The Asian monsoon - 50-7 ka BP

(1)

The Asian monsoon – 50 -7 ka BP

Akkaneewut Chabangborn

(2)

Abstract

The Asian monsoon is one of the largest climatic systems on Earth. It covers an area from the Arabian Sea to the South China Sea and from northern Australia to northern China with the world’s highest population density. Moreover, the Asian monsoon transports heat energy and humidity to higher latitudes. In order to better understand the behaviour of the Asian monsoon and its environmental impact, its variability between 50 and 7 ka BP is analysed using paleo-data compilation, data-model comparisons, and lake sediment analysis.

The main results presented here are from the compilation of the Asian monsoon variability during the last glacial maximum (LGM) (23 - 19 ka BP) which is presumed to be under persistence cool and dry climatic conditions. The pattern of reconstructed and simulated precipitation agrees well in most of the region. However, the data-model discrepancies show in some areas, which may come from low resolution of the model or the local topographic effect. The reconstructed SSTs are well correlation with simulated SSTs, except in the Arabian Sea. The LGM Asian monsoon changes around 20 – 19 ka BP. The simulated ITCZ varies between 5°N and 15°N in the west and the east of the Asian monsoon region. However, the reconstructed ITCZ is ~5°N in the Arabian Sea, shifts northward in the Bay of Bengal, reaches ~30°N over central of China and migrates southward in the South China Sea. The ITCZ is likely shift northward after 20 ka BP. The climatic change might have been triggered by several factors, e.g., an increased land-sea thermal contrast and a variation of Pacific water inflow.

(3)

Contents

1. Introduction 1

2 Objectives 1

3. Asian monsoon in the past and outstanding questions 3

4. Method of study 4

4.1 Data-model comparison 4

4.1.1 Paleo dataset 4

4.1.1.1. Spatial analysis 5

4.1.1.2. Temporal analysis 5

4.1.2. Climate model 6

4.1.3. The northward shift of the ITCZ 6

4.2 Lake sediment analysis 6

4.2.1 Thailand and its relationship to the Asian monsoon 6 4.2.2 Lake sediments from northern and southern Thailand 8

5. Paper I: Results and discussions 8

5.1 CCSM3 - paleodata comparison 13

5.2 CCSM3 and MARGO dataset comparison 17

5.3 Temporal variability of Asian monsoon during the LGM 19

6. Future perspectives 22

6.1 Analyze lake sediment from Nong Pha Kho 22

6.2 Compilation of the early Holocene 23

7. Acknowledgement 25

8. References 25

9. Paper I 38

(4)

1. Introduction

The monsoon is generally explained as a seasonal reversing wind associated with a corresponding change in precipitation (Ramage, 1971; Holton, 2004). The monsoon mechanism can be explained by the land-sea thermal contrast, caused by annual solar variation (Figure 1). The Asian monsoon system consists of two different air masses: the warm and humid southwest and southeast monsoon (summer monsoon), and the cool and dry northeast monsoon (winter monsoon). In summer, the southeast and southwest monsoon transport humidity from the Indian and Pacific Oceans to precipitate over the Asian continent (Figure 1). Moreover, the summer season is characterized by the northward shift of the Intertropical Convergence Zone (ITCZ) from the equator to the continent (Chao and Chen, 2001; Fleitmann et al., 2007; Clift and Plumb, 2008). During winter, cool and dry air masses migrate from the Asian mainland together with the northeast monsoon and the ITCZ moves southward.

The Asian monsoon, one of the largest climate systems on Earth, covers a large region from the Arabian Sea to the South China Sea and from northern Australia to northern China (Fig. 1). The Asian monsoon region can be divided into two sub-systems according to the differences in summer wind circulation pattern: the Indian monsoon and the East Asian Monsoon. These are divided approximately along longitude 105°E (Wang et al., 2003; Wang et al., 2005) (Fig. 1). The Indian monsoon domain is characterized by the formation of a distinct gyre over the Indian Ocean caused by wind circulation across the equator. The East Asian monsoon sub-system is a convergence of the southwest monsoon, the Pacific Trade winds and the subtropical front over China. Precipitation associated with the Asian monsoon is of great societal and economic importance for a region where more than half of the world’s population live. In addition, the Asian monsoon affects Earth’s other climatic systems through transport of heat energy and humidity to higher latitudes (Zahn, 2003; Clift and Plumb, 2008; Maher, 2008; Caley et al., 2011).

2. Objectives

A better understanding of the behaviour of the Asian monsoon during different climate states will further our knowledge on climate dynamics and will increase climate

(5)

model accuracy. Paleo records, which provide insight into climate variations in the past, are scarce for the Asian monsoon region. Most of the published terrestrial paleo records for the Asian monsoon region have low dating control and focus almost only on pollen analysis and vegetation reconstructions. This makes comparisons to other paleoproxy studies (marine records, speleothems) difficult.

Fig. 1: Monsoon precipitation (red-yellow) associated with wind circulation patterns (arrows) in summer (upper figure) and winter (lower figure) (from Wang et al., 2003, Fig 1 and 4). The Indian and East Asian monsoon sub-regions are separated along the thick

(6)

The objective of this study is to better understand Asian monsoon behaviour, both in space and time, and its environmental impact between 50 and 7 ka BP. For this we combine proxy data-model comparison, and lake sediment studies.

3. Asian monsoon in the past and outstanding questions

The main overall driver of Asian monsoon climate are changes in insolation (Fig.

2). Multi-proxy analyses of lake sediment sequences from Lake Huguang in south China and Lake Biwa in Japan, and the speleothem δ¹⁸Ο records from Hulu cave in China, for example, indicate prevailing warm and wet climate conditions from 50 to 40 ka BP and cool and dry climates from 40 to 15 ka BP (Xiao et al., 1999; Migram et al., 2004; Wang et al., 2001). However, paleorecords also show that the peak of summer monsoon intensity lagged maximum insolation forcing by around 3,000 years (Clemens et al., 1991; Overpeck et al., 1996; Beaufort et al., 2001).

It is traditionally assumed that the Asian monsoon region experienced cool and dry climatic conditions, caused by a strengthened northeast monsoon, during the Last Glacial Maximum (LGM: 19-23 ka BP) (Huang et al., 1997; Cosford et al., 2010;

Fleitmann et al., 2011). On the other hand, other paleorecords suggest a periodically strengthened summer monsoon during the LGM (Sun et al., 2000; Rashid et al., 2007;

Saher et al., 2007; Govil and Naidu, 2011; Mahesh et al., 2011). Also, the spatial variability of the Asian monsoon during different time intervals in the past is still an open question. It is for example still under discussion, why precipitation was high over Sumatra (van der Kaars et al., 2010), in the South China Sea (Sun et al., 2000; Colin et al., 2010) and in western China (Yu et al., 2000a; 2000b and 2003) during the LGM, despite overall dry conditions. The asynchronicity of the early Holocene high precipitation interval between the East Asian and Indian monsoon sub-regions (Herzschuh, 2006) is also still not resolved. These discrepancies likely indicate that not only insolation patterns are important, but that other factors, e.g. the El Niño-Southern Oscillation (ENSO), atmosphere-ocean interactions (Wang et al., 2005), the southward shift of the ITCZ (Braconnot et al., 2007; Broccoli et al, 2006; Zhang and Delworth, 2005), the Pacific Warm Pool (De Deckker et al., 2002; Partin et al., 2007), and North Atlantic circulation (Tiwari et al., 2009; Pausata et al., 2011; Stager et al., 2011;

(7)

Overpeck et al., 1996) may have played an important role in modulating Asian monsoon variability through time.

Fig. 2: Estimated amount of insolation (W/m²) at latitude 15°N (orange) compared with the δ¹⁸O records from Hulu (pink) (Wang et al., 2001) and Dongge caves (purple) (Yuan et al., 2004) in China (data set retrieved from the NOAA data center). The blue coloured intervals represent cool and dry climatic conditions.

4. Method of study

4.1 Data-model comparison 4.1.1 Paleo dataset

The paleo dataset was based on published paleo records from the Asian monsoon region and was constrained to between 20ºS and 40ºN, and to between 40 and 160º E (Wang et al., 2003; Wang et al., 2005) (Fig. 1). The paleorecords were selected by using the following criteria:

(8)

i) Each record should contain at least one ¹⁴C or U/Th dating result during the interested interval

ii) Dating results with estimated errors of > 1000 years and ¹⁴C dates obtained on carbonate bulk sediments were excluded

iii) Paleo records with more than one age estimate were used to assess the temporal variability of the Asian monsoon, while records containing only one age estimate were used as supporting information.

All published ¹⁴C ages were calibrated using the online program Calib 6.0 (http://calib.qub.ac.uk/calib/calib.html) (Reimer et al., 2009) and age-depth curves were constructed for paleorecords containing several ¹⁴C dates. The paleorecords were interpreted in terms of spatial and temporal variability of past precipitation reflecting a strengthened summer monsoon.

4.1.1.1. Spatial analysis

The paleo records were compared for each type of proxy. Pollen (>5%) was assigned to biomes by using Plant Functional Types (PFTs) that had been established for China (Yu et al., 2000), Japan (Takahara et al., 2000), South East Asia (Pickett et al., 2004), and by the Global Paleovegetation Mapping (BIOME6000) project (Prentice and Webb 1998). For the Himalayan Mountains, we used the PFTs established by Kramer et al. (2010) because the BIOME6000 project for the Indian continent is still in progress.

The speleothem and foraminifera δ¹⁸O records were averaged over the interested time interval and are assumed to represent precipitation.

4.1.1.2. Temporal analysis

Paleo proxies from individual sites with more than one dating estimate were considered for a temporal analysis. The variation of cold-dry tolerant pollen taxa, e.g.

Artemisia and Chenopodiaceae, or of wet indicators, such as Pteridophyta spores, was utilized to identify climatic changes. The series of averaged δ¹⁸O records measured on speleothem and foraminifera were established using a running mean. This was then compared with the δ¹⁸O mean value of the interested interval from the spatial analysis of the individual site.

(9)

4.1.2. Climate model

The climate model output data was simulated by the Community Climate System Model version 3 (CCSM3), which is a global, coupled ocean-atmosphere-sea ice-land surface climate model (Collins et al., 2006). The simulation analyzed here is described in Otto-Bliesner et al. (2006) and Brandefelt and Otto-Bliesner (2009). CCSM3 mean values of annual air temperature, precipitation and SSTs were calculated and used for comparisons with the qualitative precipitation estimates derived from the compiled paleo datasets and the quantitative sea surface temperatures (SST).

4.1.3. The northward shift of the ITCZ

In order to provide insight into the shift of the southwest monsoon, the northern boundary of the ITCZ was assessed using the reconstructed precipitation and calculated from climate model output according to the method outlined by Braconnot et al. (2007) as:

∑

=

= = _N

pr lat y

N

pr lat y

y pr

y lat y pr lon

ITCZ

loc _o

o

30

max) ( 30

max) (

) (

) ( ) ( )

( _

Where loc_ITCZ(lon) = location of the ITCZ in each longitude max)

( pr

lat = latitude of maximum precipitation in each longitude )

( y

pr = precipitation over latitude y

4.2 Lake sediment analysis

4.2.1 Thailand and its relationship to the Asian monsoon

Thailand occupies a strategic position between the two monsoon sub-regions (Fig.

3). Lake sediments may therefore record a climatic signal of past changes in monsoon intensity that can be compared to other studies in the Asian monsoon region, and to

(10)

Fig. 3: The position of the ITCZ (shaded area) and the possible occurrence of tropical cyclones (dash orange line) over Thailand for each month of the year. The red and blue arrows mark the southwest and northeast monsoon, respectively. The Indian and East Asian monsoon sub-regions are divided by the dashed black line (longitude 105°E). The yellow and blue circles represent Nong Tale Pron and Pa Kho (Modified from http://www.tmd.go.th/info/info.php?FileID=23).

The climate of Thailand is controlled by the southwest and northeast monsoon during the rainy and dry seasons, respectively. The rainy season starts in May and ends in October. It can be divided into two intervals; from May to June and from September to October precipitation is mainly caused by the Indian and East Asian monsoon sub-

(11)

systems. The southwest monsoon and the ITCZ reach Thailand in May and cause heavy precipitation over the area. The ITCZ shifts northward and passes over Thailand in June.

The ITCZ is situated over the South China Sea during July and causes the formation of tropical cyclones, which pass over Northeast Thailand. The ITCZ shifts southward and passes Thailand again in August leading to higher precipitation than between May and June. The heavy rainfall continues until October (Fig. 3). After that, climatic conditions become cool and dry due to the northeast monsoon, which prevails from November to January.

4.2.2 Lake sediments from northern and southern Thailand

Lake sediments were collected from sites in northern (Nong Pa Kho) and southern (Nong Tale Pron) Thailand during fieldwork in January 2009-2011 (Fig. 3). The geochemical analysis of these sediment sequences will be performed during the next two years.

Laboratory work will include detailed lithostratigraphic descriptions and correlations between overlapping core segments in order to establish a continuous sediment succession. Consecutive 1- cm samples will be analyzed for loss-on-ignition (LOI; 550°C and 950°C for organic matter and carbonate content). The total organic carbon (TOC), total nitrogen (TN), total sulphur (TS) and bulk δ¹³Corg will be measured based on the pattern of the LOI data series by a MAT Delta+ mass spectrometer.

For 14C dating analysis, sediment samples will be sieved under running water.

The sieve remains will then be examined under a binocular microscope and identifiable plant remains (seeds, charcoal, leaves, twigs and wood) will be picked out and sent for dating.

5. Paper I: Results and discussions

Paper I focuses on a data-model comparison of the last glacial maximum (LGM), which is the time interval between 23-19 ka BP. It is the most recent period in Earth’s

(12)

of 130 m (Clack et al., 2009). The LGM is presumed to have well known primary boundary parameters and to be under quasi-equilibrium conditions (Mix et al., 2001;

Schmidt, 2010). Therefore it has been used as a test period for climate models and to assess paleoproxy uncertainties. During the LGM, the Asian monsoon region is traditionally assumed to have experienced cool and dry climatic conditions, caused by a strengthened northeast monsoon (Huang et al., 1997; Hodell et al., 1999; Hope, 2001;

White et al., 2004; Cosford et al., 2010; Fleitmann et al., 2011).

Marine and terrestrial pollen, foraminifera and speleothem δ¹⁸O, and the MARGO (2009) dataset were used to assess past precipitation and sea surface temperatures (Table 1) and compared with CCSM3 data output. The LGM paleorecords used here are constrained to between 25 and 17 ka BP in order to cover the LGM Chronozone 2 (Mix et al., 2001). Moreover, the Asian monsoon region was divided into 3 sub-regions: the Sundaland, the East Asian monsoon (EAM) and the South Asian monsoon (SAM) (Fig.

4). The Sundaland was treated separately because the LGM sea level low stand led to the exposure of a vast area that differs markedly from the present environment.

(13)

Table 1: Paleorecords and paleoproxies used for the compilation. The age assignments are based on M = MARGO project members (2009), ¹⁴C and U/Th dates. The number of dates between 25-17 ka BP for each sequence is given in brackets.

Site

No. Site Name Lat

(°)

Long (°)

Elevation

(m asl.) Archive Proxy Age assignment References

1 SHI9034 -9.10 111.01 -3330 Marine δ¹⁸O M Ding et al. (2002)

2 SHI9016 -8.46 128.24 -1805 Marine δ¹⁸O M Spooner et al. (2005)

3 Bandung basin -7.00 108.00 665 Terrestrial Pollen ¹⁴C (1) van der Kaars & Dam (1997)

4 BAR94-42 -6.75 102.42 -2542 Marine Pollen ¹⁴C (2) van der Kaars et al. (2010)

5 SHI9006 -4.33 117.60 -1999 Marine δ¹⁸O M Ding et al. (2002)

6 Danau Sentarum 0.73 112.10 35-50 Terrestrial Pollen ¹⁴C (1) Anshari et al. (2001)

7 Danau di Atas -1.07 100.77 1535 Terrestrial Pollen ¹⁴C (2) Newsome & Flenley (1988)

8 Pea Sim-sim swamp 2.29 98.89 1450 Terrestrial Pollen ¹⁴C (5) Maloney (1980)

9 Pee Bullok 2.28 98.98 1400 Terrestrial Pollen ¹⁴C (4) Maloney & McCormac (1996)

10 K-12 2.69 127.74 -3510 Marine δ¹⁸O M Barmawidjaja et al. (1993)

11 Tasek Bera Basin 3.06 102.64 20-30 Terrestrial Hardwood remain ¹⁴C (1) Wüst & Bustin (2004)

12 KH92-1-5cBX 3.53 141.87 -2282 Marine Alkenone M Ohkouchi et al. (1994)

13 Gunung Buda National Park

4.00 114.00 ~1000 Terrestrial δ¹⁸O U/Th (9) Partin et al. (2007)

14 SO18302 4.15 108.57 83 Marine Pollen ¹⁴C (1) Wang et al. (2009)

15 SO18300 4.35 108.65 91 Marine Pollen ¹⁴C (1) Wang et al. (2009)

16 GIK17964-2 6.16 112.21 -1556 Marine Alkenone M Pelejero et al. (1999)

18 MD97-2142 12.69 119.47 - 1557 Marine δ¹⁸O M Chen et al. (2003)

20 31-KL 18.75 115.87 -3360 Marine δ¹⁸O M Chen et al. (1998)

21 GIK17938-2 19.79 117.54 -2840 Marine δ¹⁸O M Chen et al. (1999)

22 MD97-2148 19.80 117.54 -2830 Marine δ¹⁸O M Chen et al. (2002)

24 Core 17940 20.12 117.38 -1727 Marine Pollen ¹⁴C (2) Sun et al. (2000)

25 Tianyang Basin 20.78 110.03 120 Terrestrial Pollen ¹⁴C (2) Zheng & Lei (1999)

26 Huguang Maar lake 21.15 110.28 23 Terrestrial Pollen ¹⁴C (2) Wang et al. (2010)

(14)

Site

No. Site Name Lat

(°)

Long (°)

Elevation

27 Toushe Basin 23.82 120.88 650 Terrestrial Pollen ¹⁴C (3) Liew et al. (2006)

28 DGKS9603 28.15 127.27 -1100 Marine δ¹⁸O M Li et al. (2001)

29 DGKS9603 28.15 127.27 -1100 Marine Pollen ¹⁴C (2) Xu et al. (2010)

30 Jintanwan Cave 29.48 109.53 460 Terrestrial δ¹⁸O U/Th (3) Cosford et al. (2010)

31 Hulu Cave 32.30 119.17 100 Terrestrial δ¹⁸O U/Th (3) Wang et al. (2001)

32 Songjia Cave 32.41 107.41 ~680 Terrestrial δ¹⁸O U/Th (2) Zhou et al. (2008)

33 Beizhuangcun, Shaanxi Province

34.33 109.48 600-1100 Terrestrial Pollen ¹⁴C (2) Wang & Sun (1994)

34 Biwa Lake 35.25 136.05 85 Terrestrial Pollen ¹⁴C (2) Hayashi et al. (2010)

35 Weinan section, Loess Plateau

34.40 109.50 600-1100 Terrestrial Pollen ¹⁴C (1) Sun et al. (1997) 36 Pyonggeodong

archaeological site,

35.17 128.06 100-300 Terrestrial Pollen ¹⁴C (3) Chung et al. (2006)

37 Iwaya 35.52 135.88 20 Terrestrial Pollen ¹⁴C (1) Takahara & Takeoka (1992)

38 Lake Mikata 35.56 135.89 0 Terrestrial Pollen ¹⁴C (2) Yasuda (1982)

39 CH84-04 36.46 142.14 -2630 Marine Alkenone M Bard et al. (unpublished)

40 KH-79-3_L3 37.06 134.72 -935 Marine Alkenone M Ishiwatari et al. (2001)

41 KT94-15_PC-9 39.57 139.41 -807 Marine Alkenone M Ishiwatari et al. (2001)

42 MD85-674 3.19 50.44 -4875 Marine Alkenone M Bard et al. (1997)

43 SK-157-14 5.18 75.91 -3306 Marine δ¹⁸O ¹⁴C (1) Ahmad et al. (2008)

44 Horton Plains 6.81 80.83 2100-2300 Terrestrial Pollen ¹⁴C (2) Premathilake (2006)

45 MD77-191 7.30 76.43 -1254 Marine Alkenone M Sonzogni et al. (1998)

48 TY93905/P 10.70 51.93 ~-1500 Marine Alkenone M Sonzogni et al. (1998)

49 Nilgiri hills 11.25 76.67 2200 Terrestrial δ¹³C ¹⁴C (1) Rajagopalan et al. (1997)

50 MD77-195 11.30 74.32 ~-1200 Marine Alkenone M Sonzogni et al. (1998)

51 RC12-344 12.46 96.04 -2140 Marine δ¹⁸O ¹⁴C (3) Rashid et al. (2007)

52 Moomi Cave, Socotra Island, Yemen

12.50 54.00 ~1000 Terrestrial δ¹⁸O U/Th (8) Shakun et al. (2007)

53 TY93929/P 13.70 53.25 -2490 Marine Alkenone M Sonzogni et al. (1998)

(15)

Site

No. Site Name Lat

(°)

Long (°)

Elevation

56 GeoB3005-1 14.97 54.37 -2316 Marine Alkenone M Budziak et al. (2000)

58 MD76-131(C) 15.53 72.57 -1230 Marine δ¹⁸O M Cayre et al. (1999)

59 GeoB3007-1 16.17 59.76 -1920 Marine Alkenone M Budziak et al. (2000)

61 117-723_Site 18.05 57.61 -816 Marine δ¹⁸O M MARGO project member (2009)

64 SO93-126KL 19.97 90.03 -1250 Marine Alkenone M Sonzogni et al. (1998)

66 SO90-137KA 23.12 66.48 -573 Marine δ¹⁸O ¹⁴C (4) van Rad et al. (1999)

67 SO90-93KL 23.59 64.22 -1802 Marine Alkenone M Schulz & Emeis (unpublished)

68 Bharatpur Bird Sanctuary, Rajasthan

27.12 77.52 174 Terrestrial Pollen ¹⁴C (1) Sharma & Chatterjee (2007) 69 Kathmandu Basin 27.67 85.22 1303 Terrestrial Pollen ¹⁴C (3) Fujii & Sakai (2002) 70 Lake Shudu, Yunnan,

China

27.90 99.95 3630 Terrestrial Pollen ¹⁴C (4) Cook et al. (2011)

71 Phulara palaeolake 29.33 80.13 1500-1700 Terrestrial Pollen ¹⁴C (2) Kotilia et al. (2010)

72 Ren Co 30.73 96.68 4450 Terrestrial Pollen ¹⁴C (3) Tang et al. (2000)

(16)

5.1 CCSM3 - paleodata comparison

The pollen records from the Sundaland are assigned to a warm-temperate rainforest and a tropical deciduous broadleaf forest biome by PFTs, suggesting relatively high precipitation compared with the other sub-regions (Fig. 4). However, the warm- temperate rainforest biome in the west (#4 and #7) and the tropical deciduous broadleaf forest biome in the east (#3, #8, #9, #6, #14 and #15) may indicate slightly higher precipitation in the west than in the east. Another explanation for this discrepancy is the rain shadow caused by the Sumatra-Java mountain range. In addition, the biome assignments indicate that the exposed Sunda shelf would be mainly covered by deciduous forest, which corresponds well with other undated pollen records (Li and Sun, 1999; Sun et al., 2002) and with an abundance of C4 plants, as assessed from biomarker δ¹³C (Hu et al., 2003). The wet climatic conditions compare well with the significantly low δ¹⁸O from foraminifera and speleothem records and the high precipitation (>2000 mm/year) simulated by CCSM3 data output (Fig. 4). Moreover, the high precipitation would indicate a strengthening of the southeast and southwest monsoon during the LGM.

The relatively high precipitation over Sundaland as shown by CCSM3 compares well to the reconstructed position of the ITCZ, which covered this sub-region and indicates a prevailing southwest monsoon (Fig. 5). However, the northern limit of the ITCZ in CCSM3 at around 15°N does not compare well with reconstructed high rainfall and run-off along the Mekong River (Colin et al., 2010) and in southern China. This may come from a significant decrease in precipitation at 15°N (~> 1500 - ~<1000 mm/year) in the CCSM3 simulation.

For the EAM sub-region, a mix of steppe and warm mixed forest biomes and broadleaf evergreen/warm mixed forest are assigned to the pollen assemblages in lake sediments collected in south China (#25 and #26) and Taiwan (#27). These biomes indicate relatively low precipitation as compared to the Sundaland sub-region (Fig. 4).

Further north, near the East China Sea (#25) and the China mainland (#33, #35) the former biomes are replaced by steppe biome, and by a mix of steppe and cool mixed forest or cool mixed forest in Korea (#36) and Japan (#34, #37 and #38). These environmental changes generally agree well with the northward decrease in precipitation as simulated by CCSM3 (Fig. 4). In addition, these reconstructions correspond well with

(17)

the synthesis of Chinese paleorecords by Zheng et al. (1998), who suggested colder and drier climatic conditions in the north as compared to the south of China.

Pollen assemblages from marine cores collected in the South (#24) and East China Sea (#29) are mainly composed of non-arboreal pollen taxa, indicating relatively dry climatic conditions compared to the adjacent terrestrial records. These differing results may come from the fact that both marine cores were collected near river mouths, i.e. the Pearl and Yangtze River. Consequently, the high percentage of non-arboreal pollen taxa may have been transported from the Chinese mainland.

Speleothems from Jintanwan (#30) and Songjia caves (#32) in central China, and Hulu cave (#31) near the northeastern coast of China provide records of past precipitation. The LGM mean values of δ¹⁸O suggest that precipitation near Hulu and Jintanwan caves is lower than over the Sundaland sub-region. In contrast, the significantly low LGM mean δ¹⁸O values of speleothem records from Songjia caves suggest higher precipitation than over Sundaland. The Songjia and Jintanwan cave discrepancies caused by the LGM mean δ¹⁸O values from Songjia cave were calculated only between 20 and 19 ka BP. Consequently, the significantly low δ¹⁸O mean values of speleothems collected from Songjia cave are likely to indicate a significant climatic change from dry to wet climatic conditions after 20 ka BP.

The wet and dry climatic conditions over Jintawan cave (#30) and Songjia cave (#32) during the LGM may indicate that the northern boundary of the ITCZ may have been situated around Jintawan cave (~30°N) (Fig. 5). This agrees well with PMIP2 wind simulations, which suggest a significantly strengthened northeast monsoon wind north of 30°N (Jiang and Lang, 2010). However, the dominance of steppe and cool mixed forest biomes north of Taiwan may imply a southward shift of the ITCZ towards the South China Sea.

(18)

Fig. 4: Paleodata - CCSM3 comparison for the LGM. (A) The paleorecords used here were separated into different proxies; (B) quantitative precipitation modeled by CCSM3 and compared to the quantitative information on humidity inferred from the paleorecords.

The Asian monsoon region was divided into 3 sub-regions along the white lines.

(19)

Fig. 5: The northward shift of the Intertropical Convergence Zone (ITCZ) during the LGM in the Asian monsoon region. The orange line shows the position of the ITCZ simulated by CCSM3; the blue and green lines show the position of the ITCZ at 23-20 ka BP and 19 ka BP, respectively according to paleorecords. The squares represent the paleorecords used to interpolate the northern boundary of the ITCZ.

The paleoproxies indicate that precipitation over the SAM sub-region was significantly lower than that over the other sub-regions (Fig. 4). In addition, the paleorecords indicate that the northern SAM sub-region (#69, #70, #71) is wetter than the south sub-region (#44, #49), which does not correspond with the CCSM3 simulation.

However, the assumption of arid conditions in the western and central part of India and relatively wet conditions in the east may find support in the distribution of Paleolithic hunter-gatherer sites (Misra, 2001). These were preferentially located in eastern India during the Upper Paleolithic (30 - 10 ka BP), where more food may have been available due to wetter climatic conditions. The relatively high precipitation over the north SAM sub-region is likely a local effect, which may be explained by ascending air masses along the lee side of the Himalayan Mountains (Boos and Kuang, 2010; Cane, 2010).

Moreover, the LGM mean δ¹⁸O values of foraminifera obtained from the Bay of Bengal (#51) and southern India (#43) indicate wet climatic conditions, which are comparable

(20)

with those over Sundaland. In contrast, the LGM mean δ¹⁸O values of speleothem (#52) and foraminifera (#66) show a significantly lower precipitation over the Arabian Sea.

In the CCSM3 simulation, the northern boundary of the ITCZ is located at ~5°N in the Arabian Sea, which compares to the reconstruction. However, further to the east, the simulated ITCZ is situated at ~15°N, i.e. over southern India, because of a sudden decrease in precipitation. However, the reconstructed precipitation here suggests that the ITCZ might not have moved north of 5°N in the Arabian Sea and may have been situated south of the Indian subcontinent during the LGM (Fig. 5). This may be explained by the weak land-sea thermal contrast during the LGM due to sea level lowering, which could have prevented the development of low pressures south of India during the summer. In contrast, the ITCZ shifts northward in the Bay of Bengal towards Southwest China. The northward migration of the reconstructed ITCZ in the Bay of Bengal may have been caused by many factors, e.g. the exposure of the Sunda shelf, surrounding high mountain ranges, as well as a decreased inflow from the Pacific Ocean. These would have allowed the development of a low-pressure cell in the Bay of Bengal. The southwest monsoon may thus have moved along the mountain ranges and supplied humidity from the Pacific trade wind to form the EAM sub-regions.

5.2 CCSM3 and MARGO dataset comparison

Reconstructed SSTs (MARGO, 2009) based on alkenone and foraminifera δ¹⁸O collected in the Sundaland, the EAM sub-region, the Bay of Bengal and the eastern Arabian Sea show a good correlation with CCSM3 SSTs. The correlation coefficient is 0.91, although the MARGO (2009) data set gives values of ~2°C higher than CCSM3 SSTs (Fig. 6).

In contrast, the SSTs reconstructed from foraminifera δ¹⁸O obtained near LGM active rivers e.g. the Pearl River (#18, #20, #21, #22) the Yangtze River (#28) and the Gulf of Cambay (#58) are relatively high compared with those based on alkenone from adjacent areas (Fig. 7). This likely indicates inflow of freshwater rather than SSTs. The effect of run-off is significant in the EAM sub-region, where reconstructed and simulated SSTs decrease northward. Alkenone-based SSTs show a stepwise decrease from the southern part of the South China Sea (~25°C) (#16, #17), to the northern part of the South

(21)

China Sea (~22°C) (#23) and to the ocean around Japan (~20°C) (#39, #40, #41), while foraminifera δ¹⁸O based SSTs (#20, #21, #22 and #28) suggest slight differences between the South and the East China Sea. The relatively high reconstructed SSTs in the Sundaland suggest that a connection (Indonesian Gateway) between the Pacific Ocean and the Indian Ocean, where warmer and lower salinity waters could pass through, even during low sea level. However, reconstructed SSTs near Lombok Strait (~26°C) (#1) are lower than those from other areas in the Indonesian Gateway (27°C - 28°C) (#2, #5, #10,

#12), which may reflect the lowering of the sea level, a narrowing of the strait and decreased through-flow during the LGM (Žuvela-Aloise, 2005). Reconstructed SSTs from the Bay of Bengal and along the southwest coast of India are similar to those from the Lombok Strait. The relatively high SSTs off the southwest coast of India have been explained by a flow of water from the Bay of Bengal to the Arabian Sea due to a strengthened Northeast Monsoon (Chodankar et al., 2005).

Fig. 6: Comparison between MARGO (2009) reconstructed SSTs and CCSM3 simulated

(22)

In the western Arabian Sea, the reconstructed SSTs are relatively low (~23°C) in the north, and relatively high in the south (~26°C), while CCSM3 SSTs are slightly different.

Fig. 7: Sea surface temperatures in the Asian monsoon region reconstructed by δ¹⁸O and alkenones (MARGO 2009).

5.3 Temporal variability of Asian monsoon during the LGM

The paleorecords with more than one dating result between 25 and 17 ka BP were used to analyze the temporal variability of the Asian monsoon. These indicate that climatic conditions changed from relatively wet to dry in Sumatra (#4 and #7), Borneo (#13), the south of China (#26), Taiwan (#27) and the west of China (#70) between 20 ND 19 ka BP (Fig. 8). In contrast, an opposite scenario is reconstructed for north of Taiwan in the EAM sub-region (#29, #30, #31, #34, #38) and in the SAM sub-region (#44, #52, #69 and #71) at almost the same period. The change in climatic conditions from dry to wet over the Sundaland sub-region appears only in sites located behind the Sumatra-Java mountain range, indicating the effect of a rain shadow associated with a

(23)

weaker southwest monsoon. The north-south discrepancies in the EAM sub-region may have been caused by the position of the northern boundary of the ITCZ extending from 30°N in central China and to the south near Taiwan. This implies that the southwest monsoon never reached north of 30°N until 20 ka BP, but became strengthened thereafter. The shift from dry to wetter climate conditions may be explained by the gradual rise in sea level at ~20 ka BP, which was followed by a rapid increase at ~19 ka BP (Yokoyama et al., 2000; Hanebuth et al., 2009; Hanebuth et al., 2011). The rise in sea level increased the water flow from the Pacific Ocean and pushed high humidity air masses that had been concentrated to the Bay of Bengal, to the west. The westward movement of high humidity air masses after ~20 ka BP increased the precipitation amount over India and the Arabian Sea, and led to a decrease in precipitation in the Sundaland, along the southern edge of the Himalayan Mountains and over the South China Sea. Moreover, it initiated a northward shift of the ITCZ between 20 and 19 ka BP (Fig. 5). These changes seem to have occurred slightly earlier than the changes described by Herzschuh (2006) for China, who noted a change towards wetter conditions between 18.5 and 17.5 ka BP.

(24)

Fig. 8: Quantitative humidity change inferred from pollen assemblages in the Asian monsoon region spanning the time interval between 25 and 17 ka BP. Thicker lines represent the calibrated age points and the thin lines are interpolations between these. The black and grey colours represent relatively wet and dry climatic condition, respectively during the time interval.

(25)

6. Future perspectives

6.1 Analysis of the lake sediments from Nong Pha Kho

Nong Pha Kho (17°06’N, 102°56’E) is situated in northeastern Thailand (Fig. 8).

As mentioned above, the area is influenced by both the Indian and East Asian monsoon.

Nong Pha Kho is around 175 m above sea level and 20 km southwest of Nong Han Kumphawapi. According to Penny et al. (2001), the pollen records obtained from this lake can trace climatic conditions back to ~40 ka BP.

Figure 9: Nong Pa Kho in northeastern Thailand.

(26)

The sediment cores from Nong Pha Kho were collected in January 2009 by using a modified Russian corer (7.5 cm diameter, 1 m length). The sediment cores were taken with an overlap of 50 cm in order to achieve a continuous sequence. Detailed lithostratigraphic descriptions and loss-on-ignition (LOI; 550°C and 950°C for organic matter and carbonate content) have already done. ¹⁴C dating shows that the bottom sediments are older than 50 ka BP. The comparison of the %LOI curve from Pa Kho with the δ¹⁸O records from Hulu Cave (Wang et al. 2001), the North Greenland ice core record (North Greenland Ice Core Project members, 2004) and the EPCIA Dronning Maud Land (EDML) ice core from Antarctica (EPICA Community members, 2006) shows an interesting relationship (Fig. 10). Lower values of LOI match with higher δ¹⁸O values in EDML suggesting that warmer periods over Antarctica compare to higher precipitation over Pa Kho. However the opposite is the case for Greenland and Hulu Cave. Warmer climate conditions over Greenland, and higher precipitation over Hulu Cave compare to lower values of LOI.

The future work will be to obtain a detailed ¹⁴C chronology for the lower 7 m of the Pa Kho sequence, to analyze total organic carbon (TOC), total nitrogen (TN), total sulfide (TS) and bulk δ¹³C_org.

(27)

Fig. 10: Comparison of the LOI (%) curve for Pa Kho and the δ¹⁸O records of Hulu Cave, the North GRIP ice core and the Antarctic EDML ice core.

6.2 Compilation of proxy data for the early Holocene and data-model comparisons The early Holocene was characterized by a rapid sea level rise as well as by abruptly warm and wet climatic conditions. Speleothem, loess, and lacustrine records in various parts of the Asian monsoon region indicate that the strengthened summer monsoon followed insolation patterns, with an increase in summer monsoon intensity during the early Holocene (Kutzbach, 1981; Wang et al., 2005). Precipitation

(28)

precipitation interval over the Indian monsoon sub-region occurred from 10.9 to 7.0 ka BP. However, in the East Asian monsoon sub-region wet climatic conditions started already 11.5 ka BP and reached their maximum between 8.3 and 5.5 ka BP. The asynchronicity of wet climatic intervals between the Indian and East Asian monsoon sub- regions will be the focus of the next data-model comparison, which will emphasize on the early Holocene between 10.5 and 7.0 ka BP.

The SSTs will be reconstructed using only Mg/Ca and alkenone data series and well-dated marine records (Fig. 11). The terrestrial record will be analyzed using the same approach as for the LGM data-model comparison.

Fig. 11: Early Holocene (11 – 7 ka BP) mean SSTs reconstructed from Mg/Ca of planktonic foraminifera and unsatuated alkenone data series.

7. Acknowledgement

First of all, I would like to thank my main supervisor, Barbara Wohlfarth, who offers me a challenging research topic, encourages and supports me throughout this

(29)

research. I am very grateful to Jenny Brandefelt, my co-supervisor, who helped with the climate model data output, and gave valuable advice and comments on manuscript I.

Swedish Research Council (VR) grants 621-2008-2855, 348-2008-6071 and 621- 2011-4684 provide financial support for this research. I also thank Hildred Crill for language advice, Jason Cosford for providing the Jintanwan speleothem dataset, and Ludvig Löwemark, Agatha de Boer and Sakonwan Chawchai for helpful discussions.

8. References

1. Ahmad, S.M., Babu, G.A., Padmakumari, V.M., Raza, W., 2008. Surface and deep water changes in the northeast Indian ocean during the last 60 ka inferred from carbon and oxygen isotopes of planktonic and benthic foraminifera. Palaeogeography, Palaeoclimatology, Palaeoecology 262, 182- 188.

2. Anshari, G., Peter Kershaw, A., van der Kaars, S., 2001. A Late Pleistocene and Holocene pollen and charcoal record from peat swamp forest, Lake Sentarum wildlife reserve, west Kalimantan, Indonesia. Palaeogeography, Palaeoclimatology, Palaeoecology 171, 213-228.

3. Barmawidjaja, B.M., Rohling, E.J., van der Kaars, W.A., Vergnaud Grazzini, C., Zachariasse, W.J., 1993. Glacial conditions in the northern Molucca sea region (Indonesia). Palaeogeography, Palaeoclimatology, Palaeoecology 101, 147-167.

4. Beaufort, L., de Garidel-Thoron, T., Mix, A.C., Pisias, N.G., 2001. ENSO-like forcing on oceanic primary production during the Late Pleistocene. Science 293, 2440-2444.

5. Boos, W.R., Kuang, Z., 2010. Dominant control of the South Asian monsoon by orographic insulation versus plateau heating. Nature 463, 218-222.

6. Braconnot, P., Otto-Bliesner, B., Harrison, S., Joussaume, S., Peterchmitt, J.-Y., Abe-Ouchi, A., Crucifix, M., Driesschaert, E., Fichefet, T., Hewitt, C.D., Kageyama, M., Kitoh, A., Loutre, M.-F., Marti, O., Merkel, U., Ramstein, G., Valdes, P., Weber, L., Yu, Y., Zhao, Y., 2007. Results of PMIP2

(30)

2: feedbacks with emphasis on the location of the ITCZ and mid- and high latitudes heat budget. Clim. Past 3.

7. Brandefelt, J., Otto-Bliesner, B.L., 2009. Equilibration and variability in a Last Glacial Maximum climate simulation with CCSM3. Geophys. Res. Lett. 36, L19712.

8. Broccoli, A.J., Dahl, K.A., Stouffer, R.J., 2006. Response of the ITCZ to northern hemisphere cooling. Geophys. Res. Lett. 33, L01702.

9. Budziak, D., Schneider, R.R., Rostek, F., Müller, P.J., Bard, E., Wefer, G., 2000. Late Quaternary Insolation forcing on total organic carbon and C₃₇ alkenone variations in the Arabian Sea. Paleoceanography 15, 307-321.

10. Caley, T., Malaiz'e, B., Revel, M., Ducassou, E., Wainer, K., Ibrahim, M., Shoeaib, D., Migeon, S.b., Marieu, V., 2011. Orbital timing of the Indian, East Asian and African boreal monsoons and the concept of a "global monsoon". Quaternary Science Reviews 30, 3705-3715.

11. Cane, M.A., 2010. Climate: A moist model monsoon. Nature 463, 163-164.

12. Cayre, O., Beaufort, L., Vincent, E., 1999. Paleoproductivity in the equatorial Indian ocean for the last 260,000 yr: A transfer function based on planktonic foraminifera. Quaternary Science Reviews 18, 839-857.

13. Chao, W.C., Chen, B., 2001. The origin of monsoons. Journal of the Atmospheric Sciences 58, 3497-3507.

14. Chen, M.-T., Chen, Y.Y., Fang, T.S., Huang, C.Y., Lee, T.Q., 2002. The structure of the last deglaciation in South China Sea high-resolution records:

IMAGES core MD972148, In: Chen, C.-T.A. (Ed.), Marine Environment:

The Past, Present, and Future. The Fuwen Press, Kaohsiung, pp. 79-94.

15. Chen, M.-T., Huang, C.-Y., 1998. Ice-Volume forcing of winter monsoon climate in the South China Sea. Paleoceanography 13, 622-633.

16. Chen, M.-T., Shiau, L.-J., Yu, P.-S., Chiu, T.-C., Chen, Y.-G., Wei, K.-Y., 2003. 500,000-Year records of carbonate, organic carbon, and foraminiferal sea-surface temperature from the southeastern South China Sea (near Palawan Island). Palaeogeography, Palaeoclimatology, Palaeoecology 197, 113-131.

(31)

17. Chodankar, A.R., Banakar, V.K., Oba, T., 2005. Past 100 ky surface salinity- gradient response in the eastern Arabian Sea to the summer monsoon variation recorded by δ¹⁸O of G. sacculifer. Global and Planetary Change 47, 135-142.

18. Chung, C.-H., Lim, H.S., Yoon, H.I., 2006. Vegetation and climate changes during the late Pleistocene to Holocene inferred from pollen record in Jinju area, South Korea. Geosciences Journal 10, 423 - 431.

19. Clark, P.U., Dyke, A.S., Shakun, J.D., Carlson, A.E., Clark, J., Wohlfarth, B., Mitrovica, J.X., Hostetler, S.W., McCabe, A.M., 2009. The Last Glacial Maximum. Science 325, 710-714.

20. Clemens, S., Prell, W., Murray, D., Shimmield, G., Weedon, G., 1991. Forcing mechanisms of the Indian Ocean monsoon. Nature 353, 720-725.

21. Clift, P.D., Plumb, R.A., 2008. The Asian monsoon. Cambridge University Press.

22. Colin, C., Siani, G., Sicre, M.A., Liu, Z., 2010. Impact of the East Asian monsoon rainfall changes on the erosion of the Mekong River basin over the past 25,000 yr. Marine Geology 271, 84-92.

23. Collins, W.D., Bitz, C.M., Blackmon, M.L., Bonan, G.B., Bretherton, C.S., Carton, J.A., Chang, P., Doney, S.C., Hack, J.J., Henderson, T.B., Kiehl, J.T., Large, W.G., McKenna, D.S., Santer, B.D., Smith, R.D., 2006. The Community Climate System Model Version 3 (CCSM3). Journal of Climate 19, 2122-2143.

24. Cook, C.G., Jones, R.T., Langdon, P.G., Leng, M.J., Zhang, E., 2011. New insights on late Quaternary Asian palaeomonsoon variability and the timing of the Last Glacial Maximum in southwestern China. Quaternary Science Reviews 30, 808-820.

25. Cosford, J., Qing, H., Lin, Y., Eglington, B., Mattey, D., Chen, Y.G., Zhang, M., Cheng, H., 2010. The East Asian monsoon during MIS 2 expressed in a speleothem δ¹⁸O record from Jintanwan Cave, Hunan, China. Quaternary Research 73, 541-549.

(32)

26. De Deckker, P., Tapper, N.J., van der Kaars, S., 2002. The status of the Indo- Pacific Warm Pool and adjacent land at the Last Glacial Maximum. Global and Planetary Change 35, 25-35.

27. Ding, X., Guichard, F., Bassinot, F., Labeyrie, L., Fang, N., 2002. Evolution of heat transport pathways in the Indonesian archipelago during last deglaciation. Chinese Science Bulletin 47, 1912-1917.

28. EPICA Community members, 2006. One-to-one coupling of glacial climate variability in Greenland and Antarctica. Nature 444, 195-198.

29. Fleitmann, D., Burns, S.J., Mangini, A., Mudelsee, M., Kramers, J., Villa, I., Neff, U., Al-Subbary, A.A., Buettner, A., Hippler, D., Matter, A., 2007.

Holocene ITCZ and Indian monsoon dynamics recorded in stalagmites from Oman and Yemen (Socotra). Quaternary Science Reviews 26, 170-188.

30. Fleitmann, D., Burns, S.J., Pekala, M., Mangini, A., Al-Subbary, A., Al-Aowah, M., Kramers, J., Matter, A., 2011. Holocene and Pleistocene pluvial periods in Yemen, southern Arabia. Quaternary Science Reviews 30, 783-787.

31. Fujii, R., Sakai, H., 2002. Paleoclimatic changes during the last 2.5 myr recorded in the Kathmandu Basin, Central Nepal Himalayas. Journal of Asian Earth Sciences 20, 255-266.

32. Govil, P., Naidu, P.D., 2011. Variations of Indian monsoon precipitation during the last 32 kyr reflected in the surface hydrography of the Western Bay of Bengal. Quaternary Science Reviews 30, 3871-3879.

33. Hanebuth, T.J.J., Stattegger, K., Bojanowski, A., 2009. Termination of the Last Glacial Maximum sea-level lowstand: The Sunda-Shelf data revisited.

Global and Planetary Change 66, 76-84.

34. Hanebuth, T.J.J., Voris, H.K., Yokoyama, Y., Saito, Y., Okuno, J.i., 2011.

Formation and fate of sedimentary depocentres on Southeast Asia's Sunda shelf over the past sea-level cycle and biogeographic implications. Earth- Science Reviews 104, 92-110.

35. Hayashi, R., Takahara, H., Hayashida, A., Takemura, K., 2010. Millennial-scale vegetation changes during the last 40,000 yr based on a pollen record from Lake Biwa, Japan. Quaternary Research 74, 91-99.