 ± 2	178 ± 20
Uluu2-4	22.4	0.3	527 ± 1	2016	52	377.2 ± 2.3	0.0120 ± 0.0001	957 ± 9	877 ± 58	378 ± 2	815 ± 58
Uluu2-5	23.3	0.2	499.3 ± 0.9	544	177	397.6 ± 2.6	0.0117 ± 0.0001	917 ± 9	895 ± 18	399 ± 3	833 ± 18
Uluu2-6	31.8	0.3	434.5 ± 0.7	242	382	438.8 ± 2.3	0.0129 ± 0.0001	981 ± 8	970 ± 11	440 ± 2	908 ± 11
Uluu2-7	38.2	0.2	451.8 ± 0.7	609	167	426.0 ± 2.4	0.0137 ± 0.0001	1050 ± 9	1022 ± 22	427 ± 2	960 ± 22
Uluu2-8	42.4	0.4	455.2 ± 0.6	295	353	433.0 ± 2.3	0.0139 ± 0.0002	1061 ± 13	1048 ± 16	434 ± 2	986 ± 16
Uluu2-4a	47.8	0.8	454 ± 1	280	420	449 ± 4	0.0158 ± 0.0004	1191 ± 30	1179 ± 32	450 ± 4	1117 ± 32
Uluu2-9	51.0	0.5	395.5 ± 0.6	862	125	426.6 ± 4.6	0.0166 ± 0.0002	1275 ± 13	1230 ± 34	428 ± 5	1168 ± 34
Uluu2-10	57.0	0.5	404.2 ± 0.5	780	151	482.0 ± 2.2	0.0177 ± 0.0001	1312 ± 9	1274 ± 28	484 ± 2	1212 ± 28
Uluu2-11	60.4	0.3	350.2 ± 0.4	251	433	454.0 ± 2.2	0.0188 ± 0.0007	1418 ± 55	1404 ± 56	456 ± 2	1342 ± 56
Uluu2-12	68.9	0.4	415.1 ± 0.7	715	198	436.4 ± 2.3	0.0207 ± 0.0006	1584 ± 46	1549 ± 52	438 ± 2	1487 ± 52
Uluu2-13	76.0	0.4	367.7 ± 0.5	918	142	428.7 ± 2.1	0.0215 ± 0.0007	1650 ± 55	1600 ± 66	431 ± 2	1538 ± 66
Uluu2-14	79.6	0.3	391.4 ± 0.7	370	391	480.5 ± 2.6	0.0224 ± 0.0007	1665 ± 52	1646 ± 54	483 ± 3	1584 ± 54
Uluu2-5a	80.4	1.2	296 ± 1	295	383	431 ± 3	0.0232 ± 0.0004	1782 ± 28	1762 ± 32	433 ± 3	1700 ± 32
Uluu2-15	90.3	0.3	336.7 ± 0.7	332	399	440.3 ± 2.8	0.0238 ± 0.0008	1816 ± 63	1796 ± 64	443 ± 3	1734 ± 64
Uluu2-16	96.7	0.3	393.9 ± 0.7	142	1078	399.4 ± 2.3	0.0236 ± 0.0006	1857 ± 51	1850 ± 51	402 ± 2	1788 ± 51
Uluu2-17	102.9	1.0	388.7 ± 0.5	189	938	398.4 ± 2.0	0.0276 ± 0.0007	2172 ± 56	2162 ± 56	401 ± 2	2100 ± 56
Uluu2-6a	104.5	0.3	429 ± 1	529	373	400 ± 3	0.0279 ± 0.0006	2193 ± 45	2167 ± 48	402 ± 3	2105 ± 48
Uluu2-18	124.9	0.4	363.2 ± 0.3	319	762	595.7 ± 1.7	0.0405 ± 0.0008	2802 ± 54	2786 ± 55	600 ± 2	2724 ± 55
Uluu2-21	137.5	0.5	373.6 ± 0.5	1181	243	466.7 ± 1.9	0.0466 ± 0.0002	3513 ± 19	3451 ± 48	471 ± 2	3389 ± 48
Uluu2-8a	145.1	1.0	426 ± 1	606	597	538 ± 2	0.0515 ± 0.0004	3704 ± 32	3678 ± 37	543 ± 2	3616 ± 37
Uluu2-24	175.0	1.0	297.3 ± 0.4	326	972	524.1 ± 2.2	0.0646 ± 0.0003	4715 ± 22	4694 ± 27	531 ± 2	4632 ± 27

Bold values are the result of the 2-sigma calculations.

d²³⁴U = ([²³⁴U/²³⁸U]_activity – 1) × 1000.

d²³⁴U_initial was calculated based on ²³⁰Th age (T), that is, d²³⁴U_initial = d²³⁴U_measured × e^l234×T. Corrected ²³⁰Th ages assume the initial ²³⁰Th/²³²Th atomic ratio of 4.4 ± 2.2 × 10^–6. Those are the values for a material at secular equilibrium, with the bulk earth ²³²Th/²³⁸U value of 3.8. The errors are arbitrarily assumed to be 50%.

***

BP stands for ‘Before Present’ where the ‘Present’ is defined as the year AD 1950.

Sample preparation for stable isotope analyses was carried out at the Helmholtz Centre Potsdam GFZ, using a ‘Sherline 5410’ vertical milling machine with a special stage and a micrometer-drive to allow precise movement of the speleothem block, for stepwise milling at a resolution of 0.1 mm. Oxygen and carbon isotope analyses (precision of δ¹⁸O and δ¹³C on reference material is <0.06‰) were carried out on 1806 samples using an automated carbonate-extraction system (KIEL IV) interfaced with a MAT253 IRMS (ThermoFisher Scientific). All isotopic ratios were expressed in the delta notation relative to VPDB.

Optical microscopy, SEM imaging, and x-ray diffraction (XRD) were used to verify the mineral structure of the samples. Micro x-ray fluorescence (µXRF) scanning was performed using an EAGLE III XL µXRF system spectrometer (Röntgenanalytik, Germany) at GFZ, applying a 40-kV tube voltage, a 300-µA tube current, a 123-µm spot size, and a 100-µm step size. Element intensities of this nondestructive method are expressed in counts per second (cps). The consistency of replicate profiles of the same scanning line demonstrates the good repeatability of the µXRF measurements. Sulfur intensities have been used relative to calcium (S*1000/Ca) to minimize physical measurement effects.

Water samples for stable isotope analyses of oxygen (δ¹⁸O) and deuterium (δD) were collected between October 2011 and September 2014 from various locations within the cave and were stored in sealed air-tight polypropylene bottles until analysis at the Alfred Wegner Institute for Marine and Polar Research in Potsdam. A Picarro water isotope analyzer (Cavity Ring-Down Spectrometer – L2120i) was used for determination of hydrogen and oxygen isotope composition, following Meyer et al. (2000). Analytical precision of SMOW was better than 0.8‰ for δD and 0.1‰ for δ¹⁸O. Spectral analysis of Uluu-2 δ¹⁸O time-series was performed using the AnalySeries software package (Paillard et al., 1996) with a Blackman–Tukey spectral window. Considering the resolution and age uncertainties (avg. 38 years) of our record, we disregard spectral peaks below 50 years.

Results

Uluu-2 stalagmite

The Uluu-2 stalagmite is characterized by sequences of light beige to brownish gray laminations with large variations in thickness and laminae color (Figure 4). Microscopic observations and SEM diagrams on several selected thin sections, as well as XRD results, show the Uluu-2 speleothem consists of calcite. In general, laminations are thicker and darker in the lower part, changing into thinner and brighter in the upper part.

Figure 4.

Chronology and lithological units: (a) age model of the Uluu-2 stalagmite including a total of 24 samples (see Table 1); red bars depict errors of ²³⁰Th dates (2σ), and (b) image of the 176-mm-long Uluu-2 stalagmite with lithological units and locations of dated samples.

Based on thickness and lamination color and the resolution of the age model, we divided the speleothem into three lithological units (Figure 4). Unit 1 (0–20 mm) is characterized by a bright thick sequence without appreciable brownish gray layers. Unit 2 (20–110 mm) consists of finely laminated successions of brownish gray and light beige layers. Unit 3 (110–176 mm) is composed of several light beige episodes interrupted by dark gray brownish layers of variable thickness.

Chronology and growth rates

The age model yielded 24 precise U-series ages, all in stratigraphic order, which were then interpolated using the COPRA software (Breitenbach et al., 2012) to establish an ~4700-year-long chronology (Figure 4), with chronological uncertainties ranging between 11 and 66 years around an average of 38 years (Table 1).

Growth rates in stalagmites have been related to water supply and/or the temperature of drip water (Fleitmann et al., 2004; Genty et al., 2003) with higher rainfall and/or warmer temperatures resulting in higher growth rates. However, stalagmite growth rates are strongly influenced by pCO₂-level in soil and cave air, and saturation state of the infiltrating water. CO₂ degassing of drip water upon entering low-pCO₂ cave voids leads to oversaturation and carbonate precipitation. Thus, periods with either strong CO₂-gradients between infiltrating water and cave air, or low drip rates will be mirrored by increased carbonate precipitation (Baldini, 2010). The Uluu-2 age model allows constructing a highly resolved series of growth rates. Growth rates vary between ~0.0069 and ~0.327 mm/yr with an average of ~0.094, with significant changes at 100–120 mm. The upper section between 20 and 100 mm shows more than double mean growth rates (~0.07 mm/yr) than the lower part (0.03 mm/yr).

Cave drip water

Analysis of four drip water samples collected in October 2011, May 2012, April 2013, and September 2014 follows the isotopic composition of winter/spring versus summer/autumn rainfall. Seasonal comparison between October 2011 (−8.7‰) and September 2014 (−7.7‰) versus May 2012 (−11.6‰) and April 2013 (−11.3‰) reveals different isotopic compositions, with lower values in April and May, which seemingly track the isotopic seasonality of rainfall (Figure 3).

Variations in stable isotopes (δ¹⁸O, δ¹³C) and µXRF-sulfur signals

The full range of δ¹⁸O in the 4700-year-long Uluu-2 record is 5.2‰, ranging from −8.5‰ to −13.7‰ with a mean of −10.8‰ (Figure 5). Carbon-isotope and S/Ca values indicate a similar trend, ranging between −2‰ and −9‰ for δ¹³C, and 2 and 0.3 for S/Ca. Variations in laminae color and δ¹⁸O, δ¹³C, and S/Ca ratios show strong similarities. Dark sequences correspond with more positive δ¹⁸O and δ¹³C values and bright zones reflect zones with more negative values. The same trend is shown for S/Ca values; higher values correspond with darker, more grayish color zones.

Figure 5.

Lithological units and Uluu-2 stable isotope (δ¹⁸O, δ¹³C) and µXRF-sulfur analyses spanning the last 5000 years. Above: oxygen isotope data (blue line) compared with carbon isotope data (green line). Black horizontal bars are U/Th ages with 2σ error ranges. Below: oxygen isotope data (blue line, 7-point running mean) between AD 1200 and AD −2000 compared with the S/Ca µXRF scanning profile (yellow line, 7-point running mean).

Discussion

How do Uluu-2 stable isotopes and µXRF-S/Ca ratio changes translate into climate signals?

Speleothem isotope modeling studies have shown that the variation in kinetic isotope fractionation, differences in drip rate, and further cave processes are not responsible for the large range observed in the Uluu-2 δ¹⁸O record (5.2‰; Deininger et al., 2012; Dreybrodt, 2008; Mühlinghaus et al., 2009). It is more likely that changes in precipitation characteristics above the cave are responsible for fluctuations in speleothem δ¹⁸O values.

By evaluating isotopes of precipitation in the context of climate parameters in Asia, the amount effect is the dominant factor on monsoonal East Asia, while temperature mainly controls the δ¹⁸O values of precipitation in arid CA in the seasonal cycle (Araguas-Araguas et al., 1998; Yao et al., 2013). In addition, large-scale circulation changes or a shift in the balance of different moisture sources and transport trajectories can result in isotopic shifts over time (Cheng et al., 2012; Dansgaard, 1964; Rozanski et al., 1982). Today, annual rainfall in the Uluu-Too cave region is dominated by Westerlies-controlled transport of moisture from the Mediterranean and Caspian Sea regions to CA. Meteorological stations along a transect from the Eastern Mediterranean and CA (Figure 3) reveal a robust positive correlation between air temperature and δ¹⁸O of precipitation. All stations record extreme seasonal changes in the proportion of winter/spring versus summer/autumn precipitation (Figure 3) with typical minima in δ¹⁸O during winter and maxima in the summer months. If these seasonal controls also determine interannual variations in the isotopic composition of rainfall, then temperature could be a major factor explaining the reconstructed oxygen isotopic variability. However, we assume that the temperature-related effects on centennial to decadal scale at Uluu-2 are likely to be minimal because changes in the temperature-dependent fractionation between calcite and water are small, ~−0.25‰/°C (Friedman and O’Neil, 1977), and insufficient to explain the large range in δ¹⁸O Uluu-2 over the last 4700 years (5.2‰) (Figure 5). Furthermore, the constant cave temperature supports this assumption of minor effect (Figure 2). If the temperature effect is likely to remain minimal on interannual δ¹⁸O Uluu-2 composition, the linkage between climate anomalies and reconstructed δ¹⁸O Uluu-2 variability seems controlled by (1) moisture sources and transport trajectories, (2) seasonality changes, and (3) snowmelt:

Re-calculation of monthly air mass trajectories illustrates the remote provenance of moisture between November and June mostly driven by the Westerlies system (from Atlantic). In contrast, air mass movements during summer and early autumn (July to October) are limited to the regional scale within few kilometers (Figure 3; Draxler and Hess, 1998). Under the same air temperature seasonal conditions, the δ¹⁸O of winter/spring precipitation became progressively lower eastward from the Mediterranean and Caspian Sea region to CA (e.g. December data: Ras Muneef: −6.5‰ in contrast to Bishkek: −17‰), illustrating the progressive depletion of heavy isotopes with increasing distance from the coast as a consequence of the gradual rainout of the air masses (transport via Westerlies) entering the continent ‘continental effect’ (Rozanski et al., 1982).

Seasonality changes could alter the δ¹⁸O values significantly. Wet winters/springs, for example, would result in a shift toward substantially lower δ¹⁸O values in contrast to dry summers/autumns with substantially higher δ¹⁸O values. Intensified winter/spring precipitation tends to increase the winter/spring contribution to the overall recharge budget and hence results in a decrease of the annual mean oxygen isotopic ratio.

Furthermore, we suggest that a high proportional contribution of water derived from snowmelt, after relatively long and wet winters with high amounts of snowfall, can further lead a shift toward more negative δ¹⁸O Uluu-2 values. Constant temperature and humidity in the cave as well as the little amount of annual rainfall suggest that longer snow cover, which would have effects on the isotopic composition, is probably negligible.

Stalagmite carbonate δ¹³C is influenced by the δ¹³C of CO₂ in the soil lying above the cave, the amount of CO₂ available in the soil, soil microbial productivity, CO₂ degassing of drip water, and infiltration rates (Baker et al., 1997; Breitenbach et al., 2015; Genty et al., 2003; Hendy, 1971). The δ¹³C of soil CO₂ is mainly controlled by vegetation type and vegetation density above the cave, with lower values when the plant cover is predominantly of the C₃ type and higher values when C₄ plants predominate. Drier conditions in the past would have enhanced evaporation, restricted plant growth, and reduced the volume of CO₂ in the soil. In contrast, wetter conditions would lead to reduced evaporation, increased vegetation growth, and increased volume of CO₂ in the soil. Drier (wetter) conditions would also lower (enhance) drip rates in the cave, thus allowing for prolonged (reduced) CO₂ degassing and ¹³C-enrichment (depletion) in the fluid. Under conditions of limited (prolonged) exchange between hostrock limestone and infiltrating water, less (more) carbon would contribute from hostrock (δ¹³C_Uluu-Too = 1.5‰) rather than soil CO₂ (δ¹³C of C₃ soil CO₂ is ca. −23‰; McDermott, 2004) because of the lower (higher) levels of soil CO₂ and the likelihood of dissolution under more closed-system (open-system) than open-system (closed-system) dynamics. Higher δ¹³C values in Uluu-2 are interpreted as indicative of drier/warmer conditions, while lower values reflect wetter/cooler conditions at the Uluu-Too region.

Several ‘Hendy tests’ (variability of δ¹⁸O and δ¹³C along an individual laminae; Hendy, 1971) performed at different depths over the entire profile (0.8, 2.5, 3, 5.5, and 9 cm) show only minor δ¹⁸O variations (<0.5‰) and support near-equilibrium conditions during deposition. However, several workers have questioned its reliability for paleoclimatic research (Dorale and Liu, 2009) and advocate replication test. However, this was not possible because of cave conservation. Instead, to further corroborate our interpretation of the Uluu-2 δ¹⁸O record as indicator for moisture variations, we compare it with the sulfur-to-calcite (S/Ca) ratio obtained with µXRF. In general, variations in the sulfur concentrations in speleothems may reflect atmospheric sulfur variability because of sea spray, dust, volcanic activity, and/or industrial pollution (Frisia et al., 2005; Wynn et al., 2010). Sulfur aerosols in CA are mainly derived from gypsum sourced from the Caspian and Aral Sea regions (Kreutz and Sholkovitz, 2000). Enhanced sulfur concentrations in Uluu-2 are therefore likely to reflect increased atmospheric dust transport under drier conditions (Figure 5). Modern variations in dust characteristics on the northern slope of the Tien Shan reveal a complex pattern in seasonality, depending on local and hemispheric circulation dynamics, outwash of atmospheric dust, and the seasonal frequency in dust storms (Schettler et al., 2014) with higher event variability during the summer months (JJAS) and very low dust concentrations during winter months (Schettler et al., 2013, 2014). Low dust concentrations during times of enhanced wind strength (Westerlies) are likely due to enhanced moisture supply, dense vegetation, and/or agglomerated soils. To a remarkable degree, high S/Ca ratios correspond to higher δ¹⁸O values (Figure 5).

These correspondences provide an independent evidence of a broad negative relation between rainfall (δ¹⁸O) and dust input (sulfur), which supports our expectations that the Uluu-2 δ¹⁸O record reflect interannual fluctuations in mean annual precipitation, in particular the winter/spring rainfall contribution. On interannual timescales, extended winter/spring seasons would lead to abnormal wetter conditions and probably to shortened summer/autumn periods and/or denser vegetation, associated with low sulfur input, and vice versa.

Each individual proxy (δ¹⁸O, δ¹³C, and S/Ca, growth rate) shows internal indications that the Uluu-2 δ¹⁸O speleothem record can be used as proxy to indicate moisture variation and to reconstruct environmental changes over the last 5000 years. Under this assumption, the Uluu-2 stalagmite δ¹⁸O record can be interpreted as reflecting centennial variations in mean annual isotopic composition of recharging precipitation related to interannual/decadal variability in air mass dynamics, most probably during an intensified Westerlies regime, and vice versa.

Regional paleoclimate perspective on the Uluu-2 record

When compared with regional paleoclimate records, the Uluu-2 δ¹⁸O record provides insights into precipitation variability at centennial timescales and during abrupt late-Holocene climate events, which are poorly documented by available climate reconstructions from this region (Solomina and Alverson, 2004). The spatial patterns of Holocene moisture evolution in arid CA are very complex and show significant differences in climate development between the western (Kazakhstan, Tajikistan, Uzbekistan, Kyrgyzstan) and the eastern (Mongolia, NW China) parts of CA (Chen et al., 2008; Herzschuh, 2006; Rudaya et al., 2009).

The reconstructions based on the Uluu-2 record suggests that the area was characterized by a dry climate prior to 3900 yr BP, interrupted by a wet episode around 4200 yr BP. Drier conditions recurred between 4000 and 3500 yr BP (Figure 6), followed by a wet interval that lasted 500 years until 3000 yr BP. The re-establishment of wetter conditions occurred at ca. 2500 yr BP after another 500-year-long dry spell. With the exception of two dry events (1150 and 1300 yr BP), the period after 1700 yr BP shows moderate to wet conditions.

Figure 6.

Uluu-2 oxygen isotope data (5-point running mean) with other selected regional climate records over the last ~5000 years: (a) δ¹⁵N of Lake Son-Kul as a proxy for winter precipitation (Lauterbach, personal communication) and/or snowfall intensity (Lauterbach et al., 2014); (b) Lake Karakul pollen samples’ principal component analysis (PCA1) and δ¹⁸O_ostr data (Mischke et al., 2010) reflecting changes in the amount of moisture (PCA1) and long-term changes in the isotopic composition of rainfall (δ¹⁸O_ostr) (Mischke, personal communication) (Taft et al., 2014); (c) Aral Sea lake-level fluctuations (Boroffka, 2010); (d) δ¹⁸O_ostr data from Lake Issyk-Kul reflecting changes between evaporation and water input; and (e) leaf wax carbon isotope record from Lake Karakuli (Aichner et al., 2015).

Other regional climate records from CA are in line with this reconstruction (Figure 6), demonstrating generally similar patterns of Holocene environmental and climate change. We verified the regional significance of our stalagmite profile as precipitation-sensitive record by comparing the δ¹⁸O data with Aral Sea lake level of the last 4700 years (Figure 6), inferred from geomorphic and archeological evidence (Boomer et al., 2009; Boroffka, 2010). To a remarkable degree, lower isotope values coincide with higher lake level, and vice versa. The results confirm that Uluu-2 δ¹⁸O is a robust indicator of regional climate and its large-scale sensitivities. Interestingly, the dry period 4500–3900 yr BP is not reflected in the Aral lake-level record, which can be explained by dating uncertainties, sampling resolution, and/or meltwater input in Lake Aral at higher air temperatures. Pollen records from Hovsgol Lake (Prokopenko et al., 2007) indicate that a pronounced warming may have occurred around 5500 to 4000 yr BP.

However, similar to our Uluu-2 record, in Westerlies-dominated arid China, Bosten Lake experienced shallow lake level during this period (Wünnemann et al., 2006). A very pronounced drought event is also recorded at Lake Balikun (China) (An et al., 2012) and Lake Son-Kul/Kyrgyzstan (Lauterbach et al., 2014) between 4300 and 3800 yr BP (Figure 6). However, the missing indication for this assumed wet period in Lake Aral in other regional records shows that site-specific hydrological conditions, internal thresholds, the sensitivity of proxies, temporal resolution, and dating accuracy are of crucial importance for identifying such climate fluctuations. Uluu-2 climate reconstructions toward drier conditions set in between 4000 to 3500 yr BP. This broadly coincides with low lake levels and relatively dry conditions in the Aral Sea region (Figure 6; Boomer et al., 2000; Boroffka, 2010), distinct phases of reduced winter precipitation in Central Kyrgyzstan (Huang et al., 2014; Lauterbach et al., 2014), drier conditions at Lake Karakuli (Aichner et al., 2015), pollen data from Lake Karakul (Mischke et al., 2010), and evaporation versus water input data reported from Lake Issyk-Kul (Ricketts et al., 2001) (Figure 6). Furthermore, the onset of drier conditions in the Aral Sea region (Boomer et al., 2000; Boroffka, 2010) and significant water level decrease at Lake Van (Reimer et al., 2009) all correspond with the dry event observed at 3500 yr BP at Uluu-Too. An adjacent distinct wetter episode between 3500 and 3000 yr BP in the Uluu-2 record is also shown in records from Lake-Son-Kul (Lauterbach et al., 2014), Lake Karakul (Mischke et al., 2010; Taft et al., 2014), Lake Issyk-Kul (core IK97 11P) (Ricketts et al., 2001), Lake Karakuli (Aichner et al., 2015), and higher water levels of the Aral Sea (Boroffka, 2010) (Figure 6). Drier conditions between 3000 and 2500 BP agree with a significant drop in water levels of Lake Issyk-Kul (Ferronskii et al., 2003), Aral Sea (Boroffka, 2010), Bosten Lake (Mischke and Wünnemann, 2006; Wünnemann et al., 2006), as well as with reduced water inflow (δ¹⁸O_Ostra), pollen indication from Lake Karakul (Mischke et al., 2010; Taft et al., 2014) and reduced winter precipitation at Lake Son-Kul (Lauterbach et al., 2014).

Aral Sea lake-level changes between 2500 and 780 yr BP correspond well with reconstructed moisture dynamics at Uluu-Too (Figure 6). The onset of wetter conditions found in our record at 2100 yr BP also coincides with moisture evolution in the Tien Shan. Reconstructed high-stands of Lake Issyk-Kul around 1800 yr BP (Ricketts et al., 2001) and Lake Manas (Rhodes et al., 1996) as well as wet conditions at Lake Balikun (An et al., 2012) and enhanced winter precipitation in Central Kyrgyzstan (Huang et al., 2014; Lauterbach et al., 2014) correspond with the lowest Uluu-2 δ¹⁸O values over the entire record. This reported wet phase is in agreement with pollen analyses from Lake Karakul (Mischke et al., 2010) and macrophyte productivity reconstruction from Lake Karakuli (Figure 6; Aichner et al., 2015).

Global forcing mechanisms on Central Asian climate

The interplay between the main atmospheric circulation systems in central Asia – (i) the mid-latitude Westerlies, (ii) the Siberian High, and a partly/or marginal – (iii) the Indian Summer Monsoon (ISM), lead to complex interannual climate patterns (Aizen et al., 2001; Chen et al., 2008).

Modern air mass trajectories (Figure 3) show that the annual rainfall is controlled by the mid-latitude Westerlies transport of moisture from the Mediterranean and Caspian Sea region to CA during winter and spring (Figure 3; Aizen et al., 2001). Flow tracks from the Siberian High deliver cool and dry air during winter/spring and seem to have only a minimal contribution to annual rainfall. However, variation of the Siberian High intensity seems crucial for moisture transport to the Uluu-Too region. Comparison of the Uluu-2 record with two Siberian High reconstructions for the last 250 years (tree-ring reconstruction and GISP2 K⁺ ions; D’Arrigo et al., 2005; Mayewski et al., 1997) suggests that the coupling of the Siberian High with the mid-latitude Westerlies system likely contributed significantly to the observed climate variability in the study area (Figure 7). In contrast, the marginal rainfall during the summer months in the Uluu-Too region and the lack of marine salts in Muztag Ata ice cores (Northwest Tibet; Seong et al., 2009) supports that the mid-latitude Westerlies provide the main fraction of present-day precipitation while the ISM is only of minor importance. Variation in strength, duration, and main tracks of the Westerlies and its related movement of the polar front (Machalett et al., 2008) could have influenced the amount of winter and spring moisture which has reached the Uluu-Too region during the last 5000 years. Comparison of the Uluu-2 record with a reconstruction of the NAO index for the last 250 years (Luterbacher et al., 2001) shows that moisture changes in the Uluu-Too region exhibit a similar trend with the reconstructed NAO index. This correspondence corroborates the inferred North Atlantic influence on CA precipitation via the Westerlies. Modern climate data comparisons between arid CA and the NAO during 1950 and 2015 indicate the impact of the NAO leading to more winter/spring precipitation and cooler summer temperature (Figure 8). Interestingly, comparisons with temperature reconstructions from tree rings (summer indicator) in Kyrgyzstan (Esper et al., 2007) and the Northern Hemisphere (Moberg et al., 2005) for the last 250 years also match our reconstructed winter/spring data from Uluu-Too (Figure 7). Similarities between cooling events at Lake Karakuli (Figure 6; Aichner et al., 2015), northern Atlantic ice drift events (Figure 9; Bond et al., 2001), and strengthening phases of the Siberian High (Figure 7; D’Arrigo et al., 2005; Mayewski et al., 1997) are also seen in the Uluu-Too record and support our assumption that the relatively wet episodes were mainly sustained by winter/spring precipitation derived from mid-latitude Westerlies. Earlier studies (Aizen et al., 2006; Lei et al., 2014) suggested that during negative NAO modes the main pathway for westerly winds is located further south, with wetter conditions in the Mediterranean and enhanced amounts of moisture transport into Central Asian regions of the same latitude. Modeling studies propose a more complex interplay between the Eurasian and Pacific circulation systems, leading to generally higher moisture flux into Central-Southwestern Asia during episodes of intensified westerly winds and positive NAO modes (Syed et al., 2010), without inferring re-positioning of the Westerlies tracks. Interestingly, lower winter precipitation reported from the Mediterranean (Figure 9; Fletcher et al., 2012) and contrary more rainfall in the Scandinavian region (Figure 9; Bjune et al., 2005) during episodes of higher winter/spring precipitation at Uluu-Too follow this interpretation that a positive NAO mode could drive moisture delivery through CA.

Figure 7.

Comparison with Uluu-2 δ¹⁸O data and Northern Hemisphere climate records over the last 250 years: (a) Northern Hemisphere temperature anomaly reconstructions (Moberg et al., 2005); (b) reconstruction of the North Atlantic Oscillation index (Luterbacher et al., 2001); (c) tree-ring-based global reconstruction of the Siberian High Index (D’Arrigo et al., 2005); (d) reconstruction of the Siberian High inferred from the Greenland GISP2 potassium concentrations (Mayewski et al., 1997); and (e) West Tien Shan tree-ring data (Esper et al., 2007).

Figure 8.

Modern climate data comparisons between arid Central Asia and the averaged NAO (1950 and 2015) using the KNMI Climate Explorer (http://climexp.knmi.nl). Seasonal precipitation analyses show strongest impact between November and January and May to July for temperature, respectively: (a) correlation between November and January averaged NOAA CPC NAO data (Climate Prediction Center) and the time-series dataset for precipitations (CRU TS3.23) produced by the Climatic Research Unit (CRU) at the University of East Anglia between 1950 and 2014 (University of East Anglia Climatic Research Unit et al., 2008) and (b) correlation between May and July averaged NOAA CPC NAO data (Climate Prediction Center) and the NOAA National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) surface temperature between 1950 and 2015 (Kalnay et al., 1996).

Figure 9.

Uluu-2 oxygen isotope data with other selected records: (a) North Atlantic hematite grains indicating northern hemispheric cooling events (Bond et al., 2001); (b) variation in forest composition (percentage of temperate forest taxa) in the western Mediterranean as proxy for Westerlies-influenced winter precipitation (Fletcher et al., 2012); (c) reconstructed percentage of winter precipitation reflects changes in the intensity of the Westerlies regime (present-day reference is 100%; Bjune et al., 2005); (d) power spectrum on Uluu-2 oxygen isotope data for the period 1100 and 4500 yr BP. Spectral analysis of Uluu-2 speleothem record reveals several periodicities that are significant at the 95% confidence interval (gray dashed line) that closely match the periodicities found in tree-ring ∆¹⁴C record (Stuiver and Braziunas, 1993); and (e) δ¹⁸O speleothem record from Indian Central Himalaya (Sainji cave; Kotlia et al., 2015).

However, that mechanism would be contrary to our correlation within the last 250 years with more winter/spring rainfall at the Uluu-Too region to the negative mode of NAO (Figure 7).

A strengthened Siberian High would push the mid-latitude Westerlies pathways further to the south, resulting in comparably drier conditions in the northern Central Asian regions (e.g. Tien Shan) but wetter conditions in south-western CA (Pamir) (Lei et al., 2014). Recently, a similar mechanism has been invoked to explain the current moisture gradient in the Pamirs, with lower winter/spring precipitation during the positive NAO mode and more northward-oriented Westerlies, while the central Pamirs receive high winter precipitation (Aichner et al., 2015). Based on our data, we argue that the wetter episodes recorded in our speleothem data are controlled by increased winter/spring precipitation caused by strengthened Westerlies. Reconstructed wetter periods are simultaneous within the Central Asian realm (Figure 6) indicating common climate variations within the last 5000 years. Variations in the North Atlantic conditions and Siberian High both appear to align with isotopic variability in our speleothem record, including intensity and pathways of the westerly winds.

On millennial to centennial timescales, our reconstructed regional moisture variability from Uluu-Too show similar long-term climate patterns found in the Central Himalaya. Interestingly, a region influenced by the ISM (Figure 9; Kotlia et al., 2015). These complex interactions between high- and low-latitude atmospheric circulation systems can also be seen in spectral data from the Uluu-2 record. Several periodicities that are significant at the 95% confidence interval (56, 88, 104, 128, 150, 172, 208, 280, 512, and 1500 years; Figure 9) match closely the periodicities found in the tree-ring ∆¹⁴C record (Stuiver and Braziunas, 1993), emphasizing links between high- and low-latitude hydrology on millennial to centennial timescales. Still, the oceanic and atmospheric interactions over Eurasia remain poorly understood and need additional data from well-dated high-resolution proxy reconstructions to unravel the relationships between the large-scale atmospheric circulation systems of CA.

Conclusion

A U-series dated high-resolution speleothem record from Uluu-Too cave from the southern rim of the Fergana Valley/Kyrgyzstan documents environmental changes during the last 5000 years with notable excursions around 3500 yr BP, 2500 yr BP and between 1500 and 1000 yr BP. The location of the cave is ideally suited to identify past shifts in seasonal variations in winter/spring precipitation and provide evidence that the transition between regions of summer versus winter/spring precipitation could have been a key factor for local climate in the past in CA.

The reconstructed centennial to decadal scale moisture evolution in arid CA can be linked with changes in the intensity and pathways of the mid-latitude Westerlies, with substantial impact on winter/spring precipitation. This is regionally synchronous with other Westerlies-dominated paleoclimate records, revealing a hemispheric teleconnection between the North Atlantic and the Central Asian climate system and its effect on winter/spring precipitation in arid CA. We find evidence for a significant influence of the North Atlantic NAO on CA winter/spring precipitation.

However, understanding the underlying climate mechanism in full detail requires further investigations and more archives from the hydrologically sensitive region of arid CA.

Footnotes

Acknowledgements

We are grateful to H Meyer for water isotope analysis, B Richert for µXRF scanning, and R Naumann for XRD measurements. We thank H Oberhänsli and two anonymous reviewers for their constructive comments, which helped us to improve the manuscript.

Funding

This research was supported by the Helmholtz Centre Potsdam (GFZ), the Global Change Observatory of Central Asia (GCO-CA), the PROGRESS research cluster at the University of Potsdam (CW and MS), and the Central Asian Climate Dynamics Project (CADY) which is funded by Germany’s Federal Ministry of Education and Research (BMBF). HC received support from NSFC grants 41230524 and NBRPC 2013CB955902.

References

Aichner

Feakins

Lee

. (2015) High-resolution leaf wax carbon and hydrogen isotopic record of the late-Holocene paleoclimate in arid Central Asia. Climate of the past 11(4): 619–633.

Aizen

Melack

. (2001) Precipitation and atmospheric circulation patterns at mid-latitudes of Asia. International Journal of Climatology 21(5): 535–556.

Aizen

Joswiak

. (2006) Climatic and atmospheric circulation pattern variability from ice-core isotope/geochemistry records (Altai, Tien Shan and Tibet). Annals of Glaciology 43(1): 49–60.

Zhao

. (2012) A high-resolution record of Holocene environmental and climatic changes from Lake Balikun (Xinjiang, China): Implications for central Asia. The Holocene 22(1): 43–52.

Araguas-Araguas

Froehlich

Rozanski

(1998) Stable isotope composition of precipitation over southeast Asia. Journal of Geophysical Research: Atmospheres 103(D22): 28721–28742.

Baker

Ito

Smart

. (1997) Elevated and variable values of 13C in speleothems in a British cave system. Chemical Geology 136(3–4): 263–270.

Baldini

JUL

(2010) Cave atmosphere controls on stalagmite growth rate and palaeoclimate records. Geological Society, London, Special Publications 336(1): 283–294.

Barlow

Cullen

Lyon

(2002) Drought in central and southwest Asia: La Nina, the warm pool, and Indian Ocean precipitation. Journal of Climate 15(7): 697–700.

Bjune

Bakke

Nesje

. (2005) Holocene mean July temperature and winter precipitation in western Norway inferred from palynological and glaciological lake-sediment proxies. The Holocene 15(2): 177–189.

10.

Bond

Kromer

Beer

. (2001) Persistent solar influence on North Atlantic climate during the Holocene. Science 294: 2130–2136.

11.

Boomer

Aladin

Plotnikov

. (2000) The palaeolimnology of the Aral Sea: A review. Quaternary Science Reviews 19(13): 1259–1278.

12.

Boomer

Wunnemann

Mackay

. (2009) Advances in understanding the late-Holocene history of the Aral Sea region. Quaternary International 194(1–2): 79–90.

13.

Boroffka

(2010) Archaeology and its relevance to climate and water level changes: A review. In: Kostianoy

Kosarev

(eds) The Aral Sea Environment. Berlin, Heidelberg: Springer, pp. 283–303.

14.

Breitenbach

SFM

Lechleitner

Meyer

. (2015) Cave ventilation and rainfall signals in dripwater in a monsoonal setting – A monitoring study from NE India. Chemical Geology 402: 111–124.

15.

Breitenbach

SFM

Rehfeld

Goswami

. (2012) COnstructing Proxy Records from Age models (COPRA). Climate of the past 8(5): 1765–1779.

16.

Chase

Pielke

Knaff

. (2000) A comparison of regional trends in 1979–1997 depth-averaged tropospheric temperatures. International Journal of Climatology 20(5): 503–518.

17.

Chen

Yang

. (2008) Holocene moisture evolution in arid central Asia and its out-of-phase relationship with Asian monsoon history. Quaternary Science Reviews 27(3–4): 351–364.

18.

Cheng

Edwards

Shen

. (2013) Improvements in Th-230 dating, Th-230 and U-234 half-life values, and U-Th isotopic measurements by multi-collector inductively coupled plasma mass spectrometry. Earth and Planetary Science Letters 371: 82–91.

19.

Cheng

Zhang

Spotl

. (2012) The climatic cyclicity in semiarid-arid central Asia over the past 500,000 years. Geophysical Research Letters 39(1): 1705.

20.

Dansgaard

(1964) Stable isotopes in precipitation. Tellus 16(4): 436–468.

21.

D’Arrigo

Jacoby

Wilson

. (2005) A reconstructed Siberian High index since A.D. 1599 from Eurasian and North American tree rings. Geophysical Research Letters 32(5): L05705.

22.

De Martino

Carlsson

Rampolla

. (2005) Environment and Security – Transforming Risks into Cooperation, Central Asia (Ferghana/Osh/Khujand Area). Geneva: UNEP, UNDP, OSCE, NATO.

23.

Deininger

Fohlmeister

Scholz

. (2012) Isotope disequilibrium effects: The influence of evaporation and ventilation effects on the carbon and oxygen isotope composition of speleothems – A model approach. Geochimica et Cosmochimica Acta 96: 57–79.

24.

Dorale

Liu

(2009) Limitations of Hendy test criteria in judging the paleoclimatic suitability of speleothems and the need for replication. Journal of Cave and Karst Studies 71(1): 73–80.

25.

Draxler

Hess

(1998) An overview of the HYSPLIT_4 modelling system for trajectories, dispersion and deposition. Australian Meteorological Magazine 47(4): 295–308.

26.

Dreybrodt

(2008) Evolution of the isotopic composition of carbon and oxygen in a calcite precipitating H₂O–CO₂–CaCO₃ solution and the related isotopic composition of calcite in stalagmites. Geochimica et Cosmochimica Acta 72(19): 4712–4724.

27.

Dudashvili

(2012) Caves in the Region of Osh. Bishkek: Speleothem Association Publishing Group, 188 pp.

28.

Esper

Frank

Wilson

RJS

. (2007) Uniform growth trends among central Asian low- and high-elevation juniper tree sites. Trees, Structure and Function 21(2): 141–150.

29.

Ferronskii

Polyakov

Brezgunov

. (2003) Variations in the hydrological regime of Kara-Bogaz-Gol Gulf, Lake Issyk-Kul, and the Aral Sea assessed based on data of bottom sediment studies. Water Resources 30(3): 252–259.

30.

Fleitmann

Burns

Neff

. (2004) Palaeoclimatic interpretation of high-resolution oxygen isotope profiles derived from annually laminated speleothems from Southern Oman. Quaternary Science Reviews 23(7–8): 935–945.

31.

Fletcher

Debret

Sanchez Goñi

(2012) Mid-Holocene emergence of a low-frequency millennial oscillation in western Mediterranean climate: Implications for past dynamics of the North Atlantic atmospheric westerlies. The Holocene 23(2): 153–166.

32.

Friedman

O’Neil

(1977) Compilation of Stable Isotope Fractionation Factors of Geochemical Interest. Professional paper 440KK. Available at: http://pubs.usgs.gov/pp/0440kk/report.pdf.

33.

Frisia

Borsato

Fairchild

. (2005) Variations in atmospheric sulphate recorded in stalagmites by synchrotron micro-XU and XANES analyses. Earth and Planetary Science Letters 235(3–4): 729–740.

34.

Genty

Blamart

Ouahdi

. (2003) Precise dating of Dansgaard-Oeschger climate oscillations in western Europe from stalagmite data. Nature 421(6925): 833–837.

35.

Held

Soden

(2006) Robust responses of the hydrological cycle to global warming. Journal of Climate 19(21): 5686–5699.

36.

Hendy

(1971) Isotopic geochemistry of speleothems – 1. Calculation of effects of different modes of formation on isotopic composition of speleothems and their applicability as palaeoclimatic indicators. Geochimica et Cosmochimica Acta 35(8): 801.

37.

Herzschuh

(2006) Palaeo-moisture evolution in monsoonal Central Asia during the last 50,000 years. Quaternary Science Reviews 25(1–2): 163–178.

38.

Holmes

Cook

Yang

(2009) Climate change over the past 2000 years in Western China. Quaternary International 194(1–2): 91–107.

39.

Huang

Oberhansli

von Suchodoletz

. (2014) Hydrological changes in western Central Asia (Kyrgyzstan) during the Holocene as inferred from a palaeolimnological study in lake Son Kul. Quaternary Science Reviews 103: 134–152.

40.

IAEA (1992) Statistical Treatment of Data on Environmental Isotopes in Precipitation. Technical report no. 331. Vienna: International Atomic Energy Agency.

41.

IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge; New York: Cambridge University Press, 1535 pp.

42.

Kalnay

Kanamitsu

Kistler

. (1996) The NCEP/NCAR 40-year reanalysis project. Bulletin of the American Meteorological Society 77(3): 437–471.

43.

Kotlia

Singh

Joshi

. (2015) Precipitation variability in the Indian Central Himalaya during last ca. 4,000 years inferred from a speleothem record: Impact of Indian Summer Monsoon (ISM) and Westerlies. Quaternary International 371: 244–253.

44.

Kreutz

Sholkovitz

(2000) Major element, rare earth element, and sulfur isotopic composition of a high-elevation firn core: Sources and transport of mineral dust in central Asia. Geochemistry, Geophysics, Geosystems 1: 1048.

45.

Lauterbach

Witt

Plessen

. (2014) Climatic imprint of the mid-latitude Westerlies in the Central Tian Shan of Kyrgyzstan and teleconnections to North Atlantic climate variability during the last 6000 years. The Holocene 24(8): 970–984.

46.

Lei

Tian

Bird

. (2014) A 2540-year record of moisture variations derived from lacustrine sediment (Sasikul Lake) on the Pamir Plateau. The Holocene 24(7): 761–770.

47.

Luterbacher

Xoplaki

Dietrich

. (2001) Extending North Atlantic oscillation reconstructions back to 1500. Atmospheric Science Letters 2(1–4): 114–124.

48.

McDermott

(2004) Palaeo-climate reconstruction from stable isotope variations in speleothems: A review. Quaternary Science Reviews 23(7–8): 901–918.

49.

Machalett

Oches

Frechen

. (2008) Aeolian dust dynamics in central Asia during the Pleistocene: Driven by the long-term migration, seasonality, and permanency of the Asiatic polar front. Geochemistry, Geophysics, Geosystems 9: Q08Q09.

50.

Mathis

Sorrel

Klotz

. (2014) Regional vegetation patterns at lake Son Kul reveal Holocene climatic variability in central Tien Shan (Kyrgyzstan, Central Asia). Quaternary Science Reviews 89: 169–185.

51.

Mayewski

Meeker

Twickler

. (1997) Major features and forcing of high-latitude northern hemisphere atmospheric circulation using a 110,000-year-long glaciochemical series. Journal of Geophysical Research: Oceans 102(C12): 26345–26366.

52.

Meyer

Schonicke

Wand

. (2000) Isotope studies of hydrogen and oxygen in ground ice-experiences with the equilibration technique. Isotopes in Environmental and Health Studies 36(2): 133–149.

53.

Mischke

Wünnemann

(2006) The Holocene salinity history of Bosten Lake (Xinjiang, China) inferred from ostracod species assemblages and shell chemistry: Possible palaeoclimatic implications. Quaternary International 154: 100–112.

54.

Mischke

Rajabov

Mustaeva

. (2010) Modern hydrology and late-Holocene history of Lake Karakul, eastern Pamirs (Tajikistan): A reconnaissance study. Palaeogeography, Palaeoclimatology, Palaeoecology 289(1–4): 10–24.

55.

Moberg

Sonechkin

Holmgren

. (2005) Highly variable Northern Hemisphere temperatures reconstructed from low- and high-resolution proxy data. Nature 433(7026): 613–617.

56.

Morris

George Darling

Gooddy

. (2005) Assessing the extent of induced leakage to an urban aquifer using environmental tracers: An example from Bishkek, capital of Kyrgyzstan, Central Asia. Hydrogeology Journal 14(1–2): 225–243.

57.

Mühlinghaus

Scholz

Mangini

(2009) Modelling fractionation of stable isotopes in stalagmites. Geochimica et Cosmochimica Acta 73(24): 7275–7289.

58.

Paillard

Labeyrie

Yiou

(1996) Macintosh Program performs time-series analysis. Eos, Transactions American Geophysical Union 77(39): 379.

59.

Peterson

Vose

(1997) An overview of the global historical climatology network temperature database. Bulletin of the American Meteorological Society 78(12): 2837–2849.

60.

Prokopenko

Khursevich

Bezrukova

. (2007) Paleoenvironmental proxy records from Lake Hovsgol, Mongolia, and a synthesis of Holocene climate change in the Lake Baikal watershed. Quaternary Research 68(1): 2–17.

61.

Reimer

Landmann

Kempe

(2009) Lake Van, Eastern Anatolia, hydrochemistry and history. Aquatic Geochemistry 15(1–2): 195–222.

62.

Rhodes

Gasse

Lin

. (1996) A late Pleistocene-Holocene lacustrine record from Lake Manas, Zunggar (northern Xinjiang, western China). Palaeogeography, Palaeoclimatology, Palaeoecology 120(1–2): 105–121.

63.

Ricketts

Johnson

Brown

. (2001) The Holocene paleolimnology of Lake Issyk-Kul, Kyrgyzstan: Trace element and stable isotope composition of ostracodes. Palaeogeography, Palaeoclimatology, Palaeoecology 176(1–4): 207–227.

64.

Rozanski

Sonntag

Münich

(1982) Factors controlling stable isotope composition of European precipitation. Tellus 34(2): 142–150.

65.

Rudaya

Tarasov

Dorofeyuk

. (2009) Holocene environments and climate in the Mongolian Altai reconstructed from the Hoton-Nur pollen and diatom records: A step towards better understanding climate dynamics in Central Asia. Quaternary Science Reviews 28(5–6): 540–554.

66.

Schettler

Oberhansli

Stulina

. (2013) Hydrochemical water evolution in the Aral Sea Basin – Part I: Unconfined groundwater of the Amu Darya Delta – Interactions with surface waters. Journal of Hydrology 495: 267–284.

67.

Schettler

Shabunin

Kemnitz

. (2014) Seasonal and diurnal variations in dust characteristics on the northern slopes of the Tien Shan – Grain-size, mineralogy, chemical signatures and isotope composition of attached nitrate. Journal of Asian Earth Sciences 88: 257–276.

68.

Seong

Owen

. (2009) Geomorphology of anomalously high glaciated mountains at the northwestern end of Tibet: Muztag Ata and Kongur Shan. Geomorphology 103(2): 227–250.

69.

Solomina

Alverson

(2004) High latitude Eurasian paleoenvironments: Introduction and synthesis. Palaeogeography, Palaeoclimatology, Palaeoecology 209(1–4): 1–18.

70.

Sorg

Bolch

Stoffel

. (2012) Climate change impacts on glaciers and runoff in Tien Shan (Central Asia). Nature Climate Change 2(10): 725–731.

71.

Sorrel

Oberhänsli

Boroffka

. (2007) Control of wind strength and frequency in the Aral Sea basin during the late-Holocene. Quaternary Research 67(3): 371–382.

72.

Stuiver

Braziunas

(1993) Sun, ocean, climate and atmospheric 14CO2: An evaluation of causal and spectral relationships. The Holocene 3(4): 289–305.

73.

Syed

Giorgi

Pal

. (2010) Regional climate model simulation of winter climate over Central-Southwest Asia, with emphasis on NAO and ENSO effects. International Journal of Climatology 30(2): 220–235.

74.

Taft

Mischke

Wiechert

. (2014) Sclerochronological oxygen and carbon isotope ratios in Radix (Gastropoda) shells indicate changes of glacial meltwater flux and temperature since 4,200 cal yr BP at Lake Karakul, eastern Pamirs (Tajikistan). Journal of Paleolimnology 52(1–2): 27–41.

75.

Thompson

Yao

Davis

. (1997) Tropical climate instability: The last glacial cycle from a Qinghai-Tibetan ice core. Science 276(5320): 1821–1825.

76.

University of East Anglia Climatic Research Unit; Jones

Harris

(2008) Climatic Research Unit (CRU) Time-Series Datasets of Variations in Climate with Variations in Other Phenomena. NCAS British Atmospheric Data Centre. Available at: http://catalogue.ceda.ac.uk/uuid/3f8944800cc48e1cbc29a5ee12d8542d.

77.

Wünnemann

Mischke

Chen

(2006) A Holocene sedimentary record from Bosten Lake, China. Palaeogeography, Palaeoclimatology, Palaeoecology 234(2–4): 223–238.

78.

Wynn

Fairchild

Frisia

. (2010) High-resolution sulphur isotope analysis of speleothem carbonate by secondary ionisation mass spectrometry. Chemical Geology 271(3–4): 101–107.

79.

Yao

Masson-Delmotte

Gao

. (2013) A review of climatic controls on delta o-18 in precipitation over the Tibetan plateau: Observations and simulations. Reviews of Geophysics 51(4): 525–548.

Precipitation evolution of Central Asia during the last 5000 years

Abstract

Keywords

Introduction

Study site

Cave settings

Climate

Materials and methods

Results

Uluu-2 stalagmite

Chronology and growth rates

Cave drip water

Variations in stable isotopes (δ18O, δ13C) and µXRF-sulfur signals

Discussion

How do Uluu-2 stable isotopes and µXRF-S/Ca ratio changes translate into climate signals?

Regional paleoclimate perspective on the Uluu-2 record

Global forcing mechanisms on Central Asian climate

Conclusion

Footnotes

Acknowledgements

Funding

References

Variations in stable isotopes (δ¹⁸O, δ¹³C) and µXRF-sulfur signals