Data Time

fabric of the carbonate. Due to plausible factors con-

Chinese Science Bulletin 2006 Vol. 51 No. 4 441 447

trolling the drip water chemistry and concentration of

DOI: 10.1007/s11434-006-0441-9

carbon dioxide in cave air, changes in thicknesses of

Imprint of solar activity on annual laminae can be related to variations of rainfall or

cave temperature depending on different environments.

Nanjing stalagmite annual Proctor et al.[5] studied the annual growth rate in a sta-

lagmite from a Scottish cave for the past 150 years and

layer thickness sequence dur- found a weak positive correlation with mean annual

ing the Last Glacial Maximum temperature and a high negative correlation with mean

annual precipitation. The growth rate was also sug-

gested to have been driven by changes in precipitation

WU Jiangying, SHAO Xiaohua, KONG Xinggong

over the last 3000 years, and then inferred to be related

WANG Yongjin

to the winter North Atlantic Oscillation strength,

strength of thermohaline circulation and, possibly, solar

College of Geography Science, Nanjing Normal University, Nanjing

output[5]. Tan et al.[6] reconstructed a 2650-year-long

210097, China

Correspondence should be addressed to Wu Jiangying (email: wujiangy-

warm season temperature history from a stalagmite in

***@****.***.** )

Shihua Cave, Beijing based on a comparison between

Abstract A 3000-year-long stalagmite chronology the annual growth thickness and instrumental meteoro-

from Hulu Cave near Nanjing was established by logical records. They further suggested that the recon-

counting annual layers under microscope. Based on structed record is coherent with solar activity at centen-

the 230Th age, this chronology covers the period 24 nial to millennial time scales. Up to now, almost all of

21 kaBP, within the Last Glacial Maximum (LGM). published data involving the relation between stalag-

Two proxies, annual layer thickness and gray level

mite annual-growth rate and climate are limited within

were measured along the growth axis of the stalag-

the late Holocene, because available historical and in-

mite profile in order to establish a high-resolution

strumental meteorological data make it possible to re-

East Asian monsoon history during the LGM. The

veal the climatic significance of annual layer thickness

high correlation coefficient (r = 0.55) between the two

of stalagmite. Recently, a 780-year record of precipita-

proxies suggests that both of them were controlled by

tion has been derived on the basis of a comparison be-

a common factor, possibly reflecting changes in the

tween annual layer thickness and isotope oxygen data

strength of summer monsoon circulation and its pre-

of an Oman stalagmite, confirming that the 18O is a

cipitation. Low frequency variations of the annual

proxy of monsoon precipitation[7, 8]. Here we present a

layer thickness, ranging from centennial to millennial

glacial annual banding record of a stalagmite from

scales, are approximately in agreement with the

10

Hulu Cave, Nanjing, a typical East Asian monsoon in-

Be-flux recorded in the Greenland ice core, indi-

fluenced-area. A high correlation coefficient between

cating that changes in East Asian monsoon strength

might be forced by solar outputs during the LGM pe- annual layer thickness and gray level of the polished

riods. In support of this, Fourier power spectrum surface provides further evidence that changes in an-

analysis of the annual layer thickness showed certain nual layer thickness are related to climate changes at

decadal to centennial-scale cycles that agree well different time scales during the last glaciation. Subse-

with the periodicities of solar activity. quently, we discuss a possible dynamic mechanism for

the climate changes of the East Asian monsoon at sub-

Keywords: Nanjing stalagmite, annual layer sequence, gray level,

millennial time scales.

solar activity, Last Glacial Maximum.

1 Materials and methods

Previous studies on annual growth laminae in sta-

lagmites suggested that it is a useful proxy to investi- A 23-cm long stalagmite (No. 98001), collected

gate annually high-resolution climates[1 4]. The varia-

from the east edge of Hulu Cave, with a columnar

tion of annual laminae thickness has been explained as shape and an upper diameter of 10.5 cm and a 13 cm at

reflecting seasonal changes either in drip water chemis- the base. A parallel alternation of transparent and

try or in cave air carbon dioxide concentration, which milk-white laminae is clearly discernable on the pol-

alter the saturation state and, therefore, the amount and ished profile along the growth axis (Fig. 1). The growth

www.scichina.com www.springerlink.com 441

down to 230 mm were measured under an Olympus

polarizing microscope with CCD video camera con-

nected to a computer, with absolute measured errors of

1 m. A proxy of gray level was measured on a

scanned picture of the polished stalagmite surface taken

with a high-resolution scanner. The data were obtained

using ENVI software. Values of gray level, in a range

from 0 to 255, represent the optical density of reflected

light of scanned images. The higher the gray level

(Glold), the stronger the reflected light[10]. Here we use a

new gray level proxy Gl = 255 Glold, where, the gray

level increases when the measured section becomes

darker.

2 Results

The results of 230Th dates are listed in Table 1. The

difference between the corrected age and uncorrected

age is small, within 5 to 6 years, suggesting that the

initial 230Th-bearing concentration in the formation of

precipitated calcite is negligible and that the two dated

ages are accurate (relative to 2000 AD). Based on the

two dates, an average growth rate and growth duration

Fig.1. Photos of polished surface and micro-structure of stalagmite

98001. (a) Photo of polished surface; ( b) annual layer photo under the covering the dated profile were calculated to be 76.5

microscope; (c) annual layer thickness curve; (d) gray level curve of

mm/ka and 2779 85 years, respectively. Under the

polished surface.

microscope, there is no sign of any hiatuses, suggesting

a continuous deposition of carbonate precipitates with

axis is stable, leading a straight line of the growth core

respect to an annual time scale. The entire profile of the

from the base to top of the stalagmite. As a result, the

stalagmite, covering a total thickness of 232 mm, con-

diameter of 6 cm accumulated on each of lamina kept

tains 2989 bands. The average layer thickness is 77 m,

unchanged, implying that the drip waters have been

identical to the dated growth rate. The number of bands

under a steady hydrological state during the entire

in the dated sample section is 2729 31 (the error from

growth period. This kind of laminae sequence is,

the uncertainty in thickness of the two dating samples):

therefore, a good archive of climatic change. Two sub-

consistent with the growth duration determined with the

samples with 1 mm thickness for TIMS U-series dating

two 230Th dates. The agreement between the band-

were chipped off parallel to growth banding at the top

counting and the U-series dates confirms that the band-

and at the base of the stalagmite, respectively. The

230

ing lamination is of annual. As a result, we constructed

Th dating was conducted in Isotope Laboratory of

an annual-resolution time scale between 24.2 and 21.3

Department of Geology and Geophysics, University of

kaBP by using one of the dates (21345 85 aBP) as an

Minnesota, USA. The chemical treatment procedures

anchor for the annual banding sequence.

and methods were described in ref. [9]. The reported

errors are 2 . Thickness of annual layer from the top The annual layer thickness or the annual growth rate

The results of TIMS 230Th dating for Stalagmite (No. 98001)a)

Table 1

234U 234U0

238 232 230

Th/238U 230 230

U Th

Sample Depth Th age Th age

(10 9) (10 12)

No. (mm) (activity) (aBP) (uncorrected) (aBP) (corrected)

(measured) (corrected)

98001-1 10 424.7 0.6 98 4 147.5 2.2 0.2046 0.0006 213**-**-*****-** 156.7 2.3

98001-2 222.5 884.6 0.9 273 4 175.4 1.1 0.2342 0.0006 241**-**-*****-** 187.8 1.2

a) 230=9.1577 10 6a 1; 234=2.8263 10 6a 1; 238=1.55125 10 10a 1; 234U = ([234U/238U]activity 1) 1000; 234U0 was calculated based on 230

Th age

(T), i.e. 234U0 = 234Umeasured e 234T; Corrected 230Th ages assume the initial 230Th/232Th atomic ratio of 4.4 2.2 10 6.

442 Chinese Science Bulletin Vol. 51 No. 4 February 2006

ARTICLES

(Ra) is characterized by a high variance, ranging from 6 2(b)). In a further enlarged part between 22460 and

to 419 m. More than 50% of the total Ra is between 22200 aBP (Fig. 2(c)), we can identify a high correla-

tion between the two proxies within a resolution of 20

40 and 80 m per year. Fig. 2 shows a 7-year running

years. Correlation analyses (n=2949) yielded a signifi-

averaged time series of Ra curve. Changes in gray level

cantly positive correlation between Ra and Gl (linear

on the polished surface (Gl), ranging from 15 to 90,

correlation coefficient r=0.55). However, some differ-

reflect visible lithological phases on the polished pro-

file, i.e., low value of Gl is associated with a milk- ences between the Ra and Gl curves are observable and

white color calcite interval while high value of Gl is may be possibly attributed to the different resolution

with a transparent calcite section. The Gl time series between the two records. Another explanation for this

correlate well with the Ra curve (Fig. 2(a)). For exam- discrepancy is that cracks on the polished surface lead

ple, Ra increases remarkably at two intervals, from to a noise signal superimposed on the Gl time series.

23550 to 21450 aBP and from 22000 to 21250 aBP. Despite the differences, the high correlation coefficient

Correspondingly, the Gl time series shows a similar between the two proxies suggests that an identical

amplitude increase at the two intervals. In addition, physical process has controlled the variations in the two

both of Ra and Gl decrease abruptly at 23550 aBP. An proxies over time. The two proxies can thus be re-

enlarged part of Fig. 2(a) shows a good agreement of garded as reliable indicators for changes in local envi-

major fluctuations between Gl and Ra time series (Fig. ronmental conditions outside of the cave.

3 Discussion

3.1 Coherence between Ra and Gl and implication for

palaeoclimatology

Changes in Ra could be controlled by complicated

processes, involving precipitation, temperature, CO2

production rate and Ca2+ concentration of dripwater.

Thus, Ra can be interpreted as changes in the amount of

annual precipitation[11 13], mean annual tempera-

ture[14,15] and summer temperature[6] depending on dif-

ferent cave environments.

Here we regarded the Ra proxy from Hulu Cave as

an indicator of precipitation for the following two rea-

sons. First, a stalagmite (No. MSD) 18O record also

from Hulu Cave, reported by Wang et al.[16], displayed

a long-term trend similar to the Ra time series during

the same period studied here, especially an obvious

agreement between the 18O and Ra records during a

period from 24 to 23.5 kaBP. If changes in stalagmite

18O in South China indicate monsoon precipitation, as

suggested from the previous studies[17,18], the Ra record

can be also inferred to indicate variations of monsoon

precipitation. Although the 18O data from our stalag-

mite is limited to a short interval from 22000 to 21180

aBP, Kong et al.[19] it also demonstrated a negative re-

lation between Ra and 18O, further supporting our

interpretation. Second, the same conclusion was de-

Fig. 2. Comparison between gray level of polished surface (Gl) and

rived from a comparison between Ra and 18O records

annual layer thickness (Ra). (a) Data of Ra (black line) and Gl (gray line)

are running averaged by 7 years and 5years, respectively; (b) data of Ra

of stalagmites from an Oman cave, which is an Indian

and Gl are running averaged by 7 years and 5years, respectively (22680

monsoon influenced-site[7]. Alternatively, the Ra time

22200 aBP); (c) data of Ra is running averaged by 7 years and the Gl

series was used as a proxy for changes in cave tem-

time series is row data (224**-***** aBP).

www.scichina.com www.springerlink.com 443

perature. For example, Tan et al.[6] suggested that small reliable, because, under this condition, false informa-

changes in atmospheric temperature can lead to a sig- tion may come from differences in thickness of the thin

nificant changes in soil CO2, which controls amount of section and from a degree the laminae incline to the

dissolved carbonates and subsequently carbonate measured surface. We consider that the laminae bound-

thickness accumulated on the top of active stalagmite. ary should form in an environment quite different from

Therefore, different interpretations for the Ra record that of carbonate precipitate. It can be reasonably in-

would be reasonable, largely depending on different ferred that the boundary would disappear if the dripwa-

cave environments and the dissolving process of soil ter was continuous during stalagmite growth. For ex-

CO2. The modern climates of Nanjing and Beijing are ample, stalagmites from various localities in South

not remarkably different, and both areas are influenced China, where rainfall prevails in all seasons, do not

by the East Asian monsoon. Both sites are character- record continuous annual cycles. Our field investiga-

ized by high rainfall and high temperature during tions over the past several years provide a wealth of

summer. When warm and wet summer monsoon winds data about the hydrological cycle and geological setting

prevail, micro-organisms are more active and increase at Hulu Cave site. The bedrock overlying the Hulu

the partial pressure of soil CO2, leading to an enhance- Cave is thin, resulting in a short storage period for

ment of stalagmite growth[20]. Therefore, we can not seeping water in limestones[22] and subsequently an

exclude the possibility that changes in precipitation are obviously a seasonal alternations of dripwater dis-

the dominant factor controlling the annual growth of charge. Most of the dripwater will stop in the winter

stalagmites. The high coherence between 18O and Ra season (December to February). Dripwater supplied on

the top of stalagmite is an essential condition of car-

record, we observed from Hulu Cave, may suggest that

bonate precipitation. Thus, calcite precipitation only

the Ra time series are related to monsoon precipitation.

occurred under continuously supplied dripwater. Impu-

The Gl, to some extent, supports the idea that the Ra

rities (dark layer) accreted on the surface of stalagmite

time series can be used as a proxy for the East Asian

when dripwater stopped, leading to a boundary between

monsoon precipitation. Fig. 1 illustrates a high correla-

the periods of calcite precipitation. Compared with the

tion between Gl and Ra, implying a close link between

present climate conditions, the glacial period between

the two proxies. First, the observed correlation between

24 and 21 ka ago had longer dry seasons, which would

Ra and Gl depends inherently on different mineral

have been favorable for the formation of annual cycles

concentration in the laminations. When calcite layers/

in Hulu Cave stalagmites. We have found that seven

lamination groups have high concentrations of dark

stalagmites stop growing appear 10 ka ago. Among

materials, both the Ra and Gl will decrease (Fig.

them, five stalagmites, not receiving dripwater, have

1(c)), and are milk-white colored intervals on the pol-

accumulated about a 1 mm thick layer of impurities on

ished section. In contrast, Ra and Gl increase together

their top surface. The other two, although still receiving

in clean and transparent calcite layers/laminations in

drip water (unsaturated water), are clean with minor

one transparent interval on the polished surface. The

impurities. This suggests that impurities would accrete

pattern usually appears to be visible alternations of

on the top of the stalagmites without a supply of drip-

thick/thin lamination groups. Due to a relatively low

water. Longer drier periods would therefore produce

resolution of Gl time series, we identified a good rela-

thicker layers of impurities. Therefore, Gl essentially

tion between the Gl and Ra limited only at multi-de-

reflects changes in the ratio of impurities to pure calcite,

cadal or longer time scales. Therefore, we consider it to

and thus reflect the duration of growth stage/diapause

be a record of reliable decadal and longer-scale fluctua-

in one year. If so, variations of the Gl with time indi-

tions in the monsoon climate changes. Second, the dis-

cate changes in the humidity of the cave environment.

tribution of the dark material may identify their source

in the cave environment. Under the microscope, we 3.2 Comparison between Ra and solar activity proxy

observed that the dark material was concentrated on the

The production rates of cosmogenic nuclides are re-

boundary between two bands. If thickness of the thin

lated to the solar winds and solar activity. Higher pro-

section artificially reduces thinner, a range of dark ma-

duction rates are associated with weaker winds and re-

terials become thinner and finally become a dark stripe.

duced irradiance[23]. At present, two proxies, 14C and

As suggested by Tan et al.[21], the measured gray level

10

Be records, are available and thus can be used as a

data from the thin section under the microscope are not

444 Chinese Science Bulletin Vol. 51 No. 4 February 2006

ARTICLES

proxy of total solar irradiance[24]. For a detailed com-

parison, here we selected the 10Be flux record from the

GISP2 ice core because the 14C data are low-resolution

during the LGM. Fig. 3 illustrates a clear relationship

between Ra in the stalagmite from Hulu Cave and the

10

Be-flux record from GISP2 ice core with a minor age-

offset between them. The age-offset, ranging from 100

to 300 years, was also found between the 230Th dated

Hulu Cave 18O time series and the layer-counted

GISP2 18 O chronology [16]. This suggests that the

GISP2 time scale may be older than its accurate record

by 100 to 300 years. If the GISP2 time scale is shift to

make it 200 years younger, the Ra and 10Be-flux re- Fig. 3. Comparison between Ra (gray line) from stalagmite 98001 and

cords match very well. The GISP2 record has age errors 10

Be flux record (black line) from GISP2 ice core.

up to 1% during the Holocene and the uncertainties

variations at centennial to millennial time scales. We

become larger as the chronology gets older[25]. There-

also noticed that a stalagmite 18O record from Hulu

fore, we can attribute the systematical discrepancy of

Cave[16] correlates well to the Ra and 10Be-flux record.

200 years to the uncertainty of the GISP2 chronology.

Especially during 24 23.5 kaBP, the three different

As shown in Fig. 3, the two records have similar

records display a well-consistent trend. However, the

long-term trends and major peak and trough of oscilla-

18O record does not show the high amplitude oscilla-

tions. In other words, the 10Be flux decreases as the

tions that are in the Ra record between 23.5 and 21

stalagmite Ra values increase, and vice versa. This

kaBP, suggesting that the proxy of Ra is more sensitive

suggests that the monsoon strength increased when so-

to changes in precipitation than the 18O.

lar irradiance was enhanced. This solar activity-mon-

In order to further verify the correlation between the

soon relation can be also observed in a short and

solar activity and the East Asian monsoon, we con-

abruptly changed interval. For example, between 24

ducted spectral analyses of the Ra data. In a red noise

23.5 kaBP, Ra values increase from 50 m to 200 m

test, the dominant cycles passing the test at the confi-

per year as 10Be flux decreases from 0.26 to 0.16 (106

dence level of 95% are 373, 169, 60, 43, 37, 24, 19,

atom cm 2 a 1) within 200 years. The good agreement

13.5, 11 and 7 2 years (Fig. 4). Among them, peri-

between the two records suggests that changes in sta-

odicities of 373 and 169 years are within a length of the

lagmite Ra, as a proxy of East Asian monsoon climate

100 300 year cycles in the tree ring 14C records[26]. In

during the LGM, well reflect solar-induced climate

Fig. 4. The spectrum of stalagmite (No. 98001) Ra. Hamming window: 0.0357, 0.2411, 0.4466, 0.2411, 0.0357; No. of case: 2988. The solid and

dashed lines represent 95% and 99% confidence levels respectively.

www.scichina.com www.springerlink.com 445

ing of the Tibetan Plateau[32]. It is more likely that solar

a range of high frequency, a cycle of 60 years is the

strongest and common in the modern monsoon record. variability leads to changes in atmospheric or oceanic

Agnihotri et al.[27] suggested that the 60-year solar cy- circulation that amplify this initial input. GCM model-

cle might have a direct control on the intensity of the ings of the amplifying mechanisms have focused on the

atmosphere s dynamic response to solar forcing[35,36].

Indian monsoon. 24- and 19-year cycles belong to Hale

cycle in magnetic solar activity. A 11-year cycle, called Those models imply that at the time of solar irradiance

the sunspot cycle, has also been recorded in an annual maximum, a warming of the summer stratosphere was

layer profile of stalagmites from northern Italy[14] and found to strengthen easterly winds, which penetrated

in deep-sea sediments from the eastern Arabian Sea[27]. into the equatorial upper troposphere, causing poleward

shifts in the positions of the subtropical westerly jets,

Cycles at 7 2 years, high-frequency oscillations of

broadening of the tropical Hadley circulations, and

monsoon precipitation, may respond to the modern

poleward shifts of the storm tracks. By this way, the

ENSO event.

solar activity influences the global climate through the

A previous study reconstructed a 2650-year warm

interactions of coupling atmosphere-ocean in tropical

season temperature history inferred from the layer

thickness of a stalagmite from Beijing[28]. That record Pacific at time scales from annual to centennial.

matched well with oscillations in total solar irradiance,

4 Conclusions

indicating that the sun is most likely a direct forcing of

(1) An annual-resolution time scale between 24.2

climate changes at centennial to millennial scales in the

and 21.3 kaBP was established for a stalagmite from

eastern Asian continent. They further suggested that the

Hulu Cave based on precise 230Th dates and annual

solar forcing may be geographically in hemisphere

scale based on the observation of sun-climate link in banding-counting data. It provides a basis for further

the northern Atlantic[28]. A 6000-year high-resolution studying variations of global atmosphere 14C.

18O record of peat plant cellulose from northeastern (2) A high correlation between gray level and annual

China further supports that the solar activity exerts a layer thickness of the stalagmite suggests that the two

large control on the monsoon climate variations at proxies are inherently related to the monsoon climates,

timescales from decades to centuries[29]. Increasing re- supporting the idea that changes in stalagmite annual

cords provided robust evidence for a solar influence on layer thickness well reflect strength of summer mon-

climate during the last million years[27, 30, 31], Middle soon circulation and its precipitation at high frequency

Holocene[32] and even during the entire Holocene[33]. variability.

Two hypotheses have been put forward to explain the (3) Changes in stalagmite annual layer thickness

monsoon climate changes. One attributes such changes with time from Hulu Cave matched well with the 10Be

to reorganizations of the North Atlantic thermohaline flux record from the GISP2 ice core, suggesting a close

circulation and the other to changes in tropical atmos- link between solar irradiance and monsoon climate at

phere-ocean dynamics. Many studies have suggested centennial to millennial time scales. Spectral analyses

that the North Atlantic thermohaline circulation af- of the measured data of annual layer thickness demon-

fected Eurasia snow cover through the westerly jets,

strated that some of cycles emerged in annual banding

leading to a remarkable shift of the monsoon climate

data are close to the well-known periodicities of solar

during the last glaciation. Our stalagmite Ra record

activity, providing further evidence for that decadal to

from Hulu Cave demonstrates that the centennial or

millennial scale variability of East Asian monsoon cli-

shorter scale variability of the monsoon climate was

mate might be forced by sun s energy outputs even

forced by the solar activity even under the Last Glacial

during the Last Glacial Maximum.

Maximum climatic conditions. Although the variations

in solar irradiance are tiny, earth s climate system is

Acknowledgements We thank Dr. Cheng Hai at Isotope

highly sensitive to extremely weak perturbations in the Laboratory of Geology and Geophysics Department, Minne-

sun s energy output[34]. Less clear is the mechanism by sota University, USA for his measurement of TIMS-U series

which solar variability affects the East Asian monsoon. dating. This work was supported by the National Outstanding

Such minor variations in solar irradiance (0.1% 0.3%) Youth Foundation of China (Grant No. 40225007), National

Basic Research Program of China (Grant No. 2004CB720204)

at decadal to centennial time scales are unlikely to have

and Postdoctoral Founded Program of Jiangsu Province.

directly caused significant differences in sensible heat-

446 Chinese Science Bulletin Vol. 51 No. 4 February 2006

ARTICLES

References Nanjing, Chinese Science Bulletin, 2003, 48(5): 480 484.

18. Yuan, D. X., Cheng, H., Edwards, R. L. et al., Timing, Duration,

1. Genty, D., Quinif, Y., Annually laminated sequences in the internal

and Transitions of the Last Interglacial Asian Monsoon, Science,

structure of some Belgian stalagmites-Importance for paleoclima-

2004, 304: 575 578.

tology, Journal of Sedimentary Research, 1996, 66(1): 275 288.

19. Cheng, H., Edwards, R. L., Wang, X. F., Oxygen isotope records of

2. Tan, M., Qin, X. G., Liu, D. S., Microbanding of stalagmite and its

stalagmites from Southern China, Quaternary Sciences (in Chinese),

significance, Journal of Chinese Geography, 1997, 7: 16 25.

2005, 25(2): 157 163.

3. Proctor, C. J., Baker., A., Barnes, W. L. et al., A thousand year 20. Wang, X. Z., Ban, F. M., Pan, G. X., Temporal and spatial variation

speleothem proxy record of North Atlantic climate from Scotland, of cave dripwater geochemistry in Shiha Cave, Beijing, China,

Climate Dynamics, 2000, 16: 815 820. Quaternary Sciences (in Chinese), 2005, 25(2): 258 264.

4. Genty, D., Baker, A., Vokal, B., Intra- and inter-annual growth rate 21. Tan, M., Significant issues of stalagmite layer-climatology, Qua-

of modern stalagmites, Chemical Geology, 2001, 176: 191 212. ternary Sciences (in Chinese), 2005, 25(2): 164 169.

5. Proctor, C. J., Baker, A., Barnes, W. L. et al., A three thousand 22. Wang, Y. J., Kong, X. G., Shao, X. H. et al., Century-scale climatic

year record of North Atlantic climate, Climate Dynamics, 2002, 19: oscillations during the last glacial maximum recorded in a stalag-

449 453. mite from Nanjing, Quaternary Sciences (in Chinese), 2002, 22(3):

6. Tan, M., Liu, D. S., Hou, J. et al., Cyclic rapid warming on centen- 243 251.

nial-scale revealed by a 2650-year stalagmite record of warm sea- 23. Masarik, J., Beer, J., Simulation of particle fluxes and cosmogenic

son temperature, Geophysical Research Letters, 2003, 30(12): nuclide production in the Earth's atmosphere, Journal of Geophys-

1617 1621. ics Research, 1999, 104(D10): 120**-*****.

7. Burns, S. J., Fleitmann, D., Mudelsee, M. et al., A 780-year annu- 24. Bard, E., Raisbeck, G., Yiou, F. et al., Solar irradiance during the

ally resolved record of Indian Ocean monsoon precipitation from a last 1200 years based on cosmogenic nuclides, Tellus, 2000, 52B:

speleothem from south Oman, Journal of Geophysical Research, 985 992.

2002, 107(D20): 4434 4442. 25. Alley, R. B., Shuman, C. A., Meese, D. A. et al., Visual-strati-

8. Fleitmann, D., Burns, S. J., Neff, U. et al., Palaeoclimatic interpre- graphic dating of the GISP2 ice core: Basis, reproducibility, and

tation of high-resolution oxygen isotope profiles derived from an- application, Journal of Geophysical Research, 1997, 102(C12):

nually laminated speleothems from Southern Oman, Quaternary 263**-*****.

Science Reviews, 2004, 23: 935 945. 26. Stuiver, M., Reimer, P. J., Bard, E. et al., INTCAL98 Radiocarbon

9. Edwards, R. L., Chen, J. H., Wasserburg, G. J., 238U- 234U -230Th- age calibration 24000 0 cal BP, Radiocarbon, 1998, 40: 1041

232

Th systematics and the precise measurement of time over the 1083.

past 500000 years, Earth and Planetary Science Letters, 1986/87, 1: 27. Agnihotri, R., Dutta, K., Bhushan, R., Evidence for solar forcing on

175 192. the Indian monsoon during the last millennium, Earth and Planetary

10. Qi, X. G., Liu, T. S., Tan, M. et al., Characteristics of annual lami- Science Letters, 2002, 198: 521 527.

nae gray level variations in a stalagmite from Shihua Cave, Beijing 28. Tan, M., Hou, J. Z., Liu, T. S., Sun-coupled climate connection

and its climatic significance (II), Science in China, Ser. D, 2000, between eastern Asia and northern Atlantic, Geophysical Research

Letters, 2004, 31, L07207, doi: 10.1029/2003GL019085.

43(5): 521 533.

29. Hong, Y. T., Jiang, H. B., Liu, T. S. et al., Response of climate to

11. Railsback, L. B., Brook, G. A., Chen, J. et al., Environmental con-

solar forcing recorded in a 6000-year 18O time-series of Chinese

trols on the petrology of a late Holocene speleothem from Bot-

peat cellulose, The Holocene, 2000, 10 (1): 1 7.

swana with annual layers of aragonite and calcite, Journal of Sedi-

30. Speranza, A., Van Geel, B., Van der Plicht, J., Evidence for solar

mentary Research, 1994, A64(1): 147 155.

forcing of climate change at ca. 850 cal BC from a Czech peat sequ,

12. Brook, G. A., Rafter, M. A., Railsback, L. B. et al., A high-resolu-

Global and Planetary Change, 2002, 35: 51 65.

tion proxy record of rainfall and ENSO since AD 1550 from layer-

31. Paulsen, D. E., Li, H. C., Ku, T. L., Climate variability in central

ing in stalagmites from Anjohibe Cave, Madagascar, The Holocene,

China over the last 1270 years revealed by high-resolution stalag-

1999, 9: 695 705.

mite records, Quaternary Science Reviews, 2003, 22: 691 701.

13. Polyak, V., Asmerom, Y., Late Holocene climate and cultural

32. Neff, U., Burns, S. J., Mangini, A. et al., Strong coherence between

changes in the Southwestern United States, Science, 2001, 294:

solar variability and the monsoon in Oman between 9 and 6 kyr ago,

148 151.

Nature, 2001, 411: 290 293.

14. Frisia, S., Borsato, A., Preto, N. et al., Late Holocene annual

33. Wang, Y. J., Cheng, H., Edwards, R. L. et al., The Holocene Asian

growth in three Alpine stalagmites records the influence of solar

Monsoon: links to solar changes and North Atlantic climate, Sci-

activity and the North Atlantic Oscillation on winter climate, Earth

ence, 2005, 308: 854 857.

and Planetary Science Letters, 2003, 216: 411 424.

34. Bond, G., Kromer, B., Beer, J. et al., Persistent solar influence on

15. Holmgren, K., Tyson, P. D., Moberg, A. et al., A preliminary

North Atlantic climate during the Holocene. Science, 2001, 294:

3000-year regional temperature reconstruction for South Africa,

2130 2136

South African Journal of Science, 2001, 97: 49 51.

35. Haigh, J. D., The impact of solar variability on climate, Science,

16. Wang, Y. J., Cheng, H., Edwards, R. L. et al., A high-resolution

1996, 272: 981 984.

absolute-dated Late Pleistocene monsoon record from Hulu Cave,

36. Shindell, D., Rind, D., Balachandran, N., Solar cycle variability,

China, Science, 2001, 294: 2345 2348.

ozone, and climate, Science, 1999, 284: 305 308.

17. Kong, X. G., Wang, Y. J., Wu, J. Y. et al., A continuous 3000-year

precipitation record of ENSO during LGM from a stalagmite in (Received July 29, 2005; accepted September 12, 2005)

www.scichina.com www.springerlink.com 447

Contact this candidate