ARTICLES

Chinese Science Bulletin © 2007

SCIENCE IN CHINA PRESS

Springer

Changes in annual accumulation retrieved from Geladaindong ice core and its relationship to atmospheric circulation over the Tibetan Plateau ZHANG YongJun1,2, KANG ShiChang1,3†, QIN DaHe3, GRIGHOLM Bjorn4 & MAYEWSKI Paul A.4 1

Institute of Tibetan Plateau Research, Chinese Academy of Sciences (CAS), Beijing 100085, China; Graduate School of CAS, Beijing 100049, China; 3 National Key Laboratory of Cryospheric Sciences, CAS, Lanzhou 730000, China; 4 Climate Change Institute, University of Maine, Orono, ME 04469, USA

Annual accumulation records covering 1935 to 2004 were reconstructed using Geladaindong ice core in the source of Yangtze River. A significant positive correlation between annual accumulation and precipitation from nearby meteorological stations was found, suggesting ice core accumulation could be taken as a precipitation proxy in the region. In the past 70 years, precipitation in the Geladaindong region was low from 1930s to early 1960s, and the lowest value occurred in the later 1950s. Since 1960s, precipitation increased dramatically and reached the maximum around 1980s, then decreased slightly in 1990s. By using Mann-Kendall rank statistical test method, a change point for precipitation was determined in 1967. Analysis of the atmospheric circulation over the Tibetan Plateau suggested that, compared with the southwest wind during the low precipitation period (before 1967), it extended about 2 latitudes northward during high precipitation period (after 1967). Moreover, during the high precipitation, the trough over the Bal Karshi Lake was also enhanced, and both the meridional wind and vapor transporting displayed a remarkable aggrandizement. Geladaindong ice core, accumulation, atmospheric circulation, change point

Air temperature and precipitation are two significant factors in the study of climate change. Tibetan Plateau (TP) is regarded as a sensitive area to climate change, however, due to the sparse meteorological stations and short history of observed data, understanding of climate change in the region is limited. Recently, great progresses have been made in reconstructing air tempera― ― ture[1 5] and precipitation change[6 13] by using the ice core δ 18O and accumulation records, respectively. Previous work suggested that the ice core accumulation records from not only Guliya ice cap[5,6,11,12] and Dunde ― ice cap[11 13] in the northern TP but also Dasuopu Gla― cier[6 9] and East Rongbu glacier[6,10] in the southern TP could represent the regional precipitation change, and also be closely related with the atmospheric circulation. Accumulation records from the Dasuopu ice core were www.scichina.com

www.springerlink.com

related with temperature differences between atmosphere and ocean[7,8], too. Mt. Geladaindong, the source of Yangtze River, combined with climate boundary between the southern and northern TP[14,15], provides us a unique opportunity to investigate climate change in the central TP. Therefore, further expanding the knowledge of climate change in this region can establish a base for researches related to environmental variations, and be helpful for farming/ pasturing and ecology repairing in the region. In addiReceived April 10, 2007; accepted August 14, 2007 doi: 10.1007/s11434-007-0439-y † Corresponding author (email: [email protected]) Supported by the National Natural Science Foundation of China (Grant No. 40401054 and 40121101), the National Basic Research Program of China (Grant No. 2005CB422004), the “Talent Project” and Innovation Project of Chinese Academy of Sciences (Grant Nos. KZCX3-SW-339 and 334), and Dean Foundation of CAS

Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266

GEOGRAPHY

2

tion, it can provide opportunities to contribute to the economic and social development in the downstream regions of Yangtze River. For lack of observed data in the Yangtze River source, ice core records can be taken as one of the most ideal proxies for paleoclimate reconstruction. In this paper, the ice core annual accumulation (An) is used to reconstruct the precipitation history for the past 70 years. The abrupt change point of An was determined in 1967 using Mann-Kendall rank statistical ― test method[16 18]. Differences of atmospheric circulation between 1948―1967 and 1968―2004 were discussed, too.

1 Time series of An Mt. Geladaindong with an elevation of 6621 m a.s.l., located in the central TP, is the summit of Tanggula Range, and also the headstream of Yangtze River. In November, 2005, three ice cores (74, 147 and 26 m) were recovered during Sino-US cooperation expedition from the flat firn basin in the Guoqu glacier, northern slope of Mt. Geladaindong[3]. In this paper, the 74 m ice core records (33°34′37.8″N, 91°10′35.3″E, 5720 m a.s.l.) is analyzed (Figure 1). Detailed ice core sampling and dating have been discussed in our previous work[3]. Using the seasonality of δ 18O and major ions with a reference layer of β activity, the ice core dating was performed. Based on the depth-age relationship and density of the core, the time series of An was reconstructed for the period 1935―2004. Here we focus on the reconstructed time series of An (1935―2004) by using the upper 47 m of the core and its relationship to the atmospheric circulation.

Figure 2 Variations of accumulation in the Geladaindong ice core for recent 70 years. Dash line represents the 3-year smoothing average.

1950s. Since the 1960s, the series displays a dramatic rising trend. The values of An from the late 1960s to the early 1990s are about 1.5 times of those from the 1930s to 1950s. After the 1990s the series shows a downward trend. Based on Mann-Kendall rank statistical test method, the series increasing rate in the past 70 years is 29.3 mm/10 a with significance level at 0.05. Comparing Geladaindong An time series with observational annual precipitation series from nearby meteorological stations including Wudaoliang, Toutuohe, Bange, Amdo and Naqu, we find the ice core An has close correlation with Bange annual precipitation. Precipitation of Geladaindong may be mainly caused by the moisture transported southwesterly, and Bange is located at the southwest of Geladaindong. Figure 3 provides the detailed comparison between Geladaindong An and annual precipitation at Bange. These two time series share a similar trend (solid lines in Figure 3), and the correlation coefficient is 0.47 (P<0.001). Due to the dating error (±1 a)[3], we investigate the 3-year running average correlation, suggesting a significant positive correlation between two series (r = 0.66, P<0.001). Hence the reconstructed An can reflect the regional precipitation well,

Figure 1 Drilling site of Geladaindong ice core and location map of nearby meteorological stations.

Figure 2 shows the variations of An during the last 70 years. Values of An are very low from the 1930s to the early 1960s, and the lowest value occur in the later 3262

Figure 3 Comparisons of accumulation anomaly in the Geladaindong ice core with observed annual precipitation anomaly in Bange.

ZHANG YongJun et al. Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266

2 An abrupt change and its relationship to atmospheric circulation Figure 4 shows the Mann-Kendall statistical curve of Geladaindong An. A clear rising trend since the 1960s of Geladaindong An can be seen from UF curve. The rising trend reaches to the 0.05 confidence level after 1980s and to 0.001 confidence level in the mid-1980s, indicating a significant rising trend. The UF curve crosses the UB curve in 1967, thus we can conclude that this rising trend from the 1960s is an abrupt change and the change point is in 1967. Yan et al.[20] pointed out that there was

Figure 5

ARTICLES Figure 4 MK statistic curve of accumulation in the Geladaindong ice core. Beeline represents significance level at 0.05.

an abrupt change of precipitation happening in latitudinal band between the Equator and 30°N, in which a drought trend, a contrary trend with this zonal region, was found in the south and north of this zonal region. Geladaindong is located in the north of this zonal region, and wet abrupt change in 1967 is found by MannKendall method. This inverse change suggests that the abrupt change of An in Geladaindong can reflect this variation in large scale. Summer precipitation (June―September) in TP accounts for 80%―90% of the whole year[21,22]. To better understand the atmospheric circulation difference over Geladaindong between two periods separated by the abrupt year, the 500 hPa mean flow fields, using the NCEP/NCAR reanalysis data[23], during the period of 1948―1967 and 1968―2004 are presented (Figure 5). Westerly dominates this region during the period of 1968―2004 (Figure 5(a)), but Geladaindong is con-

Flow fields at 500 hPa during June―September. (a) 1948―1967; (b) 1968―2004.

ZHANG YongJun et al. Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266

3263

GEOGRAPHY

and thus can be taken as a precipitation proxy. Previous research indicated that the precipitation mechanism in the southern TP was different from the northern TP, and the trend of Dasuopu An in the southern TP was opposite to Guliya and Dunde An in the northern TP. An of Geladaindong presents a distinct trend with that of Dasuopu and Guliya. An values of Geladaindong and Guliya are lowest in early 1960s, then present a rapid rise, while An values of Geladaindong and Dasuopu tend to decrease after 1990s. Ice core records show that the An values are 400 mm, 600 mm, 700 mm for Guliya, Geladiandong, Dasupu, respectively, indicating the different precipitation mechanisms for southern and northern TP. Geladaindong, located at the northern edge of the southwest monsoon, is influenced by both southwest monsoon and inner circulation over the TP[19]. To better understand the Geladaindong precipitation variation, the atmospheric circulation over TP is also analyzed and will be discussed at the following part.

trolled by southwest monsoon during the period of 1948―1967 (Figure 5(b)). The southwest monsoon extends about 2 latitudes northward further during the period of high accumulation. Therefore, the high precipitation after 1967 in Geladaindong may be caused by the northern extending of southwest monsoon. Figure 6 presents the difference of geopotential height fields at 500 hPa between 1968―2004 and 1948―1967. As the figure shows, the values in the whole field are all positive, and the center for high value, where the highest value is above 33 gmp, is located at the nearby region of Bal Karshi Lake. In this field, the difference between the maximum value and the value of Geladaindong is 13 gpm. This reflects that the trough over Bal Karshi Lake is very clear when high precipitation occurs in Geladaindong. By analyzing 500 hPa geopotential height anomaly fields of 5 summer high precipitation years, Shi Xinghe et al.[24,25] suggested when high precipitation year in the northern and central TP, characteristics of circulation were featured by enhanced polar eddy, the major body of which leaned to west hemisphere, and remarkably enhanced trough over Ridge over among the Ural Mount, the Bar Karshi Lake and Baikal Lake. Under this atmospheric circulation, the cold air arrives at the TP, which lead by the trough over Bal Karshi Lake, and intersects the warm atmosphere over the main body of TP. This will result in high precipitation in northern and central TP. To evaluate the meridional wind and water vapor of atmosphere influencing the precipitation in TP, we present the difference of meridional wind fields at 500 hPa

(Figure 7) and difference of atmospheric water transport fields at 500 hPa for 1968―2004 and 1948―1967 (Figure 8). Figure 7 shows that the meridional wind difference between these two periods is positive, and the central value is above 0.5 m/s. This indicates that the meridional wind has a rising trend after 1967, and the meridional water vapor transportation, which offers the dynamic condition for water vapor transportation in Geladaindong, is also enforced. As the figure shows, this field of water vapor transport is a positive field, and the value is 100 g/(m·s) in Geladaindong. This contributes to higher precipitation in this region.

Figure 6 Difference of geopotential height fields at 500 hPa between 1968―2004 and 1948―1967.

Figure 8 Difference of atmospheric water transport fields at 500 hPa between 1968―2004 and 1948―1967.

3264

Figure 7 Difference of meridional wind fields at 500 hPa between 1968―2004 and 1948―1967.

ZHANG YongJun et al. Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266

1

The authors would like to thank all members of 2005 Sino-US Cooperative Expatiation Team, and special thanks are given to Q. Zhang, Z. Cong, F. Chen, Q. Ye, K. Susan, S. Sharon, and D. Qu for their assistance in ice coring and laboratory work.

Yao T D, Guo X T, Thompson L, et al. δ18O record and temperature

years in India. Sci China Ser D-Earth Sci (in Chinese), 2000, 30(6):

change over the past 100 years in ice cores on the Tibetan Plateau. Sci China Ser D-Earth Sci, 2006, 49(1): 1―9 2

619―627 10

Yao T D, Li Z X, Thompson L, et al. δ18O records from Tibetan ice

3

Kang S C, Zhang Y J, Qin D H, et al. Recent temperature increase

Glacier, Mt. Qomolangm (Everest). Ann Glaciol, 2002, 35: 278―284 11

4

Wang N, Yao T, Pu J, et al. Variations in air temperature during the last

47(20): 1746―1749 12

5

Yao T D, Qin D H, Tian L D, et al. Temperature and precipitation

Chinese), 1999, 44(21): 2336―2341 13

6

Ren J W, Qin D H, Kang S C, et al. Glacier variations and climate

16―21 14

recycling. Sci China Ser D-Earth Sci, 2001, 44(9): 790―797

49(1): 65―69 Duan K Q, Wang N L, Pu J C. The India monsoon abrupt change re-

15

Chin J Atmos Sci (in Chinese), 2003, 27(2): 157―170

Duan K Q, Yao T D, Sun W Z, et al. India monsoon variability in the Himalaya since AD 1800. Ann Glaciol, 2002, 35: 437―442

9

16

Kendall M G, Gibbons J D. Rank Correlation Methods, 5th ed. Lon-

17

Fu C B, Wang Q. The definition and detection of the abrupt climatic

don: Edward Arnold, 1981

Yao T D, Duan K Q, Tian L D, et al. Accumulation recorded in Dasuopu ice core and the precipitation variation during the past 400

Wei Z G, Huang R H, Dong W J. Interannual and interdecadal variations of air temperature and precipitation over the Tibetan Plateau.

corded by Dasuopu ice core. Chin Sci Bull, 2001, 46(24): 2069―2073 8

Tian L D, Yao T D, Sun W Z, et al. Relationship between δ D and δ 18O in precipitation on north and south of the Tibetan Plateau and moisture

warming and drying in the central Himalayas. Chin Sci Bull, 2004, 7

Yao T D, Thompson L G. Trends and features of climatic changes in the past 5000 years by the Dunde ice core. Ann Glaciol, 1992, 25:

change during the past 2000 years in Tibetan Plateau-Guliya ice core record. Sci China Ser D-Earth (in Chinese), 1996, 26(4): 348―353

Hou S G, Qin D H, Wake C P, et al. Net accumulation change recorded by ice core and its climatic significances in Everest. Chin Sci Bull (in

100 years revealed by δ18O in the Malan ice core from the Tibetan Plateau. Chin Sci Bull, 2003, 48(19): 2134―1238

Hou S G, Qin D H, Yao T D, et al. Recent change of the ice core accumulation rates on the Qinghai-Tibetan Plateau. Chin Sci Bull, 2002,

recorded in an ice core in the source region of Yangtze River. Chin Sci Bull, 2007, 52(6): 825―831

Qin D H, Hou S G, Zhang D Q, et al. Preliminary results from the chemical records of an 80m ice core recovered from the East Rongbuk

cores reveal differences in climatic change. Ann Glaciol, 2006, 43: 1―7

ARTICLES

A 70-year variation of An has been recovered using an ice core record from Mt. Geladaindong. An is correlated significantly with the precipitation from Bange meteorological station, and thus can be used as a proxy of precipitation in Geladaindong. Hence, the precipitation variation in Geladaindong is discussed during the period of 1935―2004. The results show that there is transformation from low to high accumulation period during the past 70 years. Geladaindong An was low from the 1930s to early 1960s, with the lowest value occurring at the end of 1950s. Since the 1960s, An increased dramatically and reached a maximum around the 1980s, then decreased slightly in the 1990s. Previous work indicated that the precipitation mechanism in the southern TP was different from that in the northern TP. Comparing An variation of Geladaindong and Dasuopu in southern TP as well as Guliya in northern TP, we found that An of Geladaindong presented a different trend from that of Dasuopu and Guliya. This indicated that the precipitation mechanism in Geladaindong was different from that in the northern TP and southern TP. An abrupt change point for An from the period of lower value to higher value determined by Mann-Kend-

all rank statistical test method was in 1967. The atmospheric circulation features were analyzed, and results showed that during the high precipitation period (1968―2004), the southwest summer wind extended about 2 latitudes northward, and the trough over the Bal Karshi Lake was enhanced, and the boosting meridional wind and vapor transport over the Tibetan Plateau were also remarkable. Ye et al.[26] suggested that glacier was retreating in Mt. Geladaindong during 1969―2002, this retreating trend has been becoming more obvious since 1990. Kang et al.[3] confessed this retreating agreed well with temperature increase in this region. Our results showed that not only the increasing temperature but also the decreasing precipitation from the 1990s contributed to the glacier retreat. However, it is hard to distinguish the quantitative contribution of the two major factors to the shrinkage of glacier; but that which factor is dominant is still a puzzle. Further investigation is required to reveal their inner-relationship and influence on the glacier.

ZHANG YongJun et al. Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266

3265

GEOGRAPHY

3 Conclusions and discussion

change. Chin J Atmos Sci (in Chinese), 1992, 16(4): 482―493 18

Chinese). Beijing: China Meteorological Press, 1999. 69―72 19

regions of the Yellow and Yangtze Rivers. J Glaciol Geocryol (in

Wei F Y. Climatic Statistic Diagnosis and Forecast Technology (in

Chinese), 2006, 28(1): 8―14 23

Yao T D, Pu J C, Liu J S. Research on Ice Core Climate in Tanggula Mountain. In: Glacier Climate and Environment in Tibetan Plateau (in

reanalysis project. Bull Amer Meteor Soc, 1996, 77: 437―471 24

Chinese). Beijing: Science Press, 1993. 16―20 20

Shi X H, Qin N S, Ma Y C, et al. Analyses on Qinghai rainy synoptic situation precipitation in summer of 1999. Plat Meteor (in Chinese),

Yan Z W, Ji J J, Ye D Z. The abrupt change of north hemisphere climate in 1960sⅠ: precipitation and temperature change. Sci China Ser

Kalnay E, Kanamitsu M, Kistler R, et al. The NCEP/NCAR 40 years

2002, 21(2): 222―224 25

Wang J S. Qinghai Weather and Climate (in Chinese). Beijing: China

26

Ye Q H, Kang S C, Chen F, et al. Monitoring glacier variations on

B-Chem (in Chinese), 1990, (1): 97―103

Meteorological Press, 2004. 137―342

21

Dai J X. The Tibetan Plateau Climate (in Chinese). Beijing: China Meteorological Press, 1990. 171―197

Geladandong mountain, central Tibetan Plateau, from 1969 to 2002

22

Wang K L, Cheng G, Ding Y, et al. Characteristics of water vapor

using remote-sensing and GIS technologies. J Glaciol, 2006, 52(179):

transport and atmospheric circulation for precipitation over the source

537―545

3266

ZHANG YongJun et al. Chinese Science Bulletin | December 2007 | vol. 52 | no. 23 | 3261-3266