Chapter 1 Introduction
1.3 Australian-Indonesian monsoon
Dynamics of AIM, baring strong connection with East Asian Monsoon (EAM)
system (Fig. 1-2), since the last glacial are rather less studied. Paleoclimate model
experiments (Wyrwoll et al., 2007), paleoproductivity data (Holbourn et al., 2005) and
pollen records (Kershaw, van der Kaars and Moss 2003) revealed that variation of SH
summer insolation influence Australian-Indonesian summer monsoon (AISM) rainfall
on orbital timescale. In contrast, Miller et al. (2005) suggested that the dominant AISM
forcing is the strength of the East Asian winter monsoon (Miller et al., 2005), which is
governed by the strength of air outflowing from the semi-permanent high-pressure
system over Siberia. During the Holocene epoch, the mechanisms that drive millennial-
and centennial- scale events is poorly known. The current available coral records
(Abram et al., 2007; Tudhope et al., 2001) and ocean sediment core records (Stott et al.,
2004; Visser, Thunell and Stott 2003) only provide limited information. Model
experiments simulated perturbation of regional atmospheric circulation and
millennial-scale events since the LGM (DiNezio et al., 2013; Tokinaga et al., 2012;
Zhang and Delworth, 2005), revealing significant influence of Sunda Shelf exposure on
AIM dynamics.
Previous studies have suggested that (1) the AIM system was dominated by
insolation intensity and/or sea-level on orbital time scale (Griffiths et al., 2009; Miller et
4
al., 2005), (2) exposure of the Sunda Shelf may profoundly affect regional climate, and
(DiNezio et al., 2013; Tokinaga et al., 2012) and (3) regional climate system was rather
heterogeneous than previous thought (DiNezio et al., 2013; Mohtadi et al., 2011).
Located at the modern southern border of Austral winter ITCZ, hydroclimate condition
of East Timor is dominantly affected by AM and AIM dynamics (An et al., 2000;
Griffiths et al., 2009). Therefore, by acquiring high-resolution proxy records from East
Timor stalagmites, our goal is to clarify how the AIM system responds to regional and
global changes, and have further insight on teleconnection between North Atlantic
perturbation and low-latitude climate system since the LGM.
Figure 1-2. Main atmospheric features in the modern western Pacific. Schematic view of major atmospheric system during austral winter (June to August) (a) and austral summer (December to Februarys) (b). The dominant flows in the middle and low troposphere are indicated by closed and open arrows, respectively (An et al., 2000).
a. b.
5
Chapter 2 Regional settings and methods
2.1 Study site and research material
2.1.1 Location of Lekiraka cave
Stalagmites were obtained from Lekiraka cave, Ossu, Viqueque, East Timor (Fig
2-2). This cave is located at 8o47’10.8”S, 126o23’31.1”E; 626 m above sea level, with
limestone cap layer about 20m. Lekiraka cave is about 650 km east to the Liang Luar
cave (Flores, Indonesia, 8o32’S, 120o26’E; 550 m above sea level) (Griffiths et al.,
2009).
Figure 2-1. Map showing cave and marine sediment core locations. Stars denote the locations of Lekiraka cave (yellow), Liang Luar cave (orange) (Griffiths et al., 2009), Gunung Buda (white) (Partin et al., 2007), and sediment core GeoB10053-7 (red) (Mohtadi et al., 2011).
6
Figure 2-2. Stalagmite samples, from the cave to the laboratory. (a) Stalagmites in the Lekiraka cave, highlighted by white circle. (b) Unprocessed stalagmite samples. (c) Stalagmite samples that were halved, polished, ready for subsampling.
a. b.
c.
7
2.1.2 Regional settings
The air temperature of Lekiraka cave is 25.3oC, with relative humidity of 97%
(August 3rd, 2011).The annual mean temperature of Ossu, the cave site is 26.1oC, with
mean annual rainfall of 1948 mm recorded by the Ossu weather station, for ~57%
falling during austral summer monsoon season (December–March) and ~17% falling
during austral winter monsoon season (June–September) (Fig. 2-3).
Figure 2-3. Average rainfall amount / temperature of Ossu. Long-term (1952-1974 AD) monthly average rainfall amount (grey bars) and temperature (black circles) of Ossu weather station. Data was obtained from Timor-Leste Meteorology Center.
8
2.2 Experiments
2.2.1 Subsampling
Stalagmite were halved and polished. For U-Th dating, powdered samples, 50-120
mg each, were milled along the growth layer. For stable oxygen and carbon isotope time series, powder samples, 50-100 Pg each, were milled along the growth axis. For ‘Hendy
Test’ (Hendy et al., 1971), powdered samples, 50-100 Pg each, were milled along the
coeval growth layers with interval of 1 mm. All sub-sampling procedure was performed
in class-10000 sub-sampling room and class-100 benches.
2.2.2 Labwares for U-Th dating chemical procedure
Labware used in this research included Teflon beakers, bottles, and columns;
PE/PP bottles, centrifuge tubes, vials, pipette tips, and porous PP column frits. In order
to reduce chemical blank during U-Th isotope measurement, all labwares were
acid-cleaned (Shen et al., 2003) prior to U-Th dating chemical procedure (Shen et al.,
2003; Shen et al., 2012).
2.2.3 U-Th dating chemical procedure
For reduction of potential containment, the chemical procedure for U-Th dating
was performed in a class-10,000 clean room with independent class-100 benches and
hoods. U-Th dating chemical procedure was developed in HISPEC, Department of
Geosciences, National Taiwan University (Shen et al., 2003).
9
Sample digestion
1. Stalagmite powdered sample was weighed and preserved in a 30-ml Teflon beaker.
2. Sample was covered with pure water, and dissolved gradually by adding appropriate
drops of 14N HNO3.
3. Appropriate amount of 229Th-233U-236U spike solution was added and weighed.
4. Ten drops of HClO4 were added for removal of organic matters.
5. One drop of FeCl2 was added for iron co-precipitation.
6. The sample solution was refluxed at 200oC on a hotplate over 12 hours to achieve
complete U-Th isotopic equilibrium and decomposition of organic matters.
7. The sample solution was dried at 250 oC.
Iron co-precipitation and centrifugation
1. Dehydrated sample was covered with 1-2 ml pure H2O, and dissolved with 1-2 ml 2
N HCl, then transferred from Teflon beaker into a PE centrifuge tube. The Teflon
beaker was cleaned with dilute aqua regia.
2. Appropriate amount of NH4OH was gradually added to the sample solution, until
Fe(OH)3 precipitate (for uranium and thorium co-precipitation).
3. Suspension was then separated by centrifugation and discarded, and residual was
rinsed by appropriate amount of pure H2O. This step was repeated three times.
4. Ten drops of 14 N HNO3 were added to dissolve the residual, and transferred back
10
into the original Teflon beaker.
5. Two drops of HClO4 were added to the sample solution and dried at 250 oC.
6. One drop of 14 N HNO3 was added to dissolve the dehydrated sample, and then dried
at 250 oC. This step was repeated for three times, to keep cations in nitric form.
7. Dehydrated sample was dissolved with 4-5 ml 7 N HNO3.
Uranium and thorium extraction by anion-exchange resin column
1. A 6-7 cm long column was packed with AG 1-X8 anion-exchange resin was prepared.
2. One column volume of pure H2O with one drop of 14 N HNO3 was added to remove
metal ion.
3. Repeatedly adding one drop of 7 N HNO3 was added three times. One column
volume of 7 N HNO3 was then added to condition the column.
4. Sample solution was loaded into the column.
5. Repeatedly adding one drop of 7 N HNO3 three times. One column volume of 7 N
HNO3 was then added to fully elute Fe3+.
6. The Teflon beaker was refluxed with 15ml 1% aqua regia and 0.005 N HF for 10
minutes and then rinsed with pure H2O.
7. Repeatedly adding one drop of 6 N HCl three times. One column volume of 6 N HCl
was then added. The thorium fraction was collected with the original cleaned 30-ml
Teflon beaker. Two drops of HClO4 were added to the beaker to decompose organic
11
matters.
8. Repeatedly adding one drop of pure H2O three times. One column volume of pure
H2O was then added. The uranium fraction was collected with a clean 10-ml Teflon
beaker. Two drops of HClO4 were added to the beaker to decompose organic
matters.
9. The two separated thorium and uranium fraction were dried at 250 oC for fully
dehydration.
10. One drop of HClO4 was added to dissolve dehydrated sample. The sample was then
dried at 250 oC to remove organic matter. This step was repeated for three times.
11. One drop of 14 N HNO3was added to dissolve dehydrated sample. The sample was
dried at 250 oC. This step was repeated for three times.
12. The dried thorium and uranium fractions were dissolved with 1% aqua regia and
0.005 N HF and then transferred into PE vial from Teflon beaker for instrumental
analysis.
2.2.4 U-Th dating Instrumentation
A Thermo-Fisher multiple-collector inductively coupled plasma mass spectrometer
(MC-ICP-MS), equipped with a secondary electron multiplier (SEM) and combined
with an Aridus sample introduction system was used for U-Th isotopic measurement.
U-Th dating techniques were developed in the High-Precision Mass Spectrometry and
12
Environment Change Laboratory (HISPEC), Department of Geosciences, National
Taiwan University (Shen et al., 2012). The instrument sensitivity is 2-4%, with
precision of ±1-2Å ı for abundance determinations of 50-200 fg 234U (1 – 4 ng 238U)
or 230Th. One international standard, New Brunswick Laboratories Certified Reference
Material 112A (NBL-112A) was used for calibrating mass fraction and intensity bias
(Shen et al., 2012).
2.2.5 Stable oxygen isotope analysis Instrumentation
Stable oxygen and carbon isotope analysis was performed in three laboratories.
The first is the Kano Laboratory of the Department of Environmental Changes, Kyushu
University. Isotopic measurement was conducted by Thermo-Finnigan Delta Plus mass
spectrometer with GASBENCH II gas separation system. The second is Stable Isotope
Laboratory of the Department of Earth Sciences, National Taiwan Normal University.
Isotopic measurement was conducted by Micromass IsoPrime isotope ratio mass
spectrometer (IRMS). The third is Stable Isotope Laboratory of the Department of
Geoscience, National Taiwan University. Isotopic measurement was conducted by
Thermo-Finnigan MAT 253 IRMS, with Kiel-II carbonate system. Oxygen and carbon
stable isotopic data were reported as į18O and į13C respectively, relative to the Vienna
Pee Dee Belemnite (VPDB) reference standard. The NBS-19 was used as calibration
standard (į18O = -2.20‰, į13C = +1.95‰). One-sigma external precision of the
13
measurements was better than 0.08‰ for į18O and 0.04‰ for į13C.
2.2.6 Hendy Test and Replication Test
‘Hendy Test’, a widely adapted testing method, was proposed by Hendy et. al.
(1971), In order to preclude kinetic effect dominant stalagmites from providing biased
paleoclimate proxy records. The Hendy test is based on two indicators that are
established by the į18O and į13C value of subsamples in coeval stalagmite growth layers.
First is the standard deviation of į18O value, and second is the covariation between į18O
and į13C values. If į18O within the same layers have small ı value (<0.5‰) and/or į18O
and į13C values are poorly correlated (R2<0.4), the stalagmite is considered to
precipitate without dominant kinetic fractionation, thus its į18O profile would be
considered a valid paleoclimate proxy. Another relatively new and widely accepted
testing method is ‘Replication Test’ (Dorale and Liu, 2004), by estimating the
consistency between contemporaneous į18O profiles of stalagmites in different position
of the cave. Dorale and Liu (2004) proposed that only when kinetic effect is absent or
affect the stalagmites in the exact same way, will the different stalagmite į18O profiles
be well duplicated, in which the latter case is fairly unlikely. This study adapted both
approaches, to provide a robust examination.
14
Chapter 3 Results
3.1 U-Th dating results and age model
After screening, three stalagmites, MC110803-1, MC110803-2, and 090721-2MC
were selected (Figs. 3-2 to 3-4). Age models of three selected stalagmite were
constructed by 28 dating points, Age (before 1950 AD) intervals are 16.58-0.39 ka for
MC110803-1, 14.62-0.87 ka for MC110803-2, and 13.16-11.62 ka for 090721-2MC
(Fig. 3-1, Table 3-1 to 3-3). However, there is still a time gap at 11.6-5.7 ka in the
spliced interval of 11.62 ka-5.64 ka. The average 238U concentration of these three
stalagmites is 285 ppb for MC110803-1, 266 ppb for MC110803-2, and 554 ppb for
090721-2MC, respectively. The detailed U-Th isotopic and concentration data and
dating results are available in Appendix (I). The average growth rates of three
stalagmites are 0.21, 0.092, and 0.056 mm/yr for MC110803-1, MC110803-2), and
090721-2MC, respectively.
15
Figure 3-1. Plot of sample depth versus age for three selected Lekiraka cave stalagmites. Green, blue, and orange lines denote stalagmite MC110803-1, MC110803-2, and 090721-2MC, respectively. Most of 2ı error bars are smaller than the symbols.
16
Table 3-1. 230Th age data of stalagmite MC110803-1.
Sample ID Depth(mm) Age (ka) (UURU ı
110803-1A-29 29 0.386 0.021
110803-1A-80 80 0.442 0.012
110803-1A-117 118 0.605 0.011
110803-1A-122 122 0.677 0.010
110803-1A-151 151 0.792 0.013
110803-1A-167 167 1.374 0.018
110803-1A-326.5 326.5 2.115 0.016 110803-1A-370.5 370.5 2.803 0.019 110803-1A-447.5 447.5 3.274 0.029 110803-1A-635.5 635.5 5.394 0.034 110803-1A-637.5 637.5 14.51 0.13
110803-1A-789 789 16.576 0.087
17
Figure 3-2. Stalagmite MC110803-1
Table 3-2. 230Th age data of stalagmite MC110803-2.
Sample ID Depth(mm) Age (ka) (UURU ı
110803-2A-10 10 0.874 0.014
110803-2A-142 142 1.769 0.033
110803-2A-180 180 2.062 0.025
110803-2A-240 240 2.687 0.018
110803-2A-365 365 3.425 0.023
110803-2A-415 415 3.918 0.034
110803-2A-441 441 4.383 0.022
110803-2A-524 524 5.181 0.032
110803-2A-544 544 5.64 0.032
110803-2A-550 550 11.887 0.054
110803-2A-583.5 583.5 13.004 0.061
110803-2A-650 650 14.621 0.074
19
110803-2-2.5
Table 3-3. 230Th age data of stalagmite 090721-2MC.
Figure 3-4. Stalagmite 090721-2MC
Sample ID Depth(mm) Age (ka) (UURU ı
110803-2A-3 3 11.618 0.064
3.2 Oxygen and carbon stable isotope records
3.2.1 Hendy Test and Replication Test
‘Hendy test’ (Hendy et al., 1971) was conducted along coeval growth layers, and all layers have less than 0.2‰ of į18O variation (1-sigma) (Fig. 3-5a), showing that
stalagmites deposited at an oxygen isotopic equilibrium condition. The R2for į18O and į13C value in coeval layers of three selected stalagmites MC110803-1, MC110803-2,
and 090721-2MC were depicted respectively by Figs. 3-5b to 3-5d. Several layersį18O
and į13C values show high covariance. However, various slopes of regression lines have little implications of dominant kinetic effect in the Lekiraka cave. Two
contemporaneous į18O records for MC110803-1 and MC110803-2 are well replicated
from 0.8 to 5.7ka (Fig. 3-6), combined with robust Hendy Test, this high replication
between records suggest that little kinetic fractionation and insignificant influence form
water-rock interaction.
22
Figure 3-5. Results of ‘Hendy Test’ of Lekiraka stalagmites. (a) Green, blue, and orange symbols denote the results of ‘Hendy Test’ of various stalagmite sections.
1-sigma value of all tested layers are less than 0.2‰, indicate a stable precipitation environment. (b)(c)(b) The linear regressions of subsamples from different stalagmites are subsequently color-coded. Various slopes of regression lines suggest remote possibility of dominant kinetic effect.
a.
b. c. d.
23
3.2.2 Oxygen stable isotope time series
Oxygen stable isotope time series of three selected stalagmites are given in Figure
3-7. į18O values of Lekiraka stalagmite records vary between -4.89‰ and -2.31‰
during 16.5-11.5 ka, -2.94‰ and -6.85‰ during 5.7-3.5 ka. A rapid drop of 1‰
occurred during 5.5 ka – 5.7 ka. This compiled record shows an 18O depletion tend from
late Pleistocene to Holocene (Fig 3-7).
Previous studies indicate that deglaciation process would essentially decrease the seawater į18O values for ~1‰ of decreasing (Dykoski et al., 2005; Schrag et al., 1996).
We assume that the shift of sea level is proportional to į18O change during deglaciation,
the reconstructed deglacial sea-level curve (Bard et al., 1996; Fairbanks et al., 1989;
Grant et al., 2012; Siddall et al., 2003) can be used to calculate decrease in į18O value
of ~0.5‰ during 16 to 11.5 ka, and additional ~0.5‰ decrease during the first half of Holocene (Griffiths et al., 2009). However, the fluctuation of moisture į18O values
during transformation of last deglaciation is difficult to be well estimated. The exposure
and flooding of Sunda Shelf could bring an unexpected effect on regional atmospheric
circulation and alternation of moisture amount and source to our study sites.
24
Figure 3-6. į18O records of Lekiraka stalagmites. Green, blue and orange line respectively denote į18O records of stalagmite MC110803-1, MC110803-2, and 090721-2MC. 230Th dates with 2ıġ error are color-coded with stalagmites. Vertical colorful bars indicate Heinrich event 1 (H1) stadial, Bølling-Allerød (B-A) interstadial, and Younger Dryas (YD) stadial defined in previous studies (Dykoski et al., 2005; Petit et al., 1999; Wang et al., 2001).
25
Chapter 4 Discussion
4.1 Fluctuation of hydroclimate in Australian-Indonesian monsoon territory since
the last 16,500 years
Modern (1952-1974 AD) climate data (Griffiths et al., 2009; National Directorate
of Meteorology and Geophysics, Timor-Leste), long-term (1981-2011) regional
climatology data (http://www.esrl.noaa.gov, Physical Science Division, Earth System
Research Laboratory, NOAA), and moisture-source trajectory modeling (Draxler and
Rolph, 2013) (Figs. 4-1 to 4-3) show the same hydrological condition in our Lekiraka
cave and published Liang Luar cave in Western Flores (Griffiths et al., 2009) (Figs. 4-1
to 4-2). The agreements indicate the rainfall change in Lekiraka and Liang Luar caves is
dominantly governed by same climatic condition under the Australian-Indonesian
summer monsoon (AISM). The minor offset between Lekiraka and Liang Luar records
could be attributed to different travel distance of moisture source and different
summer/winter moisture ratios (Figs. 4-3 and 4-4). Therefore, /HNLUDND į18O record is
highly consistent with that of Liang Luar (Griffiths et al., 2009), with similar peak at
Younger Dryas (YD) stadial, and relatively low į18O during the Holocene. However,
there is a prominent shift of 1‰ of į18O value during 5.5-5.7 ka, which could be
attributed to recrystallization process of the stalagmites.
26
Figure 4-1. Map showing site locations and moisture trajectories. Stars mark the cave locations of Lekiraka cave (yellow, this study), Liang Luar cave in Western Flores (orange) (Griffiths et al., 2009), Bukit Assam and Snail Shell caves in Gunung Buda National Park in Borneo (white) (Partin et al., 2007), and also a marine sediment core GeoB10053-7 in the eastern Indian Ocean (red) (Mohtadi et al., 2011). Solid lines represent five-year average (2007-2011 AD) moisture trajectories of Lekiraka cave that were constructed by the HYSLPIT moisture trajectory model (Draxler and Rolph, 2013).
Blue lines represent the moisture trajectories for rainy season (December – March), and red lines represent moisture trajectories for the other months from April to November.
27
Figure 4-2. Moisture trajectories for Laing Luar cave (orange star, Griffiths et al.,
2009). Stars mark the cave locations of Lekiraka cave (yellow, this study), Liang Luar
cave in Western Flores (orange) (Griffiths et al., 2009). Solid lines represent moisture
trajectories (September 2006 – April 2007) of Liang Luar cave that were constructed by
the HYSLPIT moisture trajectory model (Draxler and Rolph, 2013). Blue lines
represent the moisture trajectories for rainy season (December - March), and red lines
represent moisture trajectories for dry season (June - August) (Modified from Griffiths
et al., 2009).
28
Figure 4-3. Map showing site locations and seasonal climate information.
Remote-sensing SST data are averaged from January 2002 to February 2008 (Aqua MODIS, http://oceancolor.gsfc.nasa.gov). Stars mark the cave locations of Lekiraka cave (yellow, this study), Liang Luar cave in Western Flores (orange) (Griffiths et al., 2009), Bukit Assam and Snail Shell caves in Gunung Buda National Park in Borneo (white) (Partin et al., 2007), and also a marine sediment core GeoB10053-7 in the eastern Indian Ocean (red) (Mohtadi et al., 2011). Superimposed black arrows indicate the intensity and direction of scatterometer wind in (a) August and (b) February (Modified from Mohtadi et al., 2011).
a. August b. February
29
Figure 4-4. Present-day IPWP hydroclimate. Observed mean annual sea surface salinity (SSS) (Antonov et al., 2009). Solid black contour indicates the IPWP boundaries as defined by the annual mean sea surface temperature > 28ɗ (Reynolds, 2002). Dots denote the locations of marine proxy records used in DiNezio et al. (2013).
Bukit Assam & Snail Shell
GeoB 10053-7
Liang Laur Lekiraka
30
A prominent į18O shift of 1.8‰ between 16.5-11.5 ka and 5.7-0.4 ka in the
spliced Lekiraka record indicates a relatively dry glacial period and wet Holocene
in East Timor. This results generally correspond to the up-rising of eustatic sea
level since the last glacial maximum (Grant et al., 2012; Hanebuth, Stattegger and
Grootes 2000; Hanebuth et al., 2011; Sathiamurthy, 2006; Sidall et al., 2003).
However, in sediment records from core GeoB10053-7 (Mohtadi et al., 2011)
(Fig. 4-5), millennial-scale events left relatively clear signatures in its AISM
records, and the AISM strengthen rapidly in late Holocene. Lekiraka and Liang
Luar records have a constant of about 1‰ offset, which could be attributed to
different ratios of moisture (Fig. 4-1 to Fig. 4-4). The inconsistency between
ocean sediment records and stalagmite records could be attributed to complex
rainfall mechanism in this area (DiNezio et al., 2013; Griffiths et al., 2010;
Mohtadi et al., 2011). Compare Lekiraka and Liang Luar records to planktonic
foraminiferal Ʃ18O-inferred Australian-Indonesian winter monsoon (AIWM)
records of GeoB10053-7, highly consistent pattern could be observed (Fig. 4-5).
The anti-phasing relationship between GeoB10053-7 AIWM records and Liang Luar / Lekiraka į18O records could be attributed to the “pull-push” mechanism
between the AIM and the EAM (An et al., 2000; Rohling et al., 2009).
31
Figure 4-5. Paleoclimate proxy records over the past 18,000 years. D į18O record of NGRIP ice core, Greenland (North Greenland Ice Core Project members, 2004). (b) į18O record of Dongge cave, China (Dykoski et al., 2005). (c į18O record of Liang Luar cave, Flores, Indonesia (Griffiths et al., 2009). (d į18O record of Lekiraka cave, East Timor (colored line). (e) (f) ǻ18O and Lithogenic/CaCO3 record of sediment core GeoB 10053-7, offshore Java (Mohtadi et al., 2011). (g) Sea-level reconstruction (Sidall et al., 2003). (h) Insolation intensity of 30oS (January).
32
4.2 Comparison with global and regional paleoclimate proxy records
4.2.1 Late Pleistocene --- From H1 to YD
A dry/cold YD condition shown by significant increase of į18O value in the Asian
monsoon (AM) territory stalagmite records (Dykoski et al., 2005; Hu et al., 2008; Jiang
et al., 2012; Wang et al., 2001) is absent in Lekiraka and Laing Luar records. Instead, a
slightly wet/warm YD condition could be observed in Lekiraka and Laing Luar records
(Griffiths et al., 2009), corresponded to intensification of the East Asian Winter
monsoon (EAWM) (Yancheva et al., 2007). Assuming monsoon moisture trajectories
were broadly similar to present time, a large proportion of summer monsoon trajectory
would have been occupied by land during late Pleistocene (Grant et al., 2012; Hanebuth
et al., 2011; Sathiamurthy, 2006; Sidall et al., 2003) in Maritime Continent region.
Limited moisture availability of source moisture greatly confined the development of
AISM, which suggests that southern displacement of the mean ITCZ position should be
responsible for increasing rainfall over research sites during YD, agreed with modeling
results (Dykoski et al., 2005; Griffiths et al., 2009; Yancheva et al., 2007; Zhang and
Delworth, 2005).
33
Figure 4-6. Simulation result of the fully coupled ocean-atmosphere global general circulation model (CM2.0) with the slab ocean model. Simulated annual mean precipitation anomaly (m/yr) base on thermohaline reduction scenario, a slightly positive precipitation anomaly can be observed over the Southern Indian Ocean and the Southern Pacific. Yellow star denote the location of Lekiraka cave (this study). The blue contour is the annual mean sea level pressure (SLP) anomaly with an interval of 0.4 hPa (Modified from Zhang and Delworth, 2005).
34
Millennium scale climate events such as YD, Heinrich stadial 1 (HS1) , and
Millennium scale climate events such as YD, Heinrich stadial 1 (HS1) , and