REGIONAL PRECIPITATION RESPONSE TO ENHANCED MONSOON CIRCULATION
THROUGH THE HOLOCENE USING CLOSED-BASIN PALEOLAKES ON THE TIBETAN
PLATEAU
by
Adam M. Hudson
____________________________
A Dissertation Submitted to the Faculty of the
DEPARTMENT OF GEOSCIENCES
In Partial Fulfillment of the Requirements
For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2015
THE UNIVERSITY OF ARIZONA
GRADUATE COLLEGE
As members of the Dissertation Committee, we certify that we have read the dissertation
prepared by Adam M. Hudson, titled Regional Precipitation Response To Enhanced Monsoon
Circulation Through The Holocene Using Closed-Basin Paleolakes On The Tibetan Plateau and
recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor
of Philosophy.
_______________________________________________________________________
Date: April 20th, 2015
Jay Quade
_______________________________________________________________________
Date: April 20th, 2015
John W. Olsen
_______________________________________________________________________
Date: April 20th, 2015
Andrew Cohen
_______________________________________________________________________
Date: April 20th, 2015
Jonathan Overpeck
_______________________________________________________________________
Date: April 20th, 2015
Warren Beck
Final approval and acceptance of this dissertation is contingent upon the candidate’s submission
of the final copies of the dissertation to the Graduate College.
I hereby certify that I have read this dissertation prepared under my direction and recommend
that it be accepted as fulfilling the dissertation requirement.
________________________________________________ Date: April 20th, 2015
Dissertation Director: Jay Quade
2
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of the requirements for an
advanced degree at the University of Arizona and is deposited in the University Library to be
made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission, provided
that an accurate acknowledgement of the source is made. Requests for permission for extended
quotation from or reproduction of this manuscript in whole or in part may be granted by the head
of the major department or the Dean of the Graduate College when in his or her judgment the
proposed use of the material is in the interests of scholarship. In all other instances, however,
permission must be obtained from the author.
SIGNED: Adam M. Hudson
3
ACKNOWLEDGEMENTS
Thanks to my committee members, for encouraging me to pursue the many projects of
my PhD work, and advising me not only in the science, but in the process of doing science. I
especially thank my advisor Jay Quade, and my minor advisor John Olsen for spending countless
hours in the field teaching me, providing feedback on my research and writing, and helping me
develop my own scientific ideas. And lastly, thanks to my MS thesis advisor Nat Lifton for
teaching me more than I ever wanted to know about vacuum lines and radiocarbon dating.
Thanks to my family; my father Mark, who as a geologist himself taught me to love the
natural world and the methods of its study, my mother Elane, who raised me and encouraged me
to follow my own path in life, my sister Claire who will forever be my partner in crime. And
especially thanks to my wife Alicia, who has been with me literally for my entire time in
graduate school, and will always be my primary source of support.
Thanks to the friends I’ve made in the department and in Tucson, who provided
everything from help with science, to social support and general fun times. These people are too
many to name, but a special shout out to Mauricio Ibanez-Mejia, Eli Bloch, Dave Pearson, Drew
Laskowski, Devon Orme, Dan Griffin, Kendra Murray, Matt Dettinger, Amanda Pearson, Ryan
Porter, Amy Clark, Alyson Cartwright, Deanna Grimstead, Tyler Huth, Lynn Crew, Aaron
Miller, Whitney Henderson, Kat Compton, Ryan Leary, Willy Guenthner, Jess Conroy, Sarah
Truebe, and Diane Thompson.
A special thanks to the staff and scientists of the Arizona AMS Laboratory for supporting
all of the analyses that went into producing my dissertation; Todd Lange, Lori Hewitt, Rich
Cruz, Becky Watson, Marcus Lee, and Dana Biddulph.
Thanks to my collaborators and friends in China, Zhang Hucai and Lei Guoliang, without
whom my incredible experiences on the Tibetan Plateau wouldn’t have happened.
Funding for the research in this dissertation was generously provided by the Comer
Science and Education Foundation and the University of Arizona Je Tsongkhapa Endowment for
Inner Asian Studies. Additional small grants came from a Geological Society of America
student grant, and the University of Arizona Galileo Circle and Paul S. Martin Scholarships.
4
DEDICATION
To my father, Mark, who found his own love of geology, and passed it on to me.
5
TABLE OF CONTENTS
LIST OF TABLES.........................................................................................................................9
LIST OF FIGURES.....................................................................................................................10
ABSTRACT..................................................................................................................................12
INTRODUCTION.......................................................................................................................14
Research Background........................................................................................................14
Dissertation Study..............................................................................................................17
LONG-TERM EAST-WEST ASYMMETRY IN MONSOON RAINFALL ON
THE TIBETAN PLATEAU..................................................................................17
LAKE LEVEL RECONSTRUCTION FOR 12.8-2.3 KA OF THE NGANGLA
RING TSO CLOSED-BASIN LAKE SYSTEM, SOUTHWEST TIBETAN
PLATEAU.............................................................................................................18
APPLICATION OF
14
C DATING OF INTERDUNE PALEOWETLAND
DEPOSITS TO CONSTRAIN THE AGE OF MID- TO LATE HOLOCENE
MICROLITHIC
ARTIFACTS
FROM
THE
ZHONGBA
SITE,
SOUTHWESTERN QINGHAI-TIBET PLATEAU.............................................19
REFERENCES.............................................................................................................................20
APPENDIX A: LONG-TERM EAST-WEST ASYMMETRY IN MONSOON RAINFALL
ON THE TIBETAN PLATEAU.................................................................................................23
1.1 Abstract........................................................................................................................24
1.2 Introduction.................................................................................................................25
1.2.1
Modern and Paleoclimate on the Tibetan Plateau......................................25
1.2.2
Closed-Basin Lakes...................................................................................26
1.3 Dating Control Of Paleoshorelines.............................................................................27
6
1.4 Regional Pattern Of Lake Expansion..........................................................................28
1.5 Interpreting Regional Asymmetry..............................................................................30
1.6 Acknowledgements.....................................................................................................31
1.7 References...................................................................................................................31
1.8 Supplementary Methods.............................................................................................36
APPENDIX B: LAKE LEVEL RECONSTRUCTION FOR 12.8-2.3 KA OF THE
NGANGLA RING TSO CLOSED-BASIN LAKE SYSTEM, SOUTHWEST TIBETAN
PLATEAU....................................................................................................................................58
2.1 Abstract.......................................................................................................................59
2.2 Introduction.................................................................................................................60
2.3 Study Area..................................................................................................................61
2.3.1 Lake Basin Hydrography and Geology........................................................61
2.3.2 General Appearance of Lake Deposits........................................................62
2.3.3 Local Climate...............................................................................................62
2.3.4 Sampling Sites.............................................................................................62
2.4 Methods.......................................................................................................................63
2.4.1 Tufa Sampling and Preparation....................................................................63
2.4.2 Radiocarbon Dating.....................................................................................63
2.4.3 U-Th series Dating.......................................................................................64
2.5 Results.........................................................................................................................64
2.5.1
14
C and U-Th Dating Considerations...........................................................64
2.5.2 Paleoshoreline Dating Results.....................................................................67
2.5.3 Dating Results from Sedimentary Sections.................................................67
2.6 Synthesis and Discussion............................................................................................70
2.6.1 Ngangla Ring Tso Lake Level Chronology.................................................70
2.6.2 Regional climate interpretation and comparison.........................................72
2.6.3 Implications for Holocene ISM Climate in Tibet........................................76
2.7 Conclusions................................................................................................................78
2.8 Acknowledgements.....................................................................................................78
2.9 References....................................................................................................................78
7
APPENDIX C: APPLICATION OF 14C DATING OF INTERDUNE PALEOWETLAND
DEPOSITS TO CONSTRAIN THE AGE OF MID- TO LATE HOLOCENE
MICROLITHIC ARTIFACTS FROM THE ZHONGBA SITE, SOUTHWESTERN
QINGHAI-TIBET PLATEAU..................................................................................................101
3.1 Abstract.....................................................................................................................102
3.2 Introduction...............................................................................................................103
3.2.1 Previous Archaeological Research............................................................103
3.2.2 Plateau Colonization..................................................................................104
3.3 Environmental Setting And Study Area...................................................................105
3.3.1 Qinghai-Tibet Plateau Climate and Paleoclimate......................................105
3.3.2 Study Sites.................................................................................................106
3.4 Sedimentology And Appearance Of Deposits..........................................................107
3.5 Methods.....................................................................................................................108
3.5.1 Site Identification And Artifact Sampling Methods..................................108
3.5.2 Dating Methods..........................................................................................108
3.6 Results.......................................................................................................................109
3.6.1
14
C Dating of the Modern System..............................................................109
3.6.2 Zhongba 10-1 Stratigraphy and Dating......................................................110
3.6.3 Zhongba 10-9 Stratigraphy and Dating......................................................110
3.6.4 Tsangpo 18-2 Stratigraphy and Dating......................................................111
3.7 Discussion.................................................................................................................111
3.7.1 Reservoir Correction..................................................................................111
3.7.2 Site Occupation Intervals...........................................................................112
3.7.3 Zhongba Artifact Assemblages and Site Function.....................................113
3.7.4 Regional Significance................................................................................115
3.8 Conclusions...............................................................................................................117
3.9 Acknowledgements...................................................................................................118
3.10 References...............................................................................................................118
8
LIST OF TABLES
Table A1. Radiocarbon dating results............................................................................................42
Table A2. U-Th series dating results.............................................................................................43
Table A3. Modern Lakes GIS data................................................................................................44
Table A4. Paleolakes GIS data......................................................................................................47
Table A5. Composite Paleolakes GIS data....................................................................................49
Table A6. Tibetan Plateau Proxy Records.....................................................................................50
Table B1. Radiocarbon dating results for lacustrine materials......................................................89
Table B2. U-Th series dating results for lacustrine tufa................................................................90
Table C1. AMS radiocarbon dates from modern Zhongba sites.................................................123
Table C2. AMS radiocarbon dates from Zhongba and Tsangpo stratigraphic sections..............124
Table C3. Number and type of artifacts from the Zhongba archaeological site..........................125
9
LIST OF FIGURES
Figure A1. Closed-basin lakes on the Tibetan Plateau..................................................................51
Figure A2. Paleolake shoreline imagery.......................................................................................52
Figure A3. Paleolake Aw (lake area/total basin area) versus modern Aw ratios in Tibetan
lakes...............................................................................................................................................53
Figure A4. Sampling locations in the Ngangla Ring Tso and Drebyer Tsaka basins for shoreline
tufa samples...................................................................................................................................54
Figure A5. Example tufa sample cross section showing two distinct periods of tufa growth......55
Figure A6. Results of regression analysis of paleolake expansion magnitudes............................56
Figure A7. Map showing climate proxy records of monsoon rainfall strength during the
Holocene, and the time interval of inferred maximum wetness....................................................57
Figure B1. Indian Monsoon Region Overview Map.....................................................................92
Figure B2. Ngangla Ring Tso and Baqan Tso basins and study sites............................................93
Figure B3. Schematic cross section west to east showing the hydrography of the NRT basin.....94
Figure B4. Gerze meteorological station climatology for western Tibetan Plateau.....................95
Figure B5. Study area and sample photos......................................................................................96
Figure B6. Radiocarbon-dated stratigraphic sections....................................................................97
Figure B7. Photo examples of dated shoreline and streamcut sediment exposures and
stratigraphic interpretations...........................................................................................................98
Figure B8. Ngangla Ring Tso lake level chronology....................................................................99
Figure B9. Comparison of Monsoon region paleoclimate records for the period 0-13 ka..........100
Figure C1. Overview map of the Qinghai-Tibet Plateau.............................................................126
Figure C2. Tsangpo 18-2 example stratigraphy showing paleo-wetland deposit
characteristics...............................................................................................................................127
10
LIST OF FIGURES (CONTINUED)
Figure C3. Zhongba and Tsangpo stratigraphic sections.............................................................128
Figure C4. Zhongba 10-1 site photos...........................................................................................129
Figure C5. Zhongba 10-9 site photos...........................................................................................130
Figure C6. Representative lithic and ceramic artifacts from Zhongba........................................131
Figure C7. High altitude microlithic archaeological sites on the Qinghai-Tibet Plateau............132
11
ABSTRACT
The history of climatic changes in the Asian Summer Monsoon system over the Tibetan
Plateau during the Holocene has been the subject of significant research due to the importance of
the plateau as the headwaters for many major rivers providing water resources to the surrounding
large, populous countries. In general, previous research has concluded that monsoon rainfall and
summer temperatures peaked during the early Holocene (9-11 ka BP) in Tibet, coincident with
peak Northern Hemisphere summer insolation. Atmospheric teleconnections with upstream
Northern Hemisphere westerly circulation patterns influenced by North Atlantic sea surface
temperature changes have also been noted at millennial and centennial timescales. However,
recent studies have noted that the timing of peak monsoon warmth and wetness during the
Holocene are not synchronous across the entirety of the Tibetan Plateau, and studies of modern
precipitation indicate several distinct regions of monsoon precipitation variability at interannual
scales, suggesting the monsoon response to past and future climate change may be regionally
heterogeneous for the plateau. Clear assessment of this regionality within the monsoon climate
region is a topic of continuing research, but it has been hindered by lack of climate records in
remote areas, dating difficulties, and concerns over the comparability of interpreted climateproxy relationships between the many different biological, hydrological, and geochemical
proxies applied. The first part of this dissertation uses 14C and U-Th series geochronology,
sedimentology, and GIS analysis of exposed lake shoreline sediments surrounding the numerous
closed-basin lake systems of the central and western Tibetan Plateau to investigate regional
heterogeneity in monsoon rainfall, and to develop a new well-dated lake level record from the
Ngangla Ring Tso lake system in the poorly studied southwestern region. The major conclusions
are: 1) peak early Holocene monsoon rainfall, recorded by the highest paleoshorelines
surrounding 130 lake systems, intensified more relative to today in the western part (west of
86°E longitude) of the Tibetan Plateau when compared to eastern regions, closely following
regions of modern rainfall variability; 2) monsoon rainfall in the Ngangla Ring Tso region
peaked during the early Holocene insolation maximum, consistent with other records, remained
significantly higher than modern until ~6.0 ka BP, but with abrupt reductions in monsoon rainfall
associated with North Atlantic ice-rafted debris peaks.
12
The warm and wet period of the early and middle Holocene was also likely coincident
with the first major colonization of the Tibetan Plateau by prehistoric humans. Current research
suggests early foragers employing stone tools first forayed into the middle elevation areas above
3,000 m elevation on the northeastern fringe of the plateau as early as 14.8 ka BP, and therefore
the dominant hypothesis suggests plateau colonization proceeded from this direction, heading
westward through the Holocene. However, well studied and dated archaeological sites from the
high plateau are exceedingly rare, requiring further investigation. The second part of this
dissertation presents new age controls for the Holocene Zhongba microlithic site in the
southwestern Tibetan Plateau, using 14C dating of organic and carbonate-rich paleo-wetlands
sediments hosting in situ stone artifacts. The major conclusions of this study are: 1) artifacts at
the Zhongba site, which are typologically similar to microlithics across the plateau, can be no
older than 6.5 ka BP, consistent with the prevailing east-to-west colonization hypothesis, and 2)
microlithic tools continued to be important as late as 1.3 ka BP at the site, even though metal is
found in sites of similar age elsewhere in Tibet.
13
INTRODUCTION
Research Background:
The Tibetan Plateau is the largest, highest continental plateau on Earth. This exceptional
topography, attained through the continuous collision between the Indian and Eurasian plates
since the early Cenozoic, forms a formidable and important physiographic, hydrologic, and
biologic barrier separating tropical south Asia from the arid Central Asian interior. The plateau
has a unique natural landscape. The high elevation results in mean annual temperatures today at
or below 0°C, with winter-to-summer amplitude of up to 30°C.
The dominant source of
precipitation comes from the Indian Ocean, as summer monsoon rainfall associated mostly with
the Indian Monsoon subsystem of the greater Asian Summer Monsoon (Conroy and Overpeck,
2011). The extreme rain shadow of the high Himalayan ranges on the southern margin creates a
strong north-south precipitation gradient, wherein the plateau receives only 100-500 mm of
precipitation per year. Precipitation is least on the western half of the plateau, which is the
primary geographic focus of this dissertation, exhibiting a decreasing rainfall gradient from
southeast to northwest (Fig. A1).
Despite the aridity of the interior, the high glaciated ranges on the periphery of the
plateau feed many of the major rivers of Asia, including the Yellow, Yangtze, Brahmaputra,
Ganges, and Indus. These rivers flow from all sides of the plateau and provide essential water
resources to many countries, including the world’s most populous, India and China. Therefore,
there is intense interest in how the Asian Monsoon system, which brings between 60-90% of
annual precipitation to the plateau, depending on location (Bookhagen, 2010), will respond to
future warmer climate conditions.
Previous research has repeatedly shown that the high
topography of the plateau helps to drive the intense tropical summer precipitation of the Asian
Monsoon into the mid-latitudes through a combination of land-sea heat contrast, and intense
orographic uplift along the southern plateau margin (e.g. Boos et al., 2010). It is less clear,
however, what the response of monsoon precipitation, in terms of amount, seasonality, and
geographical extent, will be under future climate change. These questions are critical for future
adaptation, because the major drainages of the Tibetan Plateau provide water to many different
political states, so that differential change in the intensity of the monsoon at a regional level may
have variable effect for adaptation. By looking at past variations in the monsoon, researchers
14
aim to constrain the forcing mechanisms and the regional character of monsoon change to give
perspective on the modern conditions.
Existing paleoclimate records of monsoon variation during the Holocene in Tibet been
based on a variety of records, including ice core isotopes and pollen (e.g. Thompson et al., 1997,
Liu et al., 1998), lake core proxies for depth, hydrologic budget, and terrestrial flora (e.g.
Demske et al., 2009; Gasse et al., 1996; Li et al., 2011; Lu et al., 2011; Mischke et al., 2008;
Morrill et al., 2006; Mügler et al., 2010; Shen et al., 2005; Van Campo and Gasse, 1993;
Wischnewski et al., 2011), studies of surficial lake sediment exposures, and shoreline dating (e.g.
Kong et al., 2007; Lee et al., 2009; Pan et al., 2012; Wünneman et al., 2010) (Fig. 1). These
records show annual rainfall, mostly summer monsoon-derived, and mean annual temperature
increased as Northern Hemisphere summer insolation peaked 10.0-11.0 ka. Almost all records
suggest wetter-than-present conditions persisted until the mid-Holocene, but there is significant
debate over whether the timing was regionally homogeneous within the Tibetan Plateau (e.g.,
Mügler et al., 2010; Li et al., 2011; Wischnewski et al., 2011). This debate is hindered by dating
uncertainties (e.g. Hou et al., 2012) and differing environmental responses between the many
monsoon climate indicators that have been applied. For these reasons, a truly representative
comparison of Holocene monsoon climate requires a paleoclimate indicator that is widely and
evenly distributed across the Tibetan Plateau, and has a consistent response to climate. The first
two studies in this dissertation take advantage of the fact that the western half of the Tibetan
Plateau is internally drained, and is covered by hundreds of freshwater and saline lakes totaling
more than 40,000 km2, supported by the limited precipitation and glacial melt (Zhang et al.,
2013). These lakes are widely distributed, and because they are internally drained, their depth
and surface area is consistently and directly linked to the balance of incoming monsoon
precipitation and outgoing evaporation within their drainage basins (Mifflin and Wheat, 1979;
Benson and Paillet, 1989). Therefore records of past lake area and depth are ideally suited to
investigate the timing and forcing mechanism for Indian Summer Monsoon precipitation during
the Holocene, and to compare the magnitude of precipitation change across the entire network on
the Tibetan Plateau.
The warmer, wetter conditions of the Holocene in the Tibetan Plateau are also coincident
with expanded evidence of prehistoric human presence, typified by numerous and widely
distributed stone and ceramic artifacts (Aldenderfer and Zhang, 2004). Because the plateau is
15
such a remote, harsh, and low productivity environment, there is significant archaeological and
anthropological interest in the process by which it was colonized by prehistoric human
populations. Modern native Tibetan populations are physiologically adapted for life at high
altitude (Beall et al., 2004), and the semi-nomadic pastoralist lifestyle that is still in practice
today is well suited, and perhaps necessary, for year-round occupation of the highest elevation
areas of the Tibetan Plateau (Brantingham and Gao, 2006).
Both these physiological and
behavioral adaptations to life at high altitude require long periods of time to develop, and so
researchers look to the prehistoric archaeological record for evidence of when the plateau was
first colonized, where the first people came from, and how their population spread.
The
archaeological record of the western Tibetan Plateau is particularly poorly studied. Also, most
prehistoric archaeological sites in Tibet are non-stratified surface collections, for which age is
difficult to constrain. Therefore finding and investigating sites like Zhongba in the southwestern
plateau, studied in the third chapter of this dissertation, where artifacts can be found within a
dateable context, are critical to understanding Tibetan prehistory.
Dissertation Study Summary:
This dissertation is a significant contribution to understanding the Holocene evolution of
the Indian Summer Monsoon in the Tibetan Plateau, both in terms of the change in monsoon
intensity through time, and the differential intensification at different locations. It is also a
significant contribution to the study of Tibetan prehistoric archaeology, providing one of the first
radiometrically-dated records of prehistoric occupation for the high Tibetan Plateau, and the first
for the southwestern region. The main body of this dissertation consists of three stand-alone
manuscripts that are presented here as appendices A-C. These individual chapters have all been
previously published and are formatted in accordance with the guidelines for the journals where
they were submitted.
However, all share the common theme of using geochronology,
sedimentology, and GIS analysis to study the paleoclimate of the Asian Summer Monsoon, and
the hydrologic and anthropologic systems of the Tibetan Plateau during the Holocene. A brief
description of the study design, and major conclusions for each chapter is provided below.
16
APPENDIX A: LONG-TERM EAST-WEST ASYMMETRY IN MONSOON RAINFALL ON
THE TIBETAN PLATEAU
This manuscript uses Geographic Information Systems (GIS) analysis of geospatial
datasets and satellite imagery for 130 large closed-basin lake systems to investigate the modern
and past patterns of monsoon rainfall-driven lake area change across the entirety of the central
and western Tibetan Plateau. The study compares modern lake area ratios (the ratio of lake
surface area to total hydrographic basin area) to modern satellite-derived precipitation estimates
to investigate the way in which lake area reflects climate. Modern lake area ratios were also
compared to past lake areas reconstructed from the elevations of the highest Holocene shoreline
remnants surrounding the same lakes, to investigate how lake areas changed spatially across the
Tibetan Plateau under the insolation-driven monsoon precipitation maximum clearly
demonstrated by previous research. The experimental data collection method consisted of
digitizing the closed drainage basins for the lake systems under study, extracting modern lake
areas from historical Landsat imagery, and extracting paleolake areas using the elevations of
high paleoshorelines and the area enclosed within them in a high resolution digital elevation
model for the Tibetan Plateau. The major conclusions of this study showed that: 1) modern lake
area ratios in both glaciated and nonglaciated lake systems has a positive, statistically significant
linear relationship with mean annual precipitation, suggesting they are a good climate indicator
to be used in past climate studies; but 2) glaciated systems have higher lake area ratios overall
and a higher slope for the relationship than nonglaciated systems. The paleolake area comparison
showed that: 3) lake area ratios for the early Holocene were markedly different in geographic
pattern when compared to modern ratios, indicating a significant change to the modern east-towest gradient of decreasing precipitation across the Tibetan Plateau. We suggest this pattern
reflects more enhanced monsoon precipitation in the western zone of the Tibetan Plateau during
the early Holocene compared to modern than in the eastern zone, broadly corresponding with
observed modern areas of distinct modern precipitation variability.
This manuscript was
published in Geology in March, 2013.
17
APPENDIX B: LAKE LEVEL RECONSTRUCTION FOR 12.8-2.3 KA OF THE NGANGLA
RING TSO CLOSED-BASIN LAKE SYSTEM, SOUTHWEST TIBETAN PLATEAU
This manuscript presents a new Holocene lake level reconstruction for a previously
unstudied large closed-basin lake system in the southwestern Tibetan Plateau.
The study
combines radiocarbon and U-Th series dating of shoreline tufas, mollusk shells, and plant
macrofossils with sedimentology, stratigraphy, and geomorphology of exposed shoreline and
nearshore lake sediments to produce a record of lake level changes for the Ngangla Ring Tso
lake system over the period 12,800 to 2,300 years before present. This work showed that the
lake system was deeper and larger than present continuously over that time interval, peaking
from 10,500 to 8,500 yrs ago, when the lake was 135 meters deeper, and more than four times
larger in surface area than present..
Lake level variations, a proxy for Indian Monsoon
precipitation, varied in concert with changing Northern Hemisphere solar insolation, in
agreement with other proxy records from the southwestern Tibetan Plateau and Asian Monsoon
region. Weak monsoon intervals were also inferred from abrupt drops in lake level that are
broadly coincident with other proxy records, possibly driven by atmospheric changes in response
to North Atlantic overturning circulation weakening during the Holocene. This manuscript was
published in Quaternary Research in January, 2015.
APPENDIX C: APPLICATION OF
14
C DATING OF INTERDUNE PALEOWETLAND
DEPOSITS TO CONSTRAIN THE AGE OF MID- TO LATE HOLOCENE MICROLITHIC
ARTIFACTS FROM THE ZHONGBA SITE, SOUTHWESTERN QINGHAI-TIBET
PLATEAU
This manuscript presents a preliminary record of the prehistoric occupation of upper
Yarlung Tsangpo valley of the southwestern Tibetan Plateau by people employing microlithic
stone tools. This study defines the time of deposition of microlithic stone artifacts in the
Zhongba archaeological site based on radiocarbon dating of organic-rich sediments, plant
macrofossils, and mollusk shells in Holocene wetlands deposits. This represents one of very few
studies constraining the age of prehistoric occupation of the Tibetan Plateau with radiometric
ages, showing that in situ artifacts in the wetlands stratigraphy were deposited during the mid-to-
18
Late Holocene between 6,500 and 1,300 years before present. The two major conclusions were
that: 1) the dated occupation at the Zhongba site was later than for sites to the northeast,
tentatively supporting the current dominant hypothesis that colonization of the Tibet Plateau
proceeded from the northeast during the Holocene; and 2) microlithic industry remained an
important toolmaking strategy for plateau inhabitants as late as 1,300 years ago, despite clear
evidence of the use of metals in other sites in Tibet of similar age. This manuscript was
published in Geoarchaeology in June, 2014.
19
REFERENCES
Aldenderfer, M., Zhang Y., 2004. The prehistory of the Tibetan Plateau to the seventh century
A.D.: perspectives and research from China and west since 1950. Journal of World Prehistory,
18, 1-55.
Beall, C.M., Song, K., Elston, R.C., Goldstein, M.C., 2004. Higher offspring survival among
Tibetan women with high oxygen saturation genotypes residing at 4,000 m. Proceedings of the
National Academy of Sciences of the United States of America 101, 14300-14304.
Benson, L.V., and Paillet, F.L., 1989. The use of total lake-surface area as an indicator of
climatic change: examples from the Lahontan Basin: Quaternary Research, v. 32, p. 262-275.
Bookhagen, B., 2010. Appearance of extreme monsoonal rainfall events and their impact on
erosion in the Himalaya. Geomatics, Natural Hazards and Risk 1, 37-50.
Boos, W.R., Kuang, Z., 2010. Dominant control of the South Asian monsoon by orographic
insulation versus plateau heating. Nature 463, 218-222.
Brantingham, P.J., Gao X., 2006. Peopling of the northern Tibetan Plateau. World Archaeology
38, 387-414.
Demske, D., Tarasov, P.E., Wünneman, B., Riedel, F., 2009. Late glacial and Holocene
vegetation, Indian monsoon and westerly circulation in the Trans-Himalaya recorded in the
lacustrine pollen sequence from Tso Kar, Ladakh, NW. Palaeogeography, Palaeoclimatology,
Palaeoecology, 279, 172–185.
Gasse, F., Fontes, J.C., Van Campo, E., Wei, K., 1996. Holocene environmental changes in
Banggong Co basin (western Tibet). 4. Discussion and conclusions. Palaeogeography,
Palaeoclimatology, Palaeoecology 120, 79–92.
20
Hou J., D’Andrea, W.J., Liu Z., 2012. The influence of 14C reservoir age on interpretation of
paleolimnological records from the Tibetan Plateau. Quaternary Science Reviews 48, 67-79.
Kong, P., Na, C., Fink, D., Huang, F., Ding, L., 2007. Cosmogenic 10Be inferred lake-level
changes in Sumxi Co basin, western Tibet. Journal of Asian Earth Sciences 29, 698–703.
Lee, J., Li, S., Aitchison, J.C., 2009. OSL dating of paleoshorelines at Lagkor Tso, western
Tibet. Quaternary Geochronology 4, 335–343.
Li, Q., Lu, H., Zhu, L., Wu, N., Wang, J., Lu, X., 2011. Pollen-inferred climate changes and
vertical shifts of alpine vegetation belts on the northern slope of the Nyainqentanglha Mountains
(central Tibetan Plateau) since 8.4 kyr BP. The Holocene 21, 939–950.
Liu K.B., Yao Z., Thompson, L. G. 1998. A pollen record of Holocene climatic changes from the
Dunde ice cap, Qinghai-Tibetan Plateau. Geology 26, 135-138.
Lu X., Zhu L., Nishimura M., Morita M., Watanabe T., Nakamura T., Wang Y., 2011. A highresolution environmental change record since 19 cal ka BP in Pumoyum Co, southern Tibet.
Chinese Science Bulletin: Geology 56, 2931-2940.
Mifflin, M.D., Wheat, M.M., 1979. Pluvial lakes and estimated pluvial climates of Nevada.
Nevada Bureau of Mines and Geology, Reno, Nevada, 57 p.
Mischke, S., Kramer, M., Zhang C., Shang H., Herzschuh, U., Erzinger, J., 2008, Reduced early
Holocene moisture availability in the Bayan Har Mountains, northeastern Tibetan Plateau,
inferred from a multi-proxy lake record. Palaeogeography, Palaeoclimatology, Palaeoecology
267, 59-76.
Morrill, C., Overpeck, J.T., Cole, J.E., Liu, K.-B., Shen, C.M., Tang, L.Y., 2006. Holocene
variations in the Asian monsoon inferred from the geochemistry of lake sediments in central
Tibet. Quaternary Research 65, 232–243.
21
Mügler, I., Gleixner, G., Gunther, F., Mäusbacher, R., Daut, G., Schütt, B., Berking, J., Schwalb,
A., Schwark, L., Xu B., Yao T., Zhu L., Yi C., 2010. A multi-proxy approach to reconstruct
hydrological changes and Holocene climate development of Nam Co, central Tibet. Journal of
Paleolimnology 43, 625–648.
Pan B., Yi C., Jiang T., Dong G., Hu G., Jin Y. 2012. Holocene lake-level changes of Linggo Co
in central Tibet. Quaternary Geochronology 10, 117-122.
Shen, J., Liu, X., Wang, S., Ryo, M., 2005. Palaeoclimatic changes in the Qinghai Lake area
during the last 18,000 years. Quaternary International 136, 131–140.
Thompson, L.G., Yao T., Davis, M.E., Henderson, K.A., Mosley-Thompson, E., Lin, P.N., Beer,
J., Synal, H.-A., Cole-Dai, J., Bolzan, J. F. 1997. Tropical climate instability: The last glacial
cycle from a Qinghai-Tibetan ice core. Science 276, 1821-1825.
Van Campo, E., Gasse, F., 1993. Pollen- and diatom-inferred climatic and hydrological changes
in Sumxi Co basin (western Tibet) since 13,000 yr B.P. Quaternary Research 39, 300–313.
Wischnewski, J., Mischke, S., Wang, Y., Herzschuh, U., 2011. Reconstructing climate variability
on the northeastern Tibetan Plateau since the last Lateglacial—a multi-proxy, dual-site approach
comparing terrestrial and aquatic signals. Quaternary Science Reviews, 30, 82–97.
Wünnemann, B., Demske, D., Tarasov, P., Kotlia, B.S., Reinhardt, C., Bloemendal, J.,
Diekmann, B., Hartmann, K., Krois, J., Riedel, F., Arya, N., 2010. Hydrological evolution during
the last 15 kyr in the Tso Kar lake basin (Ladakh, India), derived from geomorphological,
sedimentological and palynological records. Quaternary Science Reviews 29, 1138-1155.
Zhang, G., Yao T., Xie H., Kang S., Lei Y., 2013. Increased mass over the Tibetan Plateau: from
lakes or glaciers? Geophysical Research Letters 40, 2125-2130.
22
APPENDIX A.
LONG-TERM EAST-WEST ASYMMETRY IN MONSOON RAINFALL ON THE TIBETAN
PLATEAU
23
LONG-TERM EAST-WEST ASYMMETRY IN MONSOON RAINFALL ON THE TIBETAN
PLATEAU
Adam M. Hudsona
Jay Quadea
a
Department of Geosciences, University of Arizona, Tucson, Arizona 85721, USA
This manuscript was published in Geology in March 2013.
1.1 ABSTRACT
Variations in the strength and duration of the Asian Monsoon affect over half the
population of the planet, and there is intense interest in where, and how much, rainfall amounts
will change regionally in response to warmer climate conditions. The modern monsoon region is
divided into distinct subsystems, so the response may be regionally heterogeneous. Northern
Hemisphere monsoon systems intensified during the early to mid-Holocene, increasing summer
rainfall and creating wetter conditions across much of eastern Asia. In this study we use the area
encompassed by high shorelines of early Holocene paleolakes in Tibet to reconstruct
paleorainfall (and hence paleomonsoon) patterns during this very wet period. We found that the
relative paleolake expansions during the early Holocene of 130 closed-basin lake systems in the
central Tibetan Plateau display a strong east-west gradient. Paleolake areas expanded by
approximately fourfold in the western plateau, compared to approximately twofold expansion in
the eastern region. This early Holocene pattern mirrors the modern west-east climate division on
the plateau: rainfall in the west is closely tied to the Indian Summer Monsoon (ISM) subsystem,
and the east to a probable mix of ISM and East Asian Summer Monsoon (EASM) subsystems.
Our results suggest that these modern climate divisions are an enduring feature of the plateau and
that ISM rainfall increased much more than EASM rainfall in response to the same insolation
forcing.
24
1.2 INTRODUCTION
The monsoon systems of Asia are arguably the most studied monsoons in the world, both
modern and in deep time. The many geologic records of the Asian monsoons are impressive for
their diversity, precision, and strong links on long timescales to Northern Hemisphere summer
insolation (Wang et al., 2008) and shorter time scales to North Atlantic sea-surface temperatures
(Gupta et al., 2003; Wang et al., 2005). Despite all this study, major debate persists over what
these proxy records mean in terms of monsoon rainfall variation across the vast monsoon region.
Modern Asian monsoon climate is divided into many regional subsystems with differing
circulation patterns and distinct precipitation variability (Conroy and Overpeck, 2011).
Paleoclimate records also show that the timing of the peak wetness during the Holocene may
differ regionally in Asia (Wang et al., 2010). In contrast, speleothem δ18O records, which
provide the most precise chronology of monsoon variation, broadly agree on the timing of major
climate changes. There is considerable recent debate, however, as to whether the cyclic changes
in δ18O values reflect change in local precipitation amount (e.g., Wang et al., 2008), or regional
changes in moisture source, with no necessary rainfall increase (e.g., Dayem et al., 2010; Pausata
et al., 2011). These controversies underscore the need for data on the timing and actual
magnitude of rainfall changes at a regional scale.
Here we examine paleohydrologic evidence for climate changes from surficial lake
deposits on the Tibetan Plateau, which today is strongly under the influence of the Asian
monsoon. It is clear that the early Holocene was wetter than today over much of the Plateau
(Mügler et al., 2010; Li et al., 2011), but the question remains as to how much monsoon rainfall
increased, how this varied regionally, and how regional variability relates to Asian Monsoon
subsystems.
1.2.1 Modern and Paleoclimate on the Tibetan Plateau
The spatial coherency of monsoon precipitation on the Tibetan Plateau is less well
understood compared to continent-scale precipitation. Climate station data and paleoclimate
records are sparse, especially in the drier western region. Rainfall seasonality follows a typical
25
monsoon pattern, peaking in July, and receiving around 60% of the annual total between May
and October (Conroy and Overpeck, 2011). The plateau is in the arid portion of the monsoon
region, and displays a gradient of decreasing precipitation from east to west (Fig. A1). Conroy
and Overpeck (2011) used empirical orthogonal function (EOF) analysis to divide modern
climate in the Tibet region into three areas of distinct precipitation variability (Fig. A1). All
regions have monsoon-like summer precipitation maxima. The northern climate region (Region
2) is mainly north of the plateau, and likely has a Westerly moisture source. Summer
precipitation in the southwest region (Region 3) is correlated with summer rainfall indices of the
Indian Summer Monsoon (ISM) on the Indian subcontinent. By contrast, that in the eastern
region (Region 1) is anti-correlated with precipitation of the western North Pacific monsoon, a
feature of the East Asian Summer Monsoon (EASM), but may be a mix of ISM and EASM
subsystems. The eastern and western plateau display different trends in annual precipitation over
the past 40 yr (Li et al., 2010), in addition to seasonal and inter-annual variability. This raises
questions as to the reason for these differences, and whether they operated in the past.
Regional syntheses of lake core and other proxy records from Tibet show that the early
Holocene peak in Northern Hemisphere summer insolation coincided with increased warmth and
wetness in Tibet associated with stronger monsoon climate. Some studies suggest there are eastwest differences in the timing of peak wetness inferred from records, with the wet interval
peaking later (Herzschuh et al., 2006) and persisting longer (Li et al., 2011) in the eastern
plateau. Others contend that the overall timing of peak wetness is similar (Mügler et al., 2010) or
that there is no pattern (Wischnewski et al., 2011). Furthermore, no attempt has been made to
assess differences in the magnitude of rainfall change because the available records have no
consistent quantifiable relationship with precipitation. For this study, we focus on shorelines
surrounding closed-basin lakes on the Tibetan Plateau to provide a consistent means of assessing
the past hydrologic budget of lake systems regionally.
1.2.2 Closed-Basin Lakes
The central and western Tibetan Plateau is divided into hundreds of closed basins, most
of which contain lakes with exposed paleoshorelines that can be mapped from satellite imagery
(Fig. A2). Closed-basin lake systems provide sensitive paleohydrologic records because lake
surface area depends on the balance of inputs from runoff and precipitation and output through
26
evaporation. Paleoshorelines are often overlooked for study because of dating challenges and
stratigraphic discontinuity. However, paleoshorelines provide a quantifiable paleolake area that
has been used to investigate the magnitude of past precipitation change (Blard et al., 2009;
Matsubara and Howard, 2009).
We employed geographic information systems (GIS) and satellite imagery to map the
modern and paleolake areas from high paleoshorelines of 130 lake systems on the central Tibetan
Plateau (Fig. A1). We use the lake-to-basin area ratio, hereafter Aw (Aw = lake area/total basin
area), to normalize lake systems of differing size to a measure of the quasi-steady-state
hydrologic budget (Fig. A2a). This ratio is directly related to annual precipitation, and is a better
indicator of water balance than lake depth, especially in lake systems with multiple sub-basins.
By comparing paleo-Aw ratios to the modern Aw ratios, we can evaluate the regional pattern of
lake expansions in response to past increased precipitation. We also compare new shoreline 14C
dates from two lake systems in central Tibet with existing age control on lake highstands and
wetter climate, and discuss the mechanism for the observed lake expansion pattern. Detailed
methods for shoreline dating and GIS can be found in the Supplemental Methods Section.
1.3 DATING CONTROL OF PALEOSHORELINES
A variety of evidence, including our new dates, suggests the highest continuous
shorelines around lakes in Tibet are early Holocene in age. This generalization includes all lakes
considered in this analysis. If valid, this facilitates our comparison of Aw ratios within the
Holocene time period across Tibet.
We used the highest elevation continuous paleoshoreline visible in satellite images for
our estimates of past lake area, which in general consisted of a well-defined ridge traceable at the
same elevation around >80% of the paleolake perimeter (Fig. A2b). Some basins contain
shoreline remnants higher in elevation, but these features are isolated, discontinuous, and
dissected, and so are easily distinguished from the younger shorelines (Fig. A2c).
To test the Holocene age assumption we dated shorelines from two lake systems,
Ngangla Ring Tso and Chabyer Tsaka (Zabuye), using
14
C and U-Th dating of shoreline tufas
(Fig. A1; Fig. A4 and Tables A1 and A2). Tufas (CaCO3) precipitated on shorelines of closedbasin lakes in high-altitude, arid conditions have been dated successfully using both techniques,
27
and the
14
C reservoir effect for tufa can be found by comparison of ages from both methods
(Placzek et al., 2006). We have assessed the reservoir effect for Ngangla Ring Tso by dating two
samples in this way. Both pairs are close in age, indicating the reservoir effect is at most ~200
cal yr (Tables A1 and A2). Six
14
C dates from the highest prominent shoreline of the Ngangla
Ring Tso system range between 8.6 and 10.4 cal ka. Two 14C dates from high shorelines in the
adjacent lake system of Chabyer Tsaka also date to the early Holocene at 11.1 and 9.9 cal ka
(Fig. A1; Table A1). The reservoir effect in this system was not determined, but the shoreline
ages agree well with those from Ngangla Ring Tso.
We have also dated tufa from a shoreline remnant above the Holocene highstand of
Ngangla Ring Tso using U-Th. The resulting age of 211.2 cal ka indicates that remnant
shorelines are much older than the Holocene features in Tibetan systems.
Our ages are supported by a variety of other shoreline and core proxy evidence for
maximum lake depths in the early Holocene (see Fig. A7). Lee et al. (2009) dated four shorelines
surrounding Lagkor Tso in western Tibet (Fig. A1) using optically stimulated luminescence
(OSL) that range in age from 5.2 to 3.2 ka. The high shoreline was not dated, but at the highstand
Lagkor Tso is connected with the Chabyer Tsaka system. Our dated shorelines from Chabyer
Tsaka are above this connection, so it is likely that the Lagkor Tso high shorelines are the same
age, and the OSL dates record the lake regression, as suggested by the authors. Kong et al.
(2007) used cosmogenic
10
Be to date the highest bedrock shoreline terraces of Sumxi Tso, and
obtained ages of 12.8 ± 1.1 ka to 6.4 ± 0.7 ka. Uncertainty in inheritance, erosion and production
rates aside, these dates are early Holocene in age within error, and compare well with the period
of lake highstand inferred in the Sumxi Tso core (Van Campo and Gasse, 1993). Many core
proxy records also suggest the early to mid-Holocene was the period when lakes were deepest
(Morrill et al., 2006) and pollen suggests greatest warmth and wetness (Van Campo and Gasse,
1993; Gasse et al., 1996; Shen et al., 2005; Herzschuh et al., 2006; Demske et al., 2009; Kramer
et al., 2010; Li et al., 2011; Wischnewski et al., 2011).
1.4 REGIONAL PATTERN OF LAKE EXPANSION
The paleolake areas define a distinct east-west gradient in the magnitude of past lake
expansion. The ratio of highstand Aw to modern Aw ranges 3.2–8.6 times modern in the extreme
28
western region (79°-81°E) to only 1.1–2.4 times modern at the eastern end (90°–92°E) (Fig. A3).
This longitudinal gradient is almost opposite the modern gradient in rainfall, where the west is
currently drier.
Notably, lakes with glaciated headwaters and those in unglaciated basins are expanded by
similar magnitude, following the same regional pattern (Fig. A3). This is significant because the
lake area-climate relationship is potentially more complicated in glaciated basins where runoff is
determined by the volume of glacial ice available to provide meltwater in addition to
precipitation and evapotranspiration. Meltwater is a function of the water budget of each glacier
and can vary depending on local climate conditions (Rupper and Roe, 2008). This complication
is reflected in modern glaciated lake systems on the plateau. Modern lake area in systems with
glaciated and unglaciated headwaters both have a significant correlation with basin-averaged
annual precipitation, but lake area in glaciated basins is higher and more variable (Fig. A3,
inset). The similar pattern of expansion for both glaciated and unglaciated basins, in contrast,
indicates a common cause for early Holocene expansion.
Both modern data sets have considerable scatter, evidenced by the low r-values, that must
be explained by hydrologic variables other than precipitation. Differing evaporation rates with
differing lake or mean basin elevation is one possible cause. Differing area-to-volume
relationship may also play a role. We conducted analysis to ensure the paleolake pattern is not
due to systematic variations in these parameters at high lake levels. We again examined glaciated
and unglaciated systems separately.
To test for differing evaporation rates, we regressed highstand and mean basin elevation
against magnitude of paleolake expansion (Figs. A6a and A6b). To investigate lake area-tovolume, we regressed the mean slope (degrees) of the region in each basin subsumed by the
paleolake against expansion magnitude (Fig. A6c). In all cases, correlation is poor and not
significant. From this we conclude the pattern of lake expansions is best explained by a regional
gradient of precipitation much different than today.
We interpret the expansion of all paleolakes in the data set to be a result of increased
monsoon rainfall. This is supported by lake records covering the plateau, including the northwest
(Van Campo and Gasse, 1993; Demske et al., 2009), and is in contrast to the later Holocene peak
wetness found in records in the Westerly-dominated regions north of the plateau (e.g., Chen et
29
al., 2008). Greater lake expansion in the western plateau indicates that monsoon rainfall
increased more under high Northern Hemisphere summer insolation there than to the east.
1.5 INTERPRETING REGIONAL ASYMMETRY
The change in lake expansion during the early Holocene, from approximately fourfold in
western Tibet to approximately twofold in the east, lies at around 86°E (Fig. A3). This compares
closely with the boundary between Region 3 and Region 1 defined by the analysis of modern
precipitation for Tibet (Conroy and Overpeck, 2011; Figure 1), suggesting the modern divisions
of monsoon climate are maintained under strong monsoon conditions during the early Holocene.
This shows that the response of Asian monsoon rainfall was spatially variable despite the
same insolation forcing, the generally accepted cause for monsoon intensification. Our spatial
reconstruction covers the Tibetan Plateau only, so we are cautious to extend it to the continental
scale, but the corresponding regions of modern precipitation variability are associated with
different monsoon subsystems (Conroy and Overpeck, 2011). If the regions are maintained in the
past, our results for the ISM-influenced western region suggest ISM rainfall increased greatly
during the early Holocene. By contrast, eastern lakes with a mix of ISM/EASM influence
indicate less precipitation increase.
Greater change in ISM precipitation relative to the EASM has been previously suggested
using both paleoclimate records and climate model results. Maher (2008) and Seki et al. (2011)
presented reconstructions of precipitation magnitude during the Holocene from eastern Tibet
suggesting the ISM drove increased precipitation. The speleothem-based magnitude
reconstruction of Hu et al., (2008) in turn indicates that precipitation in EASM-influenced China
was similar to modern during the Holocene. Our results are compatible with this hypothesis, with
the increase in precipitation associated with the ISM lessening toward the east. The climate
model results of Pausata et al. (2011) also indicate that the largest change in precipitation in the
monsoon region is found over the western Tibetan Plateau and wanes to the east. Their model
results should be cautiously applied to the Holocene, since their experiments were based on
LGM conditions perturbed by a Heinrich Event to simulate weaker monsoon conditions.
Nonetheless, our reconstruction bears remarkable spatial similarity to the model results, but with
opposite sign.
30
Furthermore, our results assist in interpreting speleothem records from Tibet, as they are
developed, and can inform the debate over the relative contribution of rainfall amount versus
moisture source to speleothem δ18O records. No Holocene speleothem records are yet published
from Tibet, but results from 75 to 130 ka (Cai et al., 2010) show very large decreases in δ18O
values during insolation maxima, which, if the early Holocene lake records are representative,
were times of increased monsoon rainfall in the region. Thus, in Tibet the case can be made that
δ18O values in speleothem calcite in part respond to rainfall amount. Quantification of the
rainfall amount-δ18O relationship awaits development of Holocene speleothem records, and of
hydrologic models of the rainfall increases implied by our results.
To conclude, our results demonstrate that for Tibet there is a strong, but spatially variable
increase in rainfall with insolation forcing that likely indicates increased ISM rainfall relative to
the EASM. Our results underscore the need to look beyond speleothem isotope records for
answers as to how rainfall amount varied in the past. Development of such records in China
more clearly under the influence of the East Asian monsoon should help resolve the debate
concerning the paleohydrologic meaning of speleothem isotope records from.
1.6 ACKNOWLEDGMENTS
This research was funded by the Comer Science and Education Foundation. We thank
Zhang Hucai and Lei Guoliang for support and collaboration. We also thank the editor and
three anonymous reviewers for comments and suggestions that greatly improved this
manuscript.
1.7 REFERENCES
An Z., Porter, S.C., Kutzbach, J.E., Wu X., Wang S., Liu X., Li X., and Zhou W., 2000,
Asynchronous Holocene optimum of the East Asian monsoon: Quaternary Science Reviews,
v. 19, p. 743-762, doi: 10.1016/S0277-3791(99)00031-1.
Blard, P.-H., Lave, J., Farley, K.A., Fornari, M., Jimenez, N., and Ramirez, V., 2009, Late local
glacial maximum in the Central Altiplano triggered by cold and locally-wet conditions
31
during the paleolake Tauca episode (17–15 ka, Heinrich 1): Quaternary Science Reviews, v.
28, p. 3414–3427, doi:10.1016/j.quascirev.2009.09.025.
Cai, Y., Cheng, H., An, Z., Edwards, R.L., Wang, X., Tan, L., and Wang, J., 2010, Large
variations of oxygen isotopes in precipitation over south-central Tibet during Marine Isotope
Stage 5: Geology, v. 38, p. 243–246, doi:10.1130/G30306.1.
Chen, F., et al., 2008, Holocene moisture evolution in arid central Asia and its out-of-phase
relationship with Asian monsoon history: Quaternary Science Reviews, v. 27, p. 351–364,
doi:10.1016/j.quascirev.2007.10.017.
Conroy, J.L., and Overpeck, J.T., 2011, Regionalization of present-day precipitation in the
greater monsoon region of Asia: Journal of Climate, v. 24, p. 4073–4095,
doi:10.1175/2011JCLI4033.1.
Dayem, K.E., Molnar, P., Battisti, D.S., and Roe, G.H., 2010, Lessons learned from oxygen
isotopes in modern precipitation applied to interpretation of speleothem records of
paleoclimate from eastern Asia: Earth and Planetary Science Letters, v. 295, p. 219–230,
doi:10.1016/j.epsl.2010.04.003.
Demske, D., Tarasov, P.E., Wünneman, B., and Riedel, F., 2009, Late glacial and Holocene
vegetation, Indian monsoon and westerly circulation in the Trans-Himalaya recorded in the
lacustrine pollen sequence from Tso Kar, Ladakh, NW India: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 279, p. 172–185, doi:10.1016/j.palaeo.2009.05.008.
Gasse, F., Fontes, J.C., Van Campo, E., and Wei, K., 1996, Holocene environmental changes in
Banggong Co basin (western Tibet). 4. Discussion and conclusions: Palaeogeography,
Palaeoclimatology, Palaeoecology, v. 120, p. 79–92, doi:10.1016/0031-0182(95)00035-6.
32
Gupta, A., Anderson, D.M., and Overpeck, J.T., 2003, Abrupt changes in the Asian southwest
monsoon during the Holocene and their links to the North Atlantic Ocean: Nature, v. 421, p.
354–357, doi:10.1038/nature01340.
Herzschuh, U., Winter, K., Wünneman, B., and Li, S., 2006, A general cooling trend on the
central Tibetan Plateau throughout the Holocene recorded by the Lake Zigetang pollen
spectra:
Quaternary
International,
v.
154–155,
p.
113–121,
doi:10.1016/j.quaint.2006.02.005.
Hu, C., Henderson, G.M., Huang, J., Xie, S., Sun, Y., and Johnson, K.R., 2008, Quantification of
Holocene Asian monsoon rainfall from spatially separated cave records: Earth and Planetary
Science Letters, v. 266, p. 221–232, doi:10.1016/j.epsl.2007.10.015.
Kong, P., Na, C., Fink, D., Huang, F., and Ding, L., 2007, Cosmogenic 10Be inferred lake-level
changes in Sumxi Co basin, western Tibet: Journal of Asian Earth Sciences, v. 29, p. 698–
703, doi:10.1016/j.jseaes.2006.04.010.
Kramer, A., Herzschuh, U., Mischke, S., and Zhang, C., 2010, Late glacial vegetation and
climate oscillations on the southeastern Tibetan Plateau inferred from the Lake Naleng
pollen profile: Quaternary Research, v. 73, p. 324–335, doi:10.1016/j.yqres.2009.12.003.
Kummerow, C., Barnes, W., Kozu, T., Shiue, J., and Simpson, J., 1998, The Tropical Rainfall
Measuring Mission (TRMM) sensor package: Journal of Atmospheric and Oceanic
Technology, v. 15, p. 809–817, doi:10.1175/15200426(1998)015<0809:TTRMMT>2.0.CO;2.
Lee, J., Li, S., and Aitchison, J.C., 2009, OSL dating of paleoshorelines at Lagkor Tso, western
Tibet: Quaternary Geochronology, v. 4, p. 335–343, doi:10.1016/j.quageo.2009.02.003.
33
Li, L., Yang, S., Wang, Z., Zhu, X., and Tang, H., 2010, Evidence of warming and wetting
climate over the Qinghai-Tibet Plateau: Arctic, Antarctic, and Alpine Research, v. 42, p.
449–457, doi:10.1657/1938-4246-42.4.449.
Li, Q., Lu, H., Zhu, L., Wu, N., Wang, J., and Lu, X., 2011, Pollen-inferred climate changes and
vertical shifts of alpine vegetation belts on the northern slope of the Nyainqentanglha
Mountains (central Tibetan Plateau) since 8.4 kyr BP: The Holocene, v. 21, p. 939–950,
doi:10.1177/0959683611400218.
Maher, B.A., 2008, Holocene variability of the East Asian summer monsoon from Chinese cave
records: A re-assessment: The Holocene, v. 18, p. 861–866,
doi:10.1177/0959683608095569.
Matsubara, Y., and Howard, A.L., 2009, A spatially explicit model of runoff, evaporation, and
lake extent: application to modern and late Pleistocene lakes in the Great Basin region,
western United States: Water Resources Research, v. 45, p. W06425,
doi:10.1029/2007WR005953.
Morrill, C., Overpeck, J.T., Cole, J.E., Liu, K.-B., Shen, C.M., and Tang, L.Y., 2006, Holocene
variations in the Asian monsoon inferred from the geochemistry of lake sediments in central
Tibet: Quaternary Research, v. 65, p. 232–243, doi:10.1016/j.yqres.2005.02.014.
Mügler, I., et al., 2010, A multi-proxy approach to reconstruct hydrological changes and
Holocene climate development of Nam Co, central Tibet: Journal of Paleolimnology, v. 43,
p. 625–648, doi:10.1007/s10933-009-9357-0.
Pausata, F.S.R., Battisti, D.S., Nisancioglu, K.H., and Bitz, C.M., 2011, Chinese stalagmite δ18O
controlled by changes in the Indian monsoon during a simulated Heinrich event: Nature
Geoscience, v. 4, p. 474–480, doi:10.1038/ngeo1169.
34
Placzek, C., Quade, J., and Patchett, P.J., 2006, Geochronology and stratigraphy of late
Pleistocene lake cycles on the southern Bolivian Altiplano: Implications for causes of
tropical climate change: Geological Society of America Bulletin, v. 118, p. 515–532,
doi:10.1130/B25770.1.
Rupper, S., and Roe, G., 2008, Glacier changes and regional climate: A mass and energy balance
approach: Journal of Climate, v. 21, p. 5384–5401, doi:10.1175/2008JCLI2219.1.
Seki, O., Meyers, P.A., Yamamoto, S., Kawamura, K., Nakatsuka, T., Zhou, W., and Zheng, Y.,
2011, Pant-wax hydrogen isotopic evidence for postglacial variations in delivery of
precipitation in the monsoon domain of China: Geology, v. 39, no., p. 875–878.
Shen, J., Liu, X., Wang, S., and Ryo, M., 2005, Palaeoclimatic changes in the Qinghai Lake area
during the last 18,000 years: Quaternary International, v. 136, p. 131–140,
doi:10.1016/j.quaint.2004.11.014.
Van Campo, E., and Gasse, F., 1993, Pollen- and diatom-inferred climatic and hydrological
changes in Sumxi Co basin (western Tibet) since 13,000 yr B.P: Quaternary Research, v. 39,
p. 300–313, doi:10.1006/qres.1993.1037.
Wang, Y., Cheng, H., Edwards, R.L., Kong, X., Shao, X., Chen, S., Wu, J., Jiang, X., Wang, X.,
and An, Z., 2008, Millenial- and orbital-scale changes in the East Asian monsoon over the
past 224,000 years: Nature, v. 451, p. 1090–1093, doi:10.1038/nature06692.
Wang, Y., Liu, X., and Herzschuh, U., 2010, Asynchronous evolution of the Indian and East
Asian Summer Monsoon indicated by Holocene moisture patterns in monsoonal central
Asia: Earth-Science Reviews, v. 103, p. 135–153, doi:10.1016/j.earscirev.2010.09.004.
Wischnewski, J., Mischke, S., Wang, Y., and Herzschuh, U., 2011, Reconstructing climate
variability on the northeastern Tibetan Plateau since the last Lateglacial—A multi-proxy,
35
dual-site approach comparing terrestrial and aquatic signals: Quaternary Science Reviews, v.
30, p. 82–97, doi:10.1016/j.quascirev.2010.10.001.
1.8 SUPPLEMENTARY METHODS
Inorganic CaCO3 tufa used for dating in this study is found in discrete laterally
continuous sub-horizontal bands associated with most shoreline ridges in both the Ngangla Ring
Tso and Drebyer Tsaka basins (Fig. A1). Full thickness samples were taken from bedrock
encrustations at two discrete sampling localities in each basin. In the Ngangla Ring Tso basin,
the six radiocarbon dates constraining the age of the paleoake highstand come from two separate
locations on opposite sides of the paleolake. The Drebyer Tsaka samples similarly come from
two locations on the high shoreline on the southwestern side of the salt pan (Fig. A4).
Tufas are generally formed at or near closed-basin lake shorelines over multiple
generations that can have large differences in age, so subsampling is necessary to correctly
determine the age of individual phases of tufa growth. In the laboratory each tufa sample was
cut into cross-section using a lapidary saw to expose the internal stratigraphy. Typically two or
three distinct generations of tufa growth ranging between 3-50 mm in thickness were observed in
each sample divided by a clear stratigraphic break (Fig. A5). Each stratigraphic unit was
subsampled for dating.
36
All tufa subsamples were rinsed ultrasonically after sampling to remove any detritus and
reacted in 2% H2O2 for 3 hours to remove any organic material. Samples were then rinsed
ultrasonically again and dried overnight in a 70°C drying oven.
A chunk of approximately 15 mg of each sample was placed in a glass y-tube with 2-3
mL of 100% H3PO4 and evacuated to <2x10-5 torr pressure. The sample was then combined with
the acid and dissolved until the reaction was visibly complete. Sample CO2 gas was extracted
under vacuum and cryogenically purified, passed through a 600°C Cu/Ag furnace to further
remove contaminant gases, frozen in liquid nitrogen (LN), split into aliquots for δ13C and AMS
measurement and sealed in glass tubes. Purified CO2 samples were then graphitized using 100
mg of zinc powder and Fe powder in a 2:1 proportion to the mass of carbon in the sample.
AMS and δ13C measurements were performed by the Arizona Accelerator Mass
Spectrometer Facility. Raw radiocarbon ages (14C yrs BP) were converted to calendar ages (cal
yrs BP where Present = AD 1950) using the Calib 6.0 software using the IntCal09 radiocarbon
calibration curve (Reimer et al., 2009).
Tufa subsamples used for U-Th dating were slabbed, sampled based on stratigraphy, and
pretreated identically to radiocarbon samples. U-Th dating was performed at the University of
Minnesota. Sample powders were drilled using a carbide-tipped drill bit from the same
stratigraphic tufa horizon from which radiocarbon samples were taken and prepared following
the methods of Cheng et al. (2000a, 2009) and Shen et al. (2002). Samples were analyzed using
a Neptune multi-collector ICP-MS. Uncertainties on concentrations are estimated as ±1% to
reflect uncertainties in spike concentration and weighing rather than only analytical uncertainties.
Ages were calculated using the
230
Th and
234
half-life of Jaffey et al. (1971). The detrital
U half-lives of Cheng et al. (2000b) and the
230
238
U
Th correction for age calculations assumes the
37
initial atomic ratio of 4.4(±2.2)x10-6. Those are values for a material at secular equilibrium with
the bulk earth 232Th/238U value of 3.8. The error in initial ratio is arbitrarily assumed to be 50%.
U-Th dates were finally corrected to AD 1950 to be consistent with the 14C dates used to assess
the reservoir effect (Table S2).
The watershed boundaries, basin-averaged precipitation, modern lake areas and paleolake
areas were measured using ESRI ArcGIS 10 software. All calculations used an equal-area conic
projection covering the Asian continent. Modern basin boundaries were digitized manually
using the 3-arc second (approximately 90m grid) Shuttle Radar Topography Mission Digital
Elevation Model (SRTM DEM) and Google Earth imagery for verification of drainage patterns.
Basin-averaged annual precipitation was determined using the mean annual TRMM (Tropical
Rainfall Measuring Mission) precipitation averaged over 1998-2009 for each 0.25°x0.25° grid
square intersecting a basin (Kummerow et al, 1997). The grid square values were then averaged
to derive a single basin-averaged estimate.
Due to a large increasing trend in modern lake areas evident since ca. 1998 identified in
most studied lake systems from sequential Landsat imagery, ‘modern’ lake areas used for
analysis were extracted from Landsat imagery taken during 1999-2001 using a maximumlikelihood image classification method. If late 2000’s images are used however, the resulting
paleolake expansion pattern is similar. Lakes within each basin with greater than 0.25 km2 area
were retained and summed to calculate the final lake area estimate (Table A3). Maximum
paleolake areas for each lake system were extracted by measuring the elevation of the highest
visible paleoshoreline with good preservation (visible at the same elevation on all sides of the
basin) in no less than three locations on the imagery. These shorelines could generally be traced
around >80% of the perimeter of the paleolake. The paleolake extent was determined by
38
extracting cells with equal or less elevation than the paleolake shoreline from the DEM similar to
the method used in the PaleolakeR tool developed by Sheng (2009). Modern lake areas not
subsumed by the highstand paleolake, including hydrologically open lakes flowing to the closed
terminal lake were added to the paleolake area for the final summed area estimates (Table A4).
The basin, ‘modern’ and highstand paleolake area data were analyzed to define the spatial
pattern of lake expansions on the interior of the Tibetan Plateau by calculating the lake area
ratios (Aw hereafter), the ratio of summed lake or paleolake areas to the total basin area for each
lake system. This includes the areas of all water bodies hydrologically open or closed within a
basin that are part of the hydrologic budget of the system.
This approach is commonly
implemented in hydrologic budget modeling (e.g. Matsubara and Howard, 2009) and effectively
normalizes lakes and basins of vastly differing size to a measure of the net hydrologic budget. It
also allows comparison between modern and paleolake systems with different basin
configurations, and has been shown to be a reliable indicator of hydrologic budget in complex
multi-lake systems (Benson et al., 1989). In a number of cases, the large paleolake areas
observed define paleolakes that spill into or subsume adjacent basins (Table A5). In this case,
the paleolake and paleobasin areas were summed for each composite system and compared
against a weighted average of the modern Aw’s for the basins included, weighted by basin area.
For all analysis, lake systems with basins less than 500 km2 were excluded to minimize the effect
of basin geometry on the lake area-to-volume relationship.
Figure A7 shows the location of Holocene lake records on the Tibetan Plateau cited in the
text, and the age range in which rainfall conditions inferred from proxy and/or shoreline
reconstructions were much greater than present. References and relevant information about each
record can be found in Table A6.
39
Benson, L.V., and Paillet, F.L., 1989, The use of total lake-surface area as an indicator of
climatic change: examples from the Lahontan Basin: Quaternary Research, v. 32, p. 262-275.
Cheng Hai., Adkins, J., Edwards, R.L., and Boyle, E.A., 2000a, U-Th dating of deep-sea corals:
Geochimica et Cosmochimica Acta, v. 64, p. 2401-2416.
Cheng Hai, Edwards, R.L., Hoff, J., Gallup, C.D., Richerds, D.A., and Asmerom, Y., 2000b, The
half-lives of uranium-234 and thorium-230: Chemical Geology, v. 169, p. 17-33.
Cheng Hai, Edwards, R.L., Broecker, W.S., Denton, G.H., Kong Xinggong, Wang Yongjin,
Zhang Rong, and Wang Xianfeng, 2009, Ice Age Terminations: Science, v. 326, p. 248-252.
Jaffey, A.H., Flynn, K.F., Glenden, L., Bentley, W.C., and Essling, A.M., 1971, Precision
measurement of half-lives and specific activities of 235U and 238U: Phys Rev C, v.4, p. 18891906.
Shen Chuanchou, Edwards, R.L., Cheng Hai, Dorale, J.A., Thomas, R.B., Moran, S.B.,
Weinstein, S.E., and Edmonds, H.N., 2002, Uranium and thorium isotopic and concentration
measurements by magnetic sector inductively coupled plasma mass spectrometry: Chemical
Geology, v. 185, p. 165-178.
40
Sheng Yongwei, 2009, PaleoLakeR: a semiautomated tool for regional-scale paleolake recovery
using geospatial information technologies: IEEE Geoscience and Remote Sensing Letters, v.
6, no. 4, p. 797-801.
Wunneman, B., and 10 others, 2010, Hydrological evolution during the last 15 kyr in the Tso
Kar lake basin (Ladakh, India), derived from geomorphological, sedimentological, and
palynological records: Quaternary Science Reviews, v. 29, p. 1138-1155.
Yao Tandong, Ren Jiawen, Xu Baiqing, Mi Desheng, Yang Mengmei, Ma Jingming, and Xu
Lanzhou, 2008, Map of glaciers and lakes on the Qinghai-Xizang (Tibet) Plateau and
adjoining regions: Xi’an Cartographic Publishing House map GS(2007)2117, scale 1:2 000
000, 1 sheet.
41
TABLE A1. RADIOCARBON RESULTS
Sample Name
Laboratory ID
Lake Basin
Shoreline elevation
(m asl)
nrc 10-92a
nrc 10-92b top
nrc 10-92b bottom
nrc 10-4-1 top
nrc 10-4-1 bottom
nrc 10-44-1
nrc 10-44-2
Kailas 21a†
Kailas 21c
dt 10-1a
dt 10-8a
AA-91437
AA-91439
AA-91434
AA-96571
AA-96570
AA-96572
AA-96573
AA-82149
AA-82150
AA-98073
AA-99058
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Ngangla Ring Tso
Drebyer Tsaka
Drebyer Tsaka
4862
4862
4862
4862
4862
4864
4864
4823
4823
4585
4599
C age
Uncertainty
2-sigma calibrated range*
Median age
(14C yrs BP)
8910
8110
8370
7840
8450
8540
8760
7100
7340
8830
9780
(14C yrs BP)
30
30
30
20
40
30
30
40
40
40
50
(calendar yrs BP)
(calendar yrs BP)
9,920-10,176
8,986-9134
9,306-9470
8,577-8,646
9,422-9,532
9,479-9,549
9,626-9,903
7,848-8,000
8,022-8,299
9,703-10,152
11,121-11,266
10,048
9,060
9,388
8,612
9,477
9,514
9,765
7,924
8,160
9,928
11,193
14
*Calibration was performed using the Calib 6.0 software and the IntCal09 calibration curve (Reimer et al., 2009), median age is the arithmetic median of the 2sigma range
†
Samples shown in bold were used for determining the 14C reservoir effect for tufa in the Ngangla Ring Tso lake system
41
42
TABLE A2. U-Th SERIES RESULTS
238
232
Sample
Number
U
(ppb)
Th
(ppb)
Kailas 21a
Kailas 21c
nrc 10-93-2
19255.9 ±89.0
15696.7 ±80.9
2908.0 ±4.5
328.9 ±6.6
279.5 ±5.6
20.9 ±0.4
230
Th / 232Th
(activity)
19.80 ±0.69
18.59 ±0.65
476.89 ±16.60
234
U/238U
(activity)
1.514 ±0.003
1.502 ±0.003
1.246 ±0.002
230
Th / 238U
(activity)
0.1104 ±0.0006
0.1081 ±0.0006
1.1183 ±0.0022
230
Th Age (yr)
(uncorrected)
8224 ±51
8112 ±52
211396 ±1548
230
Th Age (yr)*
(corrected)
7898 ±236
7769 ±248
211251 ±1549
230
Th Age (yr BP)
(Before AD 1950)
7839 ±236
7710 ±248
211191 ±1549
*Corrected 230Th ages assume the initial 230Th/232Th atomic ratio of 4.4±2.2 x10-6. Those are the values for a material at secular equilibrium, with the bulk earth
232
Th/238U value of 3.8. The errors are arbitrarily assumed to be 50%.
43
44
Nonglaciated Tso Nag
Headwaters Qagong Co
Unnamed
Baqan Co
Tuoheping Lake
Kyahu Co
Wanquan Lake
Larbu Co
Lagkor Co
Chagbo Co
Xin Lake
Rena Co
Domar Co
Unnamed
Wuming Lake
Nagding Co/Nanbei Co
Ningri Co
Gyado Co
Unnamed
Gom Caka
Bezi Co
Rigain Punco
Marye Co
BunSum Co
Unnamed
Chaoyang Lake
Deyu Co
Qoiden Co
Xiangyang Lake
Pibi Lake
Xo Caka
Haobo Co/Yingwu Co
Pongyin Co
Longwei Co
Bandao Lake
Pongce Co/Naiqam Co
Belog Co
Geigain Co
Dongyue Lake
Zainzong Co
Pangkog Co
Angdar Co
Lake Name†
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Lake ID
84.46
84.70
85.33
85.45
85.58
85.61
85.77
85.81
86.17
86.24
86.34
86.43
87.21
87.26
87.26
87.49
87.55
87.70
87.78
88.03
88.26
88.31
88.44
88.69
88.84
89.19
89.21
89.48
89.51
89.58
32.97
34.53
33.95
32.67
33.32
34.04
32.45
33.66
33.03
32.58
33.53
33.21
31.91
35.29
35.69
34.36
34.00
34.33
33.06
34.41
32.92
33.87
34.16
32.32
32.90
32.40
34.38
32.19
31.75
32.71
(°E)
82.33
82.34
82.61
82.78
82.95
82.97
83.68
83.81
84.13
84.12
84.25
84.26
(°N)
31.63
34.43
34.03
31.93
33.98
33.39
34.08
32.95
32.03
33.33
34.39
32.73
4790
5095
4482
4964
4947
4772
4871
4547
4467
4513
4813
4594
4675
5058
4813
4869
5035
4880
4595
4678
4841
4672
4860
4915
4492
4744
4852
4883
5011
4894
4800
4830
4727
4946
4917
4522
4818
4667
4846
4516
4527
4839
(m asl)
5124
5255
5358
5404
5404
5431
5456
5512
5518
5555
5569
5576
5577
5582
5589
5590
5620
5621
5627
5636
5638
5645
5647
5648
5653
5659
5693
5695
5728
5763
5772
5778
5781
5786
5819
5823
5826
5828
5862
5865
5870
5870
(m asl)
Terminal Lake Peak Basin
Elevation
Elevation
TABLE A3. MODERN LAKES
Centroid
Longitude
Centroid
Latitude
5178
5311
5095
5237
5240
5150
5135
4790
4871
4943
5165
4832
4832
5227
5081
5076
5275
5095
4993
5020
5091
4963
5064
5250
4754
4958
5021
5091
5194
5102
5069
5022
5006
5159
5087
4864
5005
4872
5060
4639
4717
4992
(m asl)
Mean Basin
Elevation
225
125
108
239
126
147
138
156
245
161
116
200
200
136
152
241
209
171
298
202
332
251
228
217
284
156
148
214
217
199
245
209
295
241
235
301
268
282
341
312
313
272
(mm/yr)
Mean Annual§
Precipitation
(km2)
58.17
22.6
21.67
13.19
102.92
27.56
121.42
16.36
105.08
57.36
41.86
18.65
12.93
22.97
11.36
19.71
15.66
40.74
12.91
66.34
17.75
42.12
7.01
10.08
12.13
44.2
69.86
30.73
11.78
35.45
24.35
42.99
52.76
48.77
36.29
103.64
25.03
26.6
20.57
10.23
129.97
43.16
Lake Area
(km2)
1100.84
643.15
633.07
688.48
3495.20
2525.15
5345.51
1365.23
4193.48
6026.83
2225.22
508.39
531.88
1233.09
809.41
1231.98
903.33
1036.05
549.21
6729.68
1592.48
1390.24
467.68
956.75
697.11
3385.39
2109.41
1328.33
894.82
1651.12
1513.68
1227.27
1611.05
1085.39
1217.20
4645.50
773.31
1487.06
548.40
610.29
2466.30
1377.30
Basin Area
5.28
3.51
3.42
1.92
2.94
1.09
2.27
1.20
2.51
0.95
1.88
3.67
2.43
1.86
1.40
1.60
1.73
3.93
2.35
0.99
1.11
3.03
1.50
1.05
1.74
1.31
3.31
2.31
1.32
2.15
1.61
3.50
3.27
4.49
2.98
2.23
3.24
1.79
3.75
1.68
5.27
3.13
(%)
Aw
45
Glaciated
Headwaters
Yunbo Co/Garing Co
Dawa Co
Larung Co/Laxong Co
Zhari Namco
Unnamed
Bura Co
Larxang Co
Punze Co
Monco Bunnyi
Xum Co
Garkung Caka
Tangra Yumco
Yumgen Co/Gyaro Co/Gyungmo Co
Zhamcomarqong
Namka Co
Yaggain Canco
Kyebxang Co
Dungqag Co
Mingjing Lake
Bam Co
Zigetang Co
S. Salijlang Kol
Chem Co
Aksayqin Lake/Gozha Co
Sumxi Co/Longmu Co
Gyeze Caka
Bamgdog Co/Pur Co/Yaeya Lake
Lumachangdong Co
Qingche Lake/Luotuo Lake
Argog Co
Memar Co/Aru Co
Nyer Co/A'ong Co
N. Heishi Lake
Bero Zeco
Ngangla Ring Co
Jianshui Lake/Bairab Co
Gopug Co
Darab Co
Unnamed
Rinchen Shubtso
Ca Co
Palung Co
Yang Lake
Taro Co/Drebyer Tsaka
Dong Co/Zhaxi Co
Gyesar Co
Camar Co/Burgar Co
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
33.15
31.86
33.01
32.45
31.77
35.06
31.25
32.08
34.68
34.15
35.31
34.61
33.95
34.95
34.02
34.74
30.98
34.22
32.55
35.56
34.73
31.54
25.30
31.86
32.47
33.96
31.28
32.11
30.88
35.41
31.42
32.18
30.20
33.31
30.77
31.23
34.34
30.87
29.85
34.40
33.98
30.47
30.67
30.29
33.97
31.07
32.19
89.70
89.79
89.79
89.98
90.41
90.56
90.58
90.86
79.69
79.78
79.83
80.46
80.91
81.56
81.62
81.89
82.24
82.31
82.46
82.75
82.93
83.08
83.12
83.18
83.21
83.32
83.45
83.55
83.58
84.60
84.06
84.73
84.78
84.81
84.95
85.00
85.23
85.50
85.70
85.77
86.04
86.11
86.22
86.41
86.49
86.60
86.60
4898
4537
4872
4614
4620
4801
4560
4568
5187
4961
4844
5003
4525
4904
4812
5039
5116
4920
4381
5049
4402
4727
4905
4718
4433
4884
4760
4340
5101
4792
4567
4394
5198
4581
4645
4625
4862
4612
5145
5166
4971
4964
4684
4714
4909
4535
4472
5903
5931
5955
5973
5986
5992
5994
6005
6005
6006
6015
6030
6047
6048
6055
6055
6064
6066
6068
6070
6075
6077
6084
6086
6099
6111
6118
6122
6131
6132
6133
6138
6140
6144
6152
6162
6165
6188
6190
6201
6202
6209
6219
6225
6259
5878
5888
TABLE A3. MODERN LAKES (contd.)
4990
4676
5019
4926
4801
4967
4852
4742
5407
5485
5252
5354
5098
5254
5240
5255
5390
5202
4867
5381
5008
5201
5159
5066
4808
5123
5294
5028
5439
5083
5071
4840
5558
5021
5178
5050
5164
5065
5434
5432
5191
5175
5068
5162
5129
5099
4888
258
287
274
304
361
255
370
350
132
165
135
136
190
122
145
132
264
132
184
103
227
235
103
210
190
121
233
259
281
125
246
251
238
177
265
290
165
277
415
182
172
288
316
298
174
305
286
20.35
23.55
113.16
157.02
53.19
115.47
211.75
209.22
116.85
113.84
467.12
150.26
106.03
285.77
358.44
209.7
59.71
241.72
271.21
94.17
35.5
514.43
256.39
67.29
35.23
23.26
181.855
95.12
143.43
80.97
705.16
131.5
141.97
64.68
125.73
111.23
72.63
1073.78
107.05
88.74
22.83
31.99
143.59
206.8
47.95
823.09
41.29
603.00
1360.83
2958.30
2621.31
1396.49
2760.42
3772.68
3479.00
5162.61
1813.41
17460.60
3260.50
2689.34
6991.35
8123.88
2457.80
1203.86
2339.32
46885.50
1732.49
4318.00
11549.60
9739.63
2315.84
2869.00
1949.43
2432.46
2720.65
1609.66
9141.18
15493.50
8189.23
954.36
6343.41
3275.61
2538.32
2168.28
18407.10
761.87
648.19
1430.69
638.14
820.05
1933.03
1704.65
9004.42
2498.00
3.37
1.73
3.83
5.99
3.81
4.18
5.61
6.01
2.26
6.28
2.68
4.61
3.94
4.09
4.41
8.53
4.96
10.33
0.58
5.44
0.82
4.45
2.63
2.91
1.23
1.19
7.48
3.50
8.91
0.89
4.55
1.61
14.88
1.02
3.84
4.38
3.35
5.83
14.05
13.69
1.60
5.01
17.51
10.70
2.81
9.14
1.65
46
Mang Co
Unnamed
Yangparpen Co
131
132
133
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
33.19
33.91
35.74
31.57
32.98
35.12
33.69
33.86
34.55
32.35
31.56
30.97
31.90
32.78
32.39
33.16
32.00
35.45
32.90
33.85
33.50
32.98
34.58
31.79
33.17
35.32
35.80
36.33
33.64
33.43
35.75
35.21
34.83
30.74
31.69
31.51
35.59
31.71
33.89
31.30
35.33
34.50
34.80
32.24
86.67
86.69
86.69
86.74
86.74
86.75
86.84
87.02
87.18
87.31
87.33
87.47
87.54
87.82
88.05
88.11
88.22
88.39
88.44
88.60
88.70
88.70
88.96
88.99
89.04
89.24
89.42
89.44
89.72
90.08
90.19
90.34
90.43
90.60
90.74
90.97
91.14
91.16
91.19
91.47
91.86
80.44
81.27
89.61
4861
4822
4889
4460
4560
4801
4755
4836
4930
4977
4605
4685
4465
4758
4700
4760
4515
4821
4819
5069
4964
4825
4823
4539
4848
4792
4866
4717
4952
4936
4876
4773
4855
4724
4683
4529
4890
4551
4923
4529
4785
5018
5000
-6287
6290
6293
6296
6313
6328
6332
6335
6338
6345
6353
6356
6363
6378
6387
6410
6410
6422
6447
6449
6458
6501
6511
6511
6514
6517
6531
6544
6544
6546
6568
6583
6600
6632
6653
6697
6756
6907
7057
--
6259
6263
6271
6275
TABLE A3. MODERN LAKES (contd.)
234
185
152
298
249
163
272
216
195
316
299
342
329
352
304
266
301
196
291
298
242
268
270
353
270
218
248
164
284
303
231
222
263
397
324
403
250
387
332
373
291
----
5220
5026
5151
4948
5062
5069
5020
5058
5160
5221
4929
5044
4948
5100
5066
5067
4889
5020
5053
5280
5169
5095
5088
4934
5032
5012
5050
4941
5149
5161
5045
4930
5071
5024
4934
4829
5018
4805
5072
4716
4952
5315
5173
--
10.38
28.32
73.96
53.78
167.47
201.93
24.92
90.52
102.99
43.93
98.55
483.67
319.91
61.95
71.79
76.23
105.89
102.61
24.3
103.52
82.06
63.95
417.09
3816.04
72.72
214.08
87.09
250.04
67.55
1130.66
224.03
430.87
520.26
1967.58
58.54
295.28
444.77
144.13
82.51
75.2
96.12
----
§
†
Lake names were taken from the map by Yao et al. (2008)
Basin-averaged mean annual precipitation from TRMM data 1998-2009 (Kummerow et al., 1997)
*These lake basins are subsumed by a large composite paleolake, but were not used in the analysis of modern lakes, and are only included for comparison to those systems
Composite
Basins*
Gangtang Co
Suana Lake
Yurbao Co/Dapeng Lake
Tangqung Co
Yi'bug Caka
Tanshui Lake/Margai Caku
Raggyor Caka
Margog Caka
Co Nyi/Tawa Co
Unnamed
Gomang Co/Zhangne Co
Ngangzi Co/Marxai Co
Dogze Co/Tomgo Co
Baidoi Co
Norma Co
Kungkung Caka/Qagain Co
Gyarab Co/Serbug Co
Rola Tso/Tanshui Lake
Peli Co
Linggo Co
Darngo Co/Amur Co
Ngoinyar Coqung
Baitan Lake
Seling Co
Cedo Caka/Terang Puno
Dogaicoring Qangco
Xiangyang Lake
Jingyu Lake
Meriqancomari
Tug Co
Lixi'oidaim Co
Xijie Ulan Lake/Yonghong Co
Ulan Ul Lake
Nam Co
Daru Co
Pung Co/Npen Co
Yinmar Lake, Hoh Xu Lake
Dung Co/Kyiru Co
Qoimo Co
Neri Yunco
Codarima
558.58
1040.96
1353.50
899.86
6960.87
9481.25
972.76
2606.57
3881.80
775.85
1472.25
8953.81
10521.50
2207.33
1405.43
2773.98
3431.62
4579.32
690.34
1856.60
2656.45
1347.08
7709.24
50039.30
3054.09
6086.54
1852.36
4767.46
1598.50
13078.10
1958.62
5859.59
6493.55
10836.40
589.17
3398.20
3348.31
1296.50
675.66
1168.02
649.43
400.74
512.63
638.14
--
1.86
2.72
5.46
5.98
2.41
2.13
2.56
3.47
2.65
5.66
6.69
5.40
3.04
2.81
5.11
2.75
3.09
2.24
3.52
5.58
3.09
4.75
5.41
7.63
2.38
3.52
4.70
5.24
4.23
8.65
11.44
7.35
8.01
18.16
9.94
8.69
13.28
11.12
12.21
6.44
14.80
---
47
Unglaciated Qagong Co
Headwaters
Baqan Co
Tuoheping Lake
Kyahu Co
Wanquan Lake
Larbu Co
Chagbo Co
Xin Lake
Rena Co
Domar Co
Unnamed
Wuming Lake
Nagding Co/Nanbei Co
Gyado Co
Unnamed
Rigain Punco
BunSum Co
Deyu Co
Xiangyang Lake
Pibi Lake
Xo Caka
Haobo Co/Yingwu Co
Pongyin Co
Longwei Co
Bandao Lake
Geigain Co
Angdar Co
Yaggain Canco
Kyebxang Co
Dungqag Co
Zigetang Co
Glaciated S. Salijlang Kol
Headwaters
Chem Co
Aksayqin Lake/Gozha Co
Qingche Lake/Luotuo Lake
N. Heishi Lake
Jianshui Lake/Bairab Co
Gopug Co
Unnamed
Ca Co
Palung Co
Lake Name
2
4
5
6
7
8
10
11
12
13
14
15
16
18
19
22
24
27
29
30
31
32
33
34
35
38
42
45
46
47
50
51
52
53
58
61
63
64
66
68
69
Lake ID
(°E)
82.34
82.78
82.95
82.97
83.68
83.81
84.12
84.25
84.26
84.46
84.7
85.33
85.45
85.61
85.77
86.24
86.43
87.26
87.55
87.7
87.78
88.03
88.26
88.31
88.44
89.19
89.58
89.79
89.98
90.41
90.86
79.69
79.78
79.83
81.89
82.75
83.12
83.18
83.32
83.55
83.58
(°N)
34.43
31.93
33.98
33.39
34.08
32.95
33.33
34.39
32.73
32.97
34.53
33.95
32.67
34.04
32.45
32.58
33.21
35.69
34
34.33
33.06
34.41
32.92
33.87
34.16
32.4
32.71
33.01
32.45
31.77
32.08
34.68
34.15
35.31
34.74
35.56
35.3
31.86
33.96
32.11
30.88
Centroid
Longitude
Centroid
Latitude
5095
4964
4947
4772
4871
4547
4513
4813
4594
4675
5058
4813
4869
4880
4595
4672
4915
4852
5011
4894
4800
4830
4727
4946
4917
4667
4839
4872
4614
4620
4568
5187
4961
4844
5039
5049
4905
4718
4884
4340
5101
(m asl)
5120
5020
5000
4789
4891
4572
4553
4830
4626
4695
5065
4821
4890
4892
4628
4712
4925
4854
5020
4898
4806
4831
4739
4949
4920
4680
4848
4876
4626
4635
4588
5245
5210
4860
5105
5091
4920
4768
4910
4460
5228
(m asl)
High
Shoreline
Elevation
(km2)
22.6
14.09
102.92
27.56
121.42
16.36
57.36
41.86
18.65
12.93
22.97
11.36
19.71
40.74
12.91
42.12
10.08
69.86
11.78
35.45
24.35
42.99
52.76
48.77
36.29
26.6
43.16
113.16
157.02
53.19
209.22
116.85
113.84
467.12
209.7
94.17
256.39
67.29
23.26
95.12
143.43
Lake Area
TABLE A4. PALEOLAKES
Terminal
Lake
Elevation
(km2)
60.04
72
264.27
102.23
239.75
106.74
265.44
114.59
49.03
40.8
40.05
18.4
71.37
54.61
32.83
98.14
21.38
83.25
33.45
55.14
54.28
58.24
99.93
61.71
55.91
63.98
82.17
140.95
203.32
118.09
308.43
1007.12
503.27
1515.77
606.74
256.67
597.03
305.11
71.11
339.47
548.3
Highstand
Lake Area
(km2)
643.15
376.19
3495.2
2525.15
5345.51
1365.23
6026.83
2225.22
508.39
531.88
1233.09
809.41
1231.98
1036.05
549.21
1390.24
956.75
2109.41
894.82
1651.12
1513.68
1227.27
1611.05
1085.39
1217.2
1487.06
1377.3
2958.3
2621.31
1396.49
3479
5162.61
1813.41
17460.6
2457.8
1732.49
9739.63
2315.84
1949.43
2720.65
1609.66
3.51
3.75
2.94
1.09
2.27
1.20
0.95
1.88
3.67
2.43
1.86
1.40
1.60
3.93
2.35
3.03
1.05
3.31
1.32
2.15
1.61
3.50
3.27
4.49
2.98
1.79
3.13
3.83
5.99
3.81
6.01
2.26
6.28
2.68
8.53
5.44
2.63
2.91
1.19
3.50
8.91
(%)
Total Basin Aggregate
Area
Lake Aw
9.34
19.14
7.56
4.05
4.49
7.82
4.40
5.15
9.64
7.67
3.25
2.27
5.79
5.27
5.98
7.06
2.23
3.95
3.74
3.34
3.59
4.75
6.20
5.69
4.59
4.30
5.97
4.76
7.76
8.46
8.87
19.51
27.75
8.68
24.69
14.82
6.13
13.17
3.65
12.48
34.06
(%)
Highstand
Lake Aw
2.66
5.11
2.57
3.71
1.97
6.52
4.63
2.74
2.63
3.16
1.74
1.62
3.62
1.34
2.54
2.33
2.12
1.19
2.84
1.56
2.23
1.35
1.89
1.27
1.54
2.41
1.90
1.25
1.29
2.22
1.47
8.62
4.42
3.24
2.89
2.73
2.33
4.53
3.06
3.57
3.82
Highstand Aw/
Modern Aw
1.095
1.642
1.201
1.157
1.662
0.784
0.735
0.943
1.38
1.126
0.741
1.337
1.299
2.062
2.167
1.537
0.941
1.048
0.996
0.941
0.919
0.532
0.793
0.793
0.867
0.843
0.803
0.734
0.877
0.888
1.196
1.388
4.47
1.84
1.281
1.458
1.262
1.092
0.926
2.283
2.481
(°)
Mean
Slope
5826
5620
5695
5973
5589
5590
5870
5828
5569
5870
5659
5648
5358
5512
5645
5638
5653
5693
6328
6447
7057
6332
6907
6338
6546
6055
6544
6544
5931
5955
6133
6140
5994
5636
6048
6111
5555
5404
5778
5903
5076
(m asl)
Mean Basin
Elevation
48
Yang Lake
Dong Co/Zhaxi Co
Camar Co/Burgar Co
Larung Co/Laxong Co
Bura Co
Garkung Caka
Yumgen Co/Gyaro Co/Gyungmo Co
Gangtang Co
Suana Lake
Yurbao Co/Dapeng Lake
Yi'bug Caka
Tanshui Lake/Margai Caku
Co Nyi/Tawa Co
Unnamed
Baidoi Co
Norma Co
Kungkung Caka/Qagain Co
Gyarab Co/Serbug Co
Linggo Co
Darngo Co/Amur Co
Baitan Lake
Cedo Caka/Terang Puno
Xiangyang Lake
Jingyu Lake
Meriqancomari
Tug Co
Lixi'oidaim Co
Ulan Ul Lake
Nam Co
Daru Co
Yinmar Lake, Hoh Xu Lake
Neri Yunco
70
72
74
77
80
85
87
88
89
90
92
93
96
97
101
102
103
104
107
108
110
112
114
115
116
117
118
120
121
122
124
127
35.41
32.18
33.31
34.34
34.4
33.97
32.19
33.19
33.91
35.74
32.98
35.12
34.55
32.35
32.78
32.39
33.16
32
33.85
33.5
34.58
33.17
35.8
36.33
33.64
33.43
35.75
34.83
30.74
31.69
35.59
31.3
84.6
84.73
84.81
85.23
85.77
86.49
86.6
86.67
86.69
86.69
86.74
86.75
87.18
87.31
87.82
88.05
88.11
88.22
88.6
88.7
88.96
89.04
89.42
89.44
89.72
90.08
90.19
90.43
90.6
90.74
91.14
91.47
4792
4394
4581
4862
5166
4909
4472
4861
4822
4889
4560
4801
4930
4977
4758
4700
4760
4515
5069
4964
4823
4848
4866
4717
4952
4936
4876
4855
4724
4683
4890
4529
4803
4430
4617
4878
5239
4935
4492
4890
4840
4895
4565
4812
4934
5010
4780
4720
4771
4550
5081
4965
4826
4850
4870
4720
4956
4942
4882
4865
4756
4694
4892
4544
80.97
131.5
64.68
72.63
88.74
47.95
41.29
10.38
28.32
73.96
167.47
201.93
102.99
43.93
61.95
71.79
76.23
105.89
103.52
82.06
417.09
72.72
87.09
250.04
67.55
1130.66
224.03
520.26
1967.58
58.54
444.77
75.2
TABLE A4. PALEOLAKES (contd.)
261.95
633.52
277.73
144.3
156.72
131.07
112.96
31.18
51.58
93.54
225.2
485.92
273.85
94.6726
138.09
138.56
126.72
240.35
137.97
114.12
522.89
106.07
129.18
340.31
103.53
1237.22
298.33
706.21
2574.58
91.43
505.66
146.59
9141.18
8189.23
6343.41
2168.28
648.19
1704.65
2498
558.58
1040.96
1353.5
6960.87
9481.25
3881.8
775.85
2207.33
1405.43
2773.98
3431.62
1856.6
2656.45
7709.24
3054.09
1852.36
4767.46
1598.5
13078.1
1958.62
6493.55
10836.4
589.17
3348.31
1168.02
0.89
1.61
1.02
3.35
13.69
2.81
1.65
1.86
2.72
5.46
2.41
2.13
2.65
5.66
2.81
5.11
2.75
3.09
5.58
3.09
5.41
2.38
4.70
5.24
4.23
8.65
11.44
8.01
18.16
9.94
13.28
6.44
2.87
7.74
4.38
6.66
24.18
7.69
4.52
5.58
4.96
6.91
3.24
5.13
7.05
12.20
6.26
9.86
4.57
7.00
7.43
4.30
6.78
3.47
6.97
7.14
6.48
9.46
15.23
10.88
23.76
15.52
15.10
12.55
3.24
4.82
4.29
1.99
1.77
2.73
2.74
3.00
1.82
1.26
1.34
2.41
2.66
2.16
2.23
1.93
1.66
2.27
1.33
1.39
1.25
1.46
1.48
1.36
1.53
1.09
1.33
1.36
1.31
1.56
1.14
1.95
1.233
1.391
0.972
0.891
3.589
2.022
1.005
1.285
1.586
1.269
0.949
1.049
1.011
1.759
1.188
1.299
0.872
1.998
1.448
0.72
1.108
0.967
1.438
0.8
0.939
1.129
1.084
1.111
2.001
1.266
1.248
1.027
6188
6756
6271
6410
6501
6086
6122
6162
6055
6345
6259
6353
6422
6201
6099
6144
6047
6209
6259
6084
6458
6118
5786
6005
6219
6517
6068
5878
6363
5728
6015
5763
49
Lake ID
54
55
56
60
61
64
73
80
88
102
113
125
Lake Name
Sumxi Co/Longmu Co
Gyeze Caka
Bamgdog Co/Pur Co/Yaeya Lake
Memar Co/Aru Co
Nyer Co/A'ong Co
Ngangla Ring Co
Taro Co/Chabyer Caka
Zhari Namco
Tangra Yumco
Dogze Co/Tomgo Co
Seling Co
Pung Co
(°E)
80.46
80.91
81.56
82.31
82.46
83.08
84.06
85.50
86.60
87.54
88.99
90.97
(°N)
34.61
33.95
34.95
34.22
32.55
31.54
31.42
30.87
31.07
31.90
31.79
31.51
Centroid
Longitude
Centroid
Latitude
5003
4525
4904
4920
4381
4727
4567
4612
4535
4465
4539
4529
(m asl)
Terminal
Lake
Elevation
5160
4800
5030
4967
4453
4875
4605
4750
4755
4520
4600
4575
(m asl)
(km2)
150.26
106.03
285.77
241.72
271.21
514.43
705.16
1073.78
823.09
319.91
3816.04
295.28
Lake
Area
(km2)
760.78
1580.15
997.51
709.56
2500.57
3696.99
4057.90
4993.07
2649.23
956.75
10641.28
1129.12
Highstand
Lake Area
(km2)
3260.50
10813.22
7503.98
2972.39
54072.50
16286.76
23916.95
22527.34
11837.31
12690.86
68316.28
8467.38
Total Basin
Area
TABLE A5. COMPOSITE PALEOLAKES
High
Shoreline
Elevation
4.96
4.30
3.83
8.86
0.63
5.00
4.51
6.37
9.15
3.39
6.70
7.69
(%)
Aggregate
Lake Aw
23.33
14.61
13.29
23.87
4.62
22.70
16.97
22.16
22.38
7.54
15.58
13.33
(%)
Highstand
Lake Aw
4.70
3.40
3.47
2.69
7.31
4.54
3.76
3.48
2.44
2.22
2.33
1.73
Highstand Aw/
Modern Aw
2.303
3.312
1.700
1.685
1.511
2.177
2.963
2.298
3.692
1.457
1.702
1.324
(°)
Mean
Slope
5354
5205
5254
5179
4867
5227
5070
5076
5098
4935
4927
4836
(m asl)
Mean Basin
Elevation
131
57
132
3
63, 67
1, 69
9, 75, 77
78, 81, 85
86, 93
25, 100
36, 40, 41, 44, 101, 132
49, 127
Composite Lake ID's
50
Naleng Lake
Hongyuan Peat
Kuhai Lake
10.)
11.)
12.)
n/a
n/a
n/a
50
n/a
123
n/a
(°E)
32.8
35.4
31.1
32.1
31.6
30.7
34.6
31.5
32.0
31.4
33.5
102.5
99.2
99.8
90.9
92.0
90.6
80.5
83.1
84.1
84.1
79.8
78.0
(°N)
33.3
Longitude
Latitude
n/a
8.4-present
12.8±1.1-6.4±0.7
>8.4-5.5 cal ka
7.5-4.4 cal ka
14.8-13.0, 10.8-4.4
cal ka
10.0-8.5 cal ka
7.5-6.0 cal ka
10.0-7.5 cal ka
12.0-~1.5
>14.8-present
17.7-present
10.8-present
10.0-4.0
7.3-present
13.0-present
10.0-4.4
>7.3-5.4 cal ka
n/a
n/a
n/a
~10.0-present
cal ka
n/a
15.2-present
Length of Record†
10.4-9.6 cal ka
11.1-9.9 cal ka
11.1-9.9 cal ka
9.6-6.6 cal ka
8.5-7.0 cal ka
10.8-9.0 cal ka
Inferred Wettest
Period
18
O and
13
C, ostracods
18
O and
13
C, ostracods
and
13
13.)
!"#$%&"'(&)*
n/a
36.9
100.2
7.5-5.0 cal ka
16.0-present
18
O
n-alkane D
terrestrial pollen
terrestrial pollen
13
C,
C, % TOC, TOC 13C, C/N ratio
terrestrial pollen
OM and n-alkane D and 13C
% CaCO3, % Dolomite, carbonate
lake level, 10Be exposure dating
terrestrial pollen
grain size, TOC, major elements, bulk
bulk carbonate
% CaCO3, terrestrial pollen, diatoms,
lake level, tufa dating
lake level, tufa and OSL dating
lake level, tufa dating
bulk carbonate
% CaCO3, terrestrial pollen, diatoms,
paleolake sediments/shorelines
terrestrial pollen
Methods Used
terrestrial pollen, % TOC, TOC
carbonate mineralogy
†Record length is only shown for proxy records (mostly cores). Shoreline records do not form a continuous record, and so are not treated.
Ahung Tso
Nam Tso
7.)
Zigetang Tso
Sumxi-Longmu Tso
6.)
8.)
54
Ngangla Ring Tso
Lagkor Tso
Chabyer Tsaka (Zabuye)
3.)
4.)
5.)
9.)
64
9
73
Banggong Tso
2.)
n/a
Tso Kar
1.)
Lake ID
Lake Name†
Figure A7 #
TABLE A6. TIBETAN PLATEAU PROXY RECORDS
Shen et al., 2005
Seki et al., 2011
Wischnewski et al., 2011
Kramer et al., 2010
Herzschuh et al., 2006
Morrill et al., 2006
Mugler et al., 2010
Kong et al., 2007
Li et al., 2011
Van Campo and Gasse, 1993
this study
this study, Lee et al., 2009
this study
Gasse et al., 1996
Wunneman et al., 2011
Demske et al., 2009
References
38°N
78°E
80°E
82°E
84°E
36°N
Region 2
86°E
88°E
90°E
92°E
94°E
96°E
0
10
20
0
0
34°N
30
1
0
40
Region 3
32°N
50
0
3
4
30°N
2
Region 1
26°N
28°N
Lhasa
10
N
00
100 km
Figure A1. Closed-basin lakes on the Tibetan Plateau. Modern lake areas are shown in black,
paleolake highstands in light gray. Modern precipitation contours (mm/yr) are based on the
monthly Tropical Rainfall Measuring Mission (TRMM) data from A.D. 1998–2009 (Kummerow
et al., 1998). Empirical Orthogonal Function (EOF) regions are shown by thick black lines
(modified from Conroy and Overpeck, 2011). Numbers 1–4 refer to sites discussed in the text:
1—Sumxi-Longmu Tso; 2—Ngangla Ring Tso; 3—Lagkor Tso; 4—Chabyer Tsaka.
52
51
A
548 km2 = 34.1%
1609 km2
143 km2 = 8.9%
1609 km2
N
10 km
N
B
2 km
C
Holocene
N
0.5 km
Pre-Holocene
Figure A2. Paleolake shoreline imagery. A) Palung Tso (30.9°N, 83.6°E) Aw (lake area/total
basin area); modern and paleolake areas are shown in black and white, respectively. B) Holocene
shorelines at Baqan Tso (31.9°N, 82.8°E). C) Holocene and pre-Holocene shorelines of Chabyer
Tsaka (31.4°N, 84.1°E).
53
52
9.0
20.0
8.0
Lake Aw (%)
16.0
7.0
Paleolake Aw
Modern Aw
Unglaciated
Glaciated
6.0
12.0
8.0
r=0.413
4.0
r=0.378
5.0
0.0
50
4.0
150 250 350 450
Precipitation (mm/yr)
3.0
2.0
1.0
79
81
83
85
87
89
Lake Centroid Longitude (°E)
91
93
Figure A3. Paleolake Aw (lake area/total basin area) versus modern Aw ratios in Tibetan lakes.
Glaciated systems are shown by squares; unglaciated by black diamonds. Inset: Regression of
modern Aw versus Tropical Rainfall Measuring Mission (TRMM) mean annual precipitation
averaged for grid squares (0.25° × 0.25°) intersecting the basin. Regressions are significant at
the 99% level.
53
A
Kailas 21
NRC 10-92, 93
NRC 10-44
NRC 10-4
B
DT 10-1
DT 10-8
Figure A4. Sampling locations in the Ngangla Ring Tso and Drebyer Tsaka basins for shoreline
tufa samples. Paleolake high shoreline used for analysis shown in red. Google Earth image base.
Dating results are given in Tables A1 (14C) and A2 (U-Th series).
55
54
1 cm
Figure A5. Example tufa sample cross section showing two distinct periods of tufa growth
separated by the black line. Each stratigraphic horizon was subsampled prior to dating.
55
10.0
A
9.0
9.0
8.0
8.0
7.0
7.0
Paleolake Aw/Modern Aw
Paleo Aw/Modern Aw
10.0
6.0
5.0
4.0
3.0
6.0
5.0
4.0
3.0
2.0
2.0
1.0
1.0
0.0
4400
B
0.0
4600
4800
5000
5200
5400
4500
4700
10.0
4900
5100
5300
5500
5700
Mean Basin Elevation (m asl)
Paleolake Elevation (m asl)
C
9.0
Paleolake Aw /Modern Aw
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
0
1
2
3
4
5
Paleolake Slope (°)
Figure A6. Results of regression analysis of paleolake expansion magnitudes. Glaciated basins
are shown by white squares, unglaciated basins by black diamonds. Regression lines are shown
in black. A) Paleolake elevation, B) mean basin elevation (SRTM), C) mean slope of the region
covered by the paleolake highstand. Raw data are given in Tables A4 and A5.
57
56
78°E
80°E
38°N
76°E
82°E
86°E
36°N
1.) Tso Kar
8.5-7.0 cal ka
max lake depth
10.8-9.0 cal ka
max monsoon
88°E
90°E
92°E
94°E
6.) Sumxi-Longmu Tso
10.0-4.4 cal ka
max monsoon
12.8-6.4 ka
max lake depth
96°E
2.) Banggong Tso
9.6-6.6 cal ka
98°E
100°E
102°E
104°E
12.) Qinghai Lake
7.5-5.0 cal ka
12.) Kuhai Lake
7.5-6.0 cal ka
4.) Lagkor Tso
11.0-9.9 cal ka
5.2 ka regression
34°N
32°N
84°E
11.) Hongyuan Peat
10.0-9.0 cal ka
9.) Zigetang Tso
>7.5-4.4 cal ka
28°N
30°N
10.) Naleng Lake
10.8-4.0 cal ka
8.) Ahung Tso
>9.0-7.5 cal ka
3.) Ngangla Ring Tso
10.4-8.6 cal ka
5.) Chabyer Tsaka
11.1-9.9 cal ka
7.) Nam Tso
>8.4-5.5 cal ka
>7.3-5.4 cal ka
Figure A7. Map showing climate proxy records of monsoon rainfall strength during the Holocene,
and the time interval of inferred maximum wetness. References for each record are found in Table
A6.
58
57
APPENDIX B.
LAKE LEVEL RECONSTRUCTION FOR 12.8-2.3 KA OF THE NGANGLA RING TSO
CLOSED-BASIN LAKE SYSTEM, SOUTHWEST TIBETAN PLATEAU
58
LAKE LEVEL RECONSTRUCTION FOR 12.8-2.3 KA OF THE NGANGLA RING TSO
CLOSED-BASIN LAKE SYSTEM, SOUTHWEST TIBETAN PLATEAU
Adam M. Hudsona, Jay Quadea, Tyler E. Huthb, Guoliang Leic, Hai Chengd,e, R. Lawrence
Edwardse, John W. Olsenf, Hucai Zhangg
a
Department of Geosciences, University of Arizona, Tucson, Arizona, USA
Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah, USA.
c
College of Geographical Sciences, Fujian Normal University, Fujian, 350007, China
d
Institute of Environmental Change, Xi’an Jiaotong University, Xi’an, 710049, China
e
Department of Earth Sciences, University of Minnesota, Minneapolis, Minnesota, USA
f
School of Anthropology, University of Arizona, Tucson, Arizona, USA
g
College of Tourism and Geography, Yunnan Normal University, Kunming, 650500, China
b
Corresponding Author: Adam M. Hudson
Email: amhudson@email.arizona.edu
Phone: 001-520-626-1853
Keywords: Tibetan Plateau, Holocene, Indian Summer Monsoon, radiocarbon, U-Th series
dating, shoreline dating, paleolake
This manuscript was published in Quaternary Research in January 2015.
2.1 ABSTRACT
We present a shoreline-based, millennial-scale record of lake-level changes spanning
12.8-2.3 ka for a large closed-basin lake system on the southwestern Tibetan Plateau. Fifty-three
radiocarbon and eight U-Th series dates of tufa and beach cement provide age control on
paleoshorelines ringing the basin, supplemented by nineteen dates from shell and aquatic plant
material from natural exposures generally recording lake regressions. Our results show that
paleo-Ngangla Ring Tso exceeded modern lake level (4727 m asl) continuously between ~12.8
and 2.3 ka. The lake was at its highstand 135 m (4862 m asl) above the modern lake from 10.3
ka to 8.6 ka. This is similar to other closed-basin lakes in western Tibet, and coincides with peak
Northern Hemisphere summer insolation and peak Indian Summer Monsoon intensity. The lake
experienced a series of millennial-scale oscillations centered on 11.5, 10.8, 8.3, 5.9 and 3.6 ka,
consistent with weak monsoon events in proxy records of the Indian Summer Monsoon. It is
unclear whether these events were forced by North Atlantic or Indian Ocean conditions, but
based on the abrupt lake level regressions recorded for Ngangla Ring Tso, they resulted in
59
significant periodic reductions in rainfall over the western Tibetan Plateau throughout the
Holocene.
2.2 INTRODUCTION
The rivers draining the Tibetan Plateau provide water resources to a huge and densely
populated area of Asia and there is great interest in what effect warming climate will have on
rainfall amount and distribution in the region. The high topography of the plateau, 5,000 m asl
on average, is also thought to drive the intense Indian Summer Monsoon (ISM) rains through the
combination of land-sea heat contrast over India, and intense orographic uplift along the
Himalayas (Boos and Kuang, 2010). Since monsoon rainfall during the summer half-year (MayOctober) makes up 60%->80% of annual precipitation in areas under ISM influence (Bookhagen
et al., 2010; Conroy and Overpeck, 2011), monsoon behavior under warmer climate is
particularly interesting. Numerous paleoclimate records indicate Northern Hemisphere monsoon
rainfall intensified in the past coincident with higher Northern Hemisphere summer insolation,
most recently during the early Holocene maximum (Fleitmann et al., 2003; Dykoski et al., 2005;
Severinghaus et al., 2009; Cheng et al., 2012). However, the magnitude and spatial distribution
of the precipitation maximum in both timing of peak monsoon conditions, and rainfall amount, is
regionally variable.
Paleoclimate records of monsoon variation during the Holocene in Tibet utilize a variety
of proxies, including ice core isotopes and pollen (e.g. Thompson et al., 1997, Liu et al., 1998),
lake core proxies for depth, hydrologic budget, and terrestrial flora (e.g. Demske et al., 2009;
Gasse et al., 1996; Li et al., 2011; Lu et al., 2011; Mischke et al., 2008; Morrill et al., 2006;
Mügler et al., 2010; Shen et al., 2005; Van Campo and Gasse, 1993; Wischnewski et al., 2011),
studies of surficial lake sediment exposures, and shoreline dating (e.g. Kong et al., 2007; Lee et
al., 2009; Pan et al., 2012; Wünneman et al., 2010) (Fig. B1). These records show annual
rainfall, mostly summer monsoon-derived, and mean annual temperature increased as Northern
Hemisphere summer insolation peaked 10.0-11.0 ka. Almost all records suggest wetter-thanpresent conditions persisted until the mid-Holocene, but there is significant debate over whether
the timing was regionally homogeneous within the Tibetan Plateau (e.g., Mügler et al., 2010; Li
et al., 2011; Wischnewski et al., 2011). The western plateau, where monsoon rainfall increased
60
most during the Holocene (Hudson and Quade, 2013), has fewer paleoclimate records because of
the remote location. We present a new record of Late Glacial and Holocene precipitation change
based on shoreline dating in a large closed-basin lake system, Ngangla Ring Tso, that
supplements existing knowledge for this region. In addition, we present a companion shoreline
record and paleohydrologic model for nearby lake Baqan Tso that quantifies the
paleoprecipitation conditions required to create the large lakes of the early Holocene in the
southwestern Tibetan Plateau (Huth et al., 2015).
2.3 STUDY AREA
2.3.1 Lake Basin Hydrography and Geology
The Ngangla Ring Tso lake system, located in the southwestern Tibetan Plateau (Fig.
B1), is currently divided into three separate, internally-drained basins: Tso Nag (TN), Ngangla
Ring Tso (NRT), and Rinqen Shubtso (RS), covering areas of 1100 km2, 11,550 km2, and 2435
km2, respectively (Fig. B2, B3). The TN basin (to the west) and the RS basin (to the east) are
separated from the NRT basin by spilling elevations of 4825 m asl (+98 m above modern NRT)
and 4805 m asl (+78 m above modern NRT), respectively (Fig. 3). The modern lake areas for
TN, NRT, and RS defined by measurements from Landsat imagery are 58.2 km2, 514.4 km2, and
181.9 km2, respectively.
The watershed for each lake is mainly the north flank of the Gangdese Mountains, which
have peak elevations exceeding 6,000 m asl. Drainage from the north-south trending Lunggar
Range also feeds NRT and RS (Fig. B2). NRT and RS receive meltwater runoff from alpine
glaciers, while the TN basin has no glaciers today. The NRT basin has three perennial river
systems, the largest being the Amo Tsangpo that enters the lake from the west (Fig. B2). The
lake is also fed by springs from the north and east sides. RS basin has one major river entering
from the south end, and many small meltwater- and spring-derived streams from all sides. TN
has no perennial rivers, and appears to be mostly supported by groundwater and ephemerally
through drainages mostly on the west side (Fig. B2).
Bedrock consists of volcanic rocks of the Gangdese arc in the southern part of the basins,
gneisses forming the Lunggar Range (Kapp et al., 2008), and Paleozoic to Cretaceous strata of
61
the Lhasa Terrane underlying most of the basins (Styron et al., 2013). Sedimentary rocks are
dominantly siliciclastic. We observed no carbonates in the catchment area.
2.3.2 General Appearance of Lake Deposits
Paleolake deposits are naturally exposed across the NRT basin.
Sharp-crested
paleoshoreline benches form a continuous set of concentric outlines of former lake extent
traceable for >80% of their perimeter, from the highest bench at 4862 m asl down to the modern
lakes (Fig. B2, B5a). Drowned shoreline benches are visible below the modern water level.
Calcite and aragonite tufa commonly cements beach gravel in discrete, laterally continuous
bands associated with shoreline ridges in the NRT system (Fig. B5a). Thick tufa encrustations
are also common on bedrock outcrops cut by the shorelines. Discontinuous, weathered bands of
tufa also occur above the highest shoreline bench up to 4877 m asl. Spring-deposited tufa is
found in isolated mounds often arranged linearly parallel to fault traces, especially on the
northeastern shore of NRT (Fig. B5c). Fine-grained marls are preserved between shoreline
ridges and in areas of flat topography. Abundant, thick sections of paleolake sediments have
been exposed by streamcuts throughout the basin.
2.3.3 Local Climate
Basin-averaged annual rainfall estimated from Tropical Rainfall Measurement Mission
(TRMM) gridded precipitation data (0.25° grid) 1998-2010, indicates the lake catchments
receive ~230 mm/yr (Kummerow et al., 1998). This is significantly higher than the annual mean
of meteorological station data from nearby Gerze, of 171 mm/yr, consistent with decreasing
precipitation to the north in the rain shadow of the Himalaya (Figs. B1, B4). Mean annual
temperature near the modern NRT elevation is near 0°C, based on data from Gerze. Landsat
imagery shows the lakes develop ice cover in mid-November, and are fully melted by mid-May.
Precipitation at Gerze has a sharp summer peak in precipitation in July and August (Fig. B4).
2.3.4 Sampling Sites
We sampled several materials for dating: organic and shell samples from the modern
lakes and from natural exposures of pre-modern lake deposits, and tufa and cemented beach
gravels along shoreline transects. Modern aquatic plant material was collected near the shore of
62
NRT, and modern gastropod shells (Radix sp., Taft et al., 2012) were collected from TN. We
collected two shoreline transects: one on the north shore of RS, and one in the Amo Tsangpo
valley west of NRT (Fig. B2).
We measured and sampled three stratigraphic sections in
shoreline deposits on the northwest side of RS (NRC10-7b, NRC10-21, NRC10-25), and one to
the east of TN (NRC10-107). Two streamcut sections (NRC10-96, NRC10-105) are located
along banks of the Amo Tsangpo west of NRT, and two in the banks of Amo Tsangpo tributaries
(NRC10-124, NRC JQ-5). One additional section is located in a cut bank of the stream feeding
NRT from the east (NRC10-82) (Fig. B2).
2.4 METHODS
Sampling locations were determined using a high-precision Trimble GeoXH GPS unit to
ensure accurate elevation measurements. Individual tufa samples were collected in situ on
shoreline features from bedrock encrustations to ensure no reworking from higher shorelines.
All samples collected were densely laminated or branching-type tufa, or dense cement within
gravel beach deposits. Additional samples of aquatic plant remains and gastropod shells were
collected in situ from measured depths in stratigraphic sections.
2.4.1 Tufa Sampling and Preparation
Tufas generally form at or near closed-basin lake shorelines over multiple lake
inundations that may have large differences in age (e.g. Benson et al., 1995; McGee et al., 2012),
so subsampling is necessary to determine the age of individual phases of tufa growth. Each
sample was cross-sectioned using a lapidary saw to expose the internal stratigraphy. Each
sample typically presented two distinct generations of tufa growth, ranging 3-50 mm in
thickness, divided by a clear stratigraphic break (Fig. B5b). Individual stratigraphic units were
subsampled for dating. In most cases, two samples, from the bottom and top stratigraphic levels,
were sampled for each tufa.
2.4.2 Radiocarbon Dating
Sample preparation prior to AMS measurement was performed at an in-house
14
C
laboratory at University of Arizona. All tufa subsamples, cements, and shells were rinsed
63
ultrasonically to remove detritus and reacted in 3% H2O2 to remove organic material. Each
sample was reacted with 100% H3PO4 under vacuum until dissolved. Organic samples were
subjected to a standard acid-base-acid treatment, and combusted under vacuum with Ag and CuO
at 900°C. Sample gas was extracted under vacuum, cryogenically purified and passed through a
600°C Cu/Ag furnace to remove contaminant gases. Purified CO2 samples were graphitized
using 100 mg of Zn powder and Fe powder in a 2:1 proportion to the mass of carbon in the
sample.
AMS and δ13C measurements were performed by the Arizona AMS Laboratory.
Radiocarbon dates were calibrated using the Calib 6.0 software and the IntCal09 calibration
curve (Reimer et al., 2009).
2.4.3 U-Th series Dating
Tufas used for U-Th dating were sampled and pretreated identically to 14C samples. UTh dating was performed at the University of Minnesota. Sample powders were drilled using a
carbide-tipped drill bit from the same stratigraphic tufa horizon from which
14
C samples were
taken and prepared following the methods of Cheng et al. (2000a, 2009) and Shen et al. (2002).
Samples were analyzed using a Neptune multi-collector ICP-MS.
Uncertainties on
concentrations are estimated as ±1% to reflect uncertainties in spike concentration and weighing
rather than only analytical uncertainties. Ages were calculated using the
lives of Cheng et al. (2013) and the
238
230
Th and
234
U half-
U half-life of Jaffey et al. (1971). The detrital
230
Th
-6
correction for age calculations assumes the initial atomic ratio of 4.4(±2.2)x10 for a material at
secular equilibrium with the bulk earth
232
Th/238U value of 3.8. The error in initial ratio is
arbitrarily assumed to be 50%. U-Th dates were corrected to AD 1950 to be consistent with the
14
C dates used to assess the reservoir effect (Table B2).
2.5 RESULTS
2.5.1 14C and U-Th Dating Considerations
Tufa has been successfully dated in similar high altitude, arid conditions to NRT using
both 14C and U-Th dating methods (e.g. Placzek et al., 2006; Blard et al., 2011), but tufa dating
problems due to 14C-deficient lake waters (reservoir effect) and secondary contamination during
post-regression subaerial exposure are known to cause
14
C ages appear older and younger than
64
the depositional age, respectively (Lao and Benson et al., 1988; Zimmerman et al., 2012).
Assessment of the accuracy of tufa 14C ages for the NRT system is based on two methods. First,
we
14
C dated modern aquatic plant matter from NRT (NRC10-56-5), recent tufa from NRT
(NRC11-1-1a,b; NRC11-2-1; NRC11-3-1) and RS (NRC11-6-1), and modern Radix shell from
TN (NRC10-108-2), to test if they formed in equilibrium with the
14
C content of the modern
atmosphere (Table B1). Second, we dated ten pre-modern tufas using both U-Th series and 14C
to test for potential offsets in dates between the two methods through time (Table B1, B2). U-Th
dates are considered accurate only from samples with
230
Th/232Th activity ratios >10, indicating
they have manageable detrital thorium concentrations for accurate dating (Placzek et al., 2006).
NRT tufa can be either primary calcite or aragonite, with aragonite samples yielding more
precise ages due to much higher uranium concentrations (Table B2). Eight of the pre-modern UTh dates had 230Th/232Th ratios >13, which translates into age uncertainties ≤ 4%. Only two premodern dates (NRC10-92, NRC10-101-1) failed this criterion and both were eliminated from
consideration in constructing the lake-level history (Table B2).
For NRT, the
14
C content of the modern aquatic plant matter collected from shallow
water (NRC10-56-5) has a fraction modern carbon of 1.029 (fmc, where ‘modern’ is the
14
C
content of the atmosphere in AD 1950), indicating it formed in near equilibrium with the
atmosphere. Atmospheric fmc in 2010 is likely ~1.05 (Levin et al., 2013), giving an apparent
14
C reservoir effect of ~165
14
C yr. This indicates modern NRT lake water is likely close to
equilibrium with the atmosphere near shore where tufa is formed. Radix gastropods (Taft et al.,
2012) dated in this study were only found living in TN. Modern Radix shell collected from the
shoreline of TN was dated at 5600±40 14C yr BP (Table 1, NRC10-108-2). This large age offset
is in stark contrast to that of the NRT aquatic plants, but is expected given TN has no perennial
surface inflows today and appears to be largely groundwater-fed. All attempts to date tufa
(samples NRC11-1-1a, b, NRC11-2-1, NRC11-3-1, NRC11-6-1) found at the elevation of
modern NRT and RS yielded
14
C ages ranging from 1410±20 to 10,730±50
14
C yr BP (Table
B1). These ages may reflect a variable reservoir effect, or ages of tufa precipitated during some
previous lake cycle. Samples NRC11-1-1a and -1b, which were also dated using U-Th series to
ages of 21±54 yr and 135±268 yr, respectively, indicate that at least a local reservoir effect is
most likely (Table B2).
65
Of the ten samples dated using both
14
C and U-Th series, Kailas 21a, Kailas 21c, and
NRC 10-88-1 show good agreement between the calibrated radiocarbon ages and U-Th ages.
For Kailas 21a and NRC 10-88-1, they overlap within error, and for Kailas 21c, the
14
C age is
~200 cal yr older (Table B1, B2).
The remaining seven samples do not show age agreement, and fall into two groups.
Kailas 20a, 20d, NRC11-1-1a, NRC11-1-1b, and NRC10-92b have U-Th series ages that appear
younger than the 14C ages, which is to be expected for a reservoir effect. The variable results for
the demonstrably recent tufas support this evidence (Table B1). Major element measurements
for NRT tufa indicate that the salinity of NRT lake water increased abruptly below the
connection elevation with RS (~4805 m asl) (Lei et al., 2013). All of the samples with older 14C
ages come from below this elevation with the exception of NRC10-92b (Table B1). The
increased salinity likely reflects lower river discharge, and higher proportion of spring water
discharge to the lake, which may explain the increased reservoir effect in some samples at low
lake levels. The age discrepancy for NRC10-92b can be explained by the poor precision of the
U-Th series date for calcite tufa, evidenced by the low
230
Th/232Th activity ratio and low
238
U
concentration (Table B2). Therefore, we exclude it from consideration of the reservoir effect.
The remaining samples (NRC11-1-1a, b, Kailas 20a, 20d), along with samples NRC11-2-1 and
NRC11-3-1 were all sampled near springs, or in the case of Kailas 20a and 20d, what we
interpret as sub-aqueous paleo-spring vents (Fig. B2, B4c).
The variable reservoir effects
indicated in these samples may therefore reflect local water conditions around the spring vent
and are excluded from consideration of the lake-wide shoreline tufa reservoir effect. Although
the tufa collected from the modern shoreline of RS (NRC11-6-1) is not near a known spring
discharge, the old apparent age indicates an increased proportion of spring discharge at low lake
level may be a factor for RS, also.
The second group of two samples, NRC10-87 bottom and NRC10-101, have
14
C ages
that appear younger than the U-Th ages (Table B1, B2). This either reflects difficulty in
consistent stratigraphic sampling, or secondary contamination of tufa by atmospheric carbon.
The age offset for NRC10-87 is around ~1000 yr, for the lowermost stratigraphic level of the
sample (Table 2, NRC10-87 bottom). The uppermost tufa layer is dated by 14C to 6660-6800 cal
yr BP, so the lower sample could incorporate a fraction of younger tufa. Conversely, both top
and bottom dates may reflect secondary contamination of the tufa by atmospheric carbon during
66
post-regression exposure (Zimmerman et al., 2012). The same is likely true of NRC10-101,
where multiple radiocarbon dates have yielded inconsistent results (Table B1). The U-Th series
age for NRC 10-101 also has considerable uncertainty due to high detrital thorium (Table B2).
All of the
14
C dates obtained for these two samples are excluded from defining the reservoir
effect and the lake level curve.
Based on the dating results for modern aquatic plant material, we have corrected all 14C
dates by 165
14
C yr before calibration, but employing no correction does not change the
interpretation of the lake level record appreciably. Because attempts to date recent tufa have
yielded extremely variable results, tufa ages without corroborating ages from the same elevation
elsewhere in the basin, or from other materials such as Radix shells or aquatic plants, must be
treated as maximum-limiting age constraints. Given the variable reservoir effects, it is also very
unlikely that a single reservoir age offset would apply to multiple locations and in multiple dated
materials in the lake system. Therefore, we place greater confidence in the accuracy of the lake
level reconstruction where multiple locations and materials show good age agreement.
2.5.2 Paleoshoreline Dating Results
Seventeen paleoshoreline tufas and five beach gravel cements were dated using 14C and
U-Th dating. The thirty-five
14
C dates range 4731-4864 m asl in elevation and 5135±60-
9745±30 14C yr BP in age (Table B1). Eight paleoshoreline U-Th series dates cover 4752-4877
m asl and 4868±84-211,191±1549 yr BP (Table B2). Of these, six have 14C dates on the same
sample. NRC10-90 (U-Th) is not comparable to NRC10-90 top (14C) because the subsamples
come from different stratigraphic horizons. NRC 10-93-1 has no 14C companion because the UTh age at 211,191±1549 cal yr (Table B2) is beyond the range of
14
C dating. The calculated
apparent 14C age for NRC10-93-1 is 28,210±460 14C yr BP (Table B1).
2.5.3 Dating Results from Sedimentary Sections
We have also described nine stratigraphic sections and generated nineteen
14
C dates to
supplement the shoreline chronology (Table B1, Fig. B2, B6). Radix shells were dated in four
shoreline sections in the RS and TN basins. Aquatic plant fragments, likely Potamogeton sp.
(Morrill et al., 2006), were dated in five sections in the NRT basin. We have high confidence in
67
the accuracy of aquatic plant dates due to the FMC of 1.029 for the modern material and no
observed stratigraphic inversions within the sections (Table B1, Fig. B6).
Shoreline Sections
Shoreline sections NRC10-25, NRC10-21 and NRC10-7b, northwest of RS are located at
4,819 m asl, 4,788 m asl, and 4,784 m asl, respectively. NRC10-25 and NRC10-21 are located
in the same drainage, where both sections preserve two clear lake cycles in massive silt and sand
intervals (Fig. B6, B7a). The lower lake interval in both is overlain by alluvial/colluvial gravel.
The top of the gravel horizons are then unconformably overlain by a second set of yellow lake
silts and sands, which are finally capped by well-sorted, well-rounded beach gravels and backbar
silts of the final lake regression (Fig. B6, B7a).
The
14
C dates in both sections support the interpretation of two large-scale lake cycles.
Lake sediments yielding infinite ages >41 14C ka BP (Fig. B6, NRC10-21-1, 95 cm, NRC10-251, 110 cm, and 25-2, 190 cm) pertain to the older cycle. Shells from the younger cycle record
the most recent lake regression. Among the younger samples, the higher elevation sample
NRC10-25-3 (285 cm) is dated to 6220-6400 cal yr BP, slightly older than NRC10-21-2 (140
cm) and NRC10-21-3 (250 cm) at 6000-6200 cal yr BP and 5910-6170 cal yr BP, respectively
(Table B1, Fig. B6).
NRC10-7b is located in a large drainage on the northwest side of RS. The sedimentary
sequence is similar to the younger lake phase of NRC10-21 and NRC10-25, but truncated by
erosion at the top (Fig. B6). Two
14
C dates on shells sampled from the base of the lacustrine
sequence, and from the directly below the truncated surface are 6320-6530 cal yr BP (NRC107b-1, 180 cm) and 6300-6440 cal yr BP (NRC10-7b-2, 390 cm), respectively (Table B1).
Section NRC10-107-1 is located at 4831 m asl in the northwestern study area, east of
modern TN, in a modern drainage with a small, steep catchment (Fig. B2).
The sediment
sequence consists of initial shallow lake conditions characterized by cross-bedded coarse sand
separated by two alluvial gravel horizons. This is overlain unconformably by massive sand with
an erosional lower contact containing the dated shell horizon. A thin horizon of gravel was
deposited unconformably above the lake sediments, followed by a thick layer of coarse, planarbedded sand and topped by a second gravel. A single Radix shell 14C date of 9150-9480 cal yr
68
BP (NRC10-107-1, 195 cm) was generated for the section (Fig. B6). The uppermost units are
interpreted as post-lake alluvial fill.
Streamcut Sections
NRC JQ-5 (4806 m asl), NRC10-124 (4819 m asl), 105 (4757 m asl), 82 (4754 m asl),
and 96 (4731 m asl) are all below the 4825 m asl threshold elevation for the integrated paleolake
(Fig. B3, B6). With the exception of NRC10-82, they are all located along the Amo Tsangpo
and its tributaries west of NRT. NRC10-82 is located along the drainage at the eastern end of
NRT (Fig. B2). The depositional sequence of all sections culminates with lake regression,
characterized by the transition from deep water, offshore clays, to calcareous lacustrine silts, to
silt with abundant Potamogeton remains signifying shallower, near-shore lake depths, capped by
planar and cross-bedded deltaic sands. With the exception of NRC10-105, all sections have river
or beach gravels overlain by a cap of eolian sand or loess. NRC10-105 follows this pattern, but
has a second interval of silt above the deltaic deposits, which we interpret as a minor lake
transgression. Eolian silt and sand caps this lacustrine silt (Fig. B6, B7b).
In elevation from highest to lowest, the section radiocarbon results (Table B1) are as
follows: NRC10-124 and NRC JQ-5 contain both the transgressive and regressive phases of the
lake cycle. The base of NRC10-124 has laminated sand and silt of deltaic origin, overlain by
calcareous lacustrine silt. This interval has two clear horizons of Potamogeton at the base and
top of the lacustrine silt. Two
14
C ages constrain the lake highstand interval between 12,130-
12,560 (NRC10-124-1, 65 cm) and 7320-7510 cal yr BP (NRC10-124-2, 200 cm). In NRC JQ5, the lake silts lie directly on alluvial gravels (Fig. B6). One date on marl at the base and one on
Potamogeton below the deltaic deposits are 12,630-13,070 cal yr BP (NRC JQ 5-1, 120 cm) and
5950-6210 cal yr BP (NRC JQ-5-3, 240 cm), respectively (Fig. B6).
Two
14
C dates on Potamogeton horizons in NRC10-105 are 6500-6740 cal yr BP
(NRC10-105-1, 135 cm) and 6000-6200 cal yr BP (NRC10-105-2, 300 cm), consistent with lake
regression from higher elevation, down valley (Fig. B6, B7b). Two dates from the shallow lake
sediments of NRC10-82 are 6210-6400 cal yr BP (NRC10-82-1, 25 cm) and 4100-4410 cal yr
BP (NRC10-82-3, 70 cm). Two 14C dates from similar context in NRC10-96 date to 3260-3450
cal yr BP (NRC10-96-3, 200 cm) and 2340-2480 cal yr BP (NRC10-96-4, 260 cm).
69
2.6 SYNTHESIS AND DISCUSSION
2.6.1 Ngangla Ring Tso Lake Level Chronology
Sedimentologic observations merged with
14
C and U-Th ages from shorelines tufas and
stratigraphic sections yield quite a complete picture of lake-level changes over the past 12.8 ka
(Fig. B8). Shoreline tufa ages are interpreted as the age of mean paleo-lake elevation during the
period dated, although modern lake level is observed to fluctuate tens of centimeters annually.
Radix gastropods inhabit shallow, calm waters and are an indicator of near-shore depths of a few
meters (Taft et al., 2012). Potamogeton has been shown to occur mostly at depths <1 m in a
similar Tibetan lake, Ahung Tso, and dates on this material are also considered near-surface
indicators (Morrill et al., 2006).
We place the most confidence in shoreline ages that show close agreement in multiple
locations within the basins, and those supported by the dating and stratigraphic information in
the sections. We have less certainty in single
14
C dates based on tufa or shoreline cements
because of the demonstrated dating problems. Therefore, dates with a high degree of confidence
are shown in black in Figures B8 and B9, and those with lower confidence in gray. The lake
level curve is shown as a dashed line where poor age control makes lake level uncertain.
The oldest shoreline date (Table B2, NRC10-101-1) generated by U-Th at 14.9±2.5 ka is
excluded from the lake level curve, as are the 14C dates (Table B1), based on the poor precision
and stratigraphic dating problems described above. The sample is only 25 m above modern, and
does not change the lake level interpretation, but we do not reconstruct NRT lake level prior to
12.8 ka. The lowest date in section NRC JQ-5 indicates NRT was 79 m (4806 m asl) deeper than
present by 12.8 ka, near the threshold elevation connecting NRT and RS. Rising lake level is
corroborated by the oldest date in NRC10-124 (4819 m asl) at 12.3 ka. One 14C and one U-Th
date on separate tufa samples (NRC JQ-1 bottom, NRC10-87 bottom) indicate lake level fell to
~4800 m asl at ~11.3 ka, but there is no evidence in the sections for a depositional hiatus, so this
regression is uncertain. One date from the RS shoreline transect (Table 1, NRC10-5-1 bottom)
indicates that by 11.2 ka the lake may have risen to near the Holocene highstand (4853 m asl),
integrating the three basins, but this single date should be interpreted with caution. However, a
second date (NRC10-43-1) places lake level six meters higher (4859 m asl) by 10.3 ka, following
the same rising trend. Multiple dates on the high shoreline in both the NRT and RS basins
70
indicate the highstand elevation was maintained around 4862-4864 m asl, with minor
fluctuations, from 10.0 to 8.6 ka (Table B1). The shell dated to 9.2 ka (Fig. B6, Table B1,
NRC10-107-1) in section 10-107 at 4,831 m asl falls within this interval. However, the coarse
grain size, erosive contacts and thick alluvium in the section indicates the catchment is very
active and reworking of the shell is probable (Fig. B6). We do not interpret a lake-level
oscillation for this date.
The deepest interval 10.0-8.6 ka is followed by an abrupt lowering of NRT lake level to
at least 4817 m asl, culminating at ~8.3 ka. The paired 14C (Table B1) and U-Th series (Table
B2) dates for sample NRC10-88 and two sets of paired dates from Kailas 21 indicate the lake
may have dropped to 4805 m asl, before returning to 4823 m asl by ~7.8 ka. Lake regression to
4805 m asl is uncertain because the uppermost dated lake sediments of NRC10-124 (Fig. B6), at
4819 m asl, postdate this drop at 7.4 ka (NRC10-124-2, 200 cm), and show no evidence of prior
lake regression. However, satellite imagery of the surrounding region shows the section is
located in a small closed topographic low. It is possible that this closed depression, upstream of
the main lake, was drained later after headward erosion by the Amo Tsangpo. If this is the case,
then a lake level drop to 4805 m asl is possible. Several tufa 14C dates at elevations lower than
4805 m asl fall in the interval of 8.3-7.4 ka (Table B1, Fig. B8), but none of the stratigraphic
sections below this elevation show evidence of lake retreat during this interval (Fig. B6), so it is
unlikely the lake fell below the threshold elevation connecting RS (Fig. B3).
Multiple dates indicate lake level fell significantly after 6.7 ka. One tufa date (Table B1,
NRC 10-90 top) indicates a possible small lake level rise to 4829 m asl by 6.7 ka, after which ten
14
C dates, including the uppermost dates in all stratigraphic sections in the RS basin, NRC JQ-5,
NRC10-105, and NRC10-82 in the NRT basin, and one tufa date (Table B1, NRC JQ-1 top),
show a dramatic regression of the lake system (Fig. B8). The single, lowest-elevation shoreline
tufa sample dated in this study (Table B1, NRC10-98-1) indicates lake level may have reached
4736 m asl at 5.9 ka. Results from the RS shoreline sections compare closely with the NRT
streamcut sections, including the drop below the threshold elevation at 4805 m asl in both the
NRT and RS basins. There is, however, a clear channel visible on satellite imagery leading from
the spilling threshold of RS draining to NRT, so overflow likely occurred.
This major regression was followed by a minor lake level rise to 4775 m asl by 4.9 ka,
based on two U-Th dates on spring tufa (Table B2, Kailas 20a, 20d). Although undated, the
71
upper 1 m of section NRC10-105 (4757 m asl) contains eroded lake sediments that likely
correspond to this transgression (Fig. B6, B7b). The lacustrine interval in NRC10-82, located on
the other side of the basin, also produced an uppermost 14C date (NRC10-82-3) corresponding to
the regression from this peak at 4.0 ka (Fig. B6).
The dates (NRC10-96-3, 96-4) from NRC10-96 (4732 m asl), indicate lake level dropped
to ~5 m above modern from 3.3 to 2.4 ka (Fig. B6). The dated silt horizon conformably overlies
deep water clay, indicating lake level did not reach this elevation prior to 3.3 ka. We obtained no
14
C dates younger than 2.4 ka, with the exception of two dates from the RS basin at 4824 m asl
(NRC10-39-1) and 4839 m asl (NRC10-40-1). These dates, produced from soil carbonate rinds
on beach pebbles, are not related to the lake record (Table B1). NRC10-39-1 and 40-1 have δ18O
values (PDB) of -12.1‰ and -14.2‰, respectively, much more negative than tufa in the NRT
lake system, which ranges -2.9‰ to -7.0‰, supporting this conclusion. Therefore NRT has been
at or below the modern elevation of 4727 m asl since ~2.4 ka. The drowned shorelines ridges
visible in NRT likely record this recent lowstand.
2.6.2 Regional climate interpretation and comparison
Lake-level changes in the NRT system record changes in the water budget of the basin.
The ratio of lake area:basin area (Aw hereafter) is a better measure of the hydrologic budget of a
closed-basin lake than depth (Benson and Paillet, 1989, Mifflin and Wheat, 1979), but for the
NRT system the two variables are highly linearly correlated (r2=0.996), so we included an Aw
ratio scale bar along with lake level for comparison to other records (Fig. B8, B9).
We interpret changes in NRT lake area to dominantly reflect changes in annual precipitation
based on the close agreement in the modern and highstand Aw values and the magnitude of Aw
expansion for the NRT system during the early Holocene (4.54 times modern), the nearby,
nonglaciated Baqan Tso system (Fig. B1, B2; 4.35 times modern) (Huth et al., 2015), and most
other closed basin lake systems in western Tibet (Hudson and Quade, 2013). The similar
expansion of glaciated and nonglaciated lake systems in Tibet (Hudson and Quade, 2013)
indicates a common climatic forcing for lake area change, ruling out glacial melt as the driver for
rising lake level during the Late Glacial/Holocene transition. Furthermore, modeled equilibrium
line altitudes Tibetan Plateau glaciers indicate melting was likely lower than modern during the
early and mid-Holocene due to increased cloud cover and relative humidity (Rupper et al., 2009).
72
Since summer solar insolation was higher during the early Holocene (Berger and Loutre, 1991),
the potential for evaporation and soil evapotranspiration was increased, although higher summer
cloud cover associated with stronger monsoon climate could have reduced this effect. Any
decrease in evaporative flux from the lake surface would have been recouped by the more than
fourfold increase in available lake area.
Because the plateau interior is arid, annual soil
evapotranspiration is limited by and closely follows annual rainfall amount (Xu et al., 2005). It
is unlikely that evapotranspiration rates could have been lowered enough to generate runoff to
increase lake areas fivefold without increased precipitation. Therefore, precipitation must be
responsible for creating and sustaining high lake areas in the NRT record.
The timing of NRT lake area change is broadly similar to the majority of Holocene
paleoclimate records from Tibet. The overall pattern follows the trend of Northern Hemisphere
summer insolation, peaking around 11.0 ka and falling towards the present (Fig. B9). This
agrees well with the lake area reconstruction from Baqan Tso to the north (Fig. B2, B9; Huth et
al., this volume), providing further confidence in the lake level chronologies and interpretation as
regional precipitation records. Because of the uncertainty in published age models, particularly
for lake core reconstructions (Wischnewski et al., 2011; Hou et al., 2012), it is impossible to say
whether NRT lake area changes lead or lag inferred changes elsewhere.
However, our
interpretation of increased ISM precipitation during the early and mid-Holocene is in agreement
with the majority of core-based records (Van Campo and Gasse, 1993; Kashiwaya et al., 1995;
Gasse et al., 1996; Shen et al., 2005; Herzschuh et al., 2006; Morrill et al., 2006; Demske et al.,
2009; Kramer et al., 2010; Li et al., 2011; Lu et al., 2011; Wischnewski et al., 2011). Abrupt
millennial-scale changes in NRT Aw superimposed on the orbital-scale trend, indicate
precipitation varied significantly in response to additional climate forcing during key time
periods.
Late Glacial-Holocene Transition (12.8-11.2 ka)
The oldest reliable age for the most recent lake cycle is 12.8 ka, indicating rising lake
levels, with a lake~ 3x modern area, and therefore increasing rainfall. This matches pollen
evidence from Tso Kar, located in the Ladakh Range (Fig. B1), indicating increased monsoon
circulation between 12.9 and 12.5 ka (Demske et al., 2009). In the Tso Kar record, this was
followed by a dry phase after 12.5 ka and centered at 12.1 ka with increased Westerly-
73
transported pollen, which the authors correlate to reduced monsoon strength during the Younger
Dryas stade (YD), 12.8-11.7 ka (Stuiver and Grootes, 2000). By the late YD, Westerly-derived
pollen disappears, indicating dry conditions in the northwest Himalaya, which coupled with
decreased monsoon strength, could result in a deficit in annual lake hydrologic budgets.
Decreased temperature and monsoon strength is also inferred starting at ~12.5-12.0 ka in
proglacial lake sediments deposited in Himalayan drainages south of NRT (Juyal et al., 2004;
Beukema et al., 2011). The NRT record shows a possible minor lake regression after 12.4 ka,
with a low point centered on 11.5 ka, but there is no abrupt change associated with the YD.
Reduced precipitation under lower temperatures may have only slowed NRT lake level rise
during this time period, consistent with stratigraphic evidence from NRC10-124 showing no
regression until the mid-Holocene (Fig. B6).
Reduced monsoon precipitation as early as 13.0 ka is also inferred in the speleothem δ18O
record from Timta Cave (Sinha et al., 2005), located due south of NRT, in close agreement with
the Dongge Cave reconstruction (Fig. B9d, Dykoski et al., 2005).
Early Holocene Highstand (11.2-8.6 ka)
The rise to the Holocene highstand of NRT from 11.2-10.4 ka resembles other records
from western Tibet, showing increase in temperature and precipitation associated with a stronger
ISM.
This is inferred in pollen records from Sumxi Tso (Van Campo and Gasse, 1993),
Bangong Tso (Gasse et al., 1996) and Tso Kar (Demske et al., 2009) (Fig. B1).
10
Be exposure
ages from the high shorelines at Sumxi Tso (Kong et al., 2007) fall in this interval, and sediment
proxy evidence indicates Bangong Tso overflowed into the Indus Basin around this time (Gasse
et al., 1996). The NRT highstand was maintained until 8.6 ka. For Baqan Tso, maximum lake
level is dated at 10.9-9.5 ka, with significant fluctuations (Huth et al., this issue). Two tufa dates
(DT10-1, 9.9 ka; DT10-8, 11.2 ka) for the high shoreline in the Chabyer Tsaka basin east of
NRT fall in this time period (Hudson and Quade, 2013). Tso Kar reached maximum lake level
only by ~8.0 ka (Wünneman et al., 2010), but because of tectonic changes, it was not in a closed
basin prior to ~12.0 ka and so lake level history may differ from lakes with consistent basin
hydrography. Further east, OSL dating of shoreline sands surrounding Linggo Tso in central
Tibet also indicate the most recent highstand occurred at ~9.6 ka (Pan et al., 2012). Tangra
74
Yumtso is a notable outlier in this pattern, where OSL dating of lake sediments indicate lake
level did not rise until ~7.6 ka (Long et al., 2012).
A mid-Holocene peak in effective moisture has been indicated by some lake proxy
records in the central and eastern plateau, including Zigetang Tso (Herzschuh et al., 2006), Cuoe
Lake (Wu et al., 2006), and Qinghai Lake (Shen et al., 2005), but other records in the region
including Pumo Yumtso (Lu et al., 2011), Kuhai Lake (Wischnewski et al., 2011), Naleng Lake
(Kramer et al., 2010), Ximencuo (Zhang et al., 2009), and Ahung Tso (Morrill et al., 2006)
variably indicate early Holocene peak moisture or a broad peak spanning the early and midHolocene (Fig. B1). At face value, this indicates periods of higher precipitation in Tibet are
regionally heterogeneous. However, uncertainty in age models and proxy interpretations could
explain the disagreement between records (Wischnewski et al., 2011; Hou et al., 2012).
The NRT highstand period coincides broadly with the lowest δ18O values in the
speleothems for lowland China (Dongge Cave, Wang et al., 2005; Dykoski et al., 2005), central
Tibet (Tianmen Cave, Cai et al., 2012) and the initial peak in the Qunf Cave record in Oman
(Fleitmann et al., 2003) (Fig. B1, B9). It is also similar to the period of highest δ18O values
found in the Guliya ice core record (Thompson et al., 1997), and strongest monsoon upwelling
along the Oman margin (Gupta et al., 2003) (Fig. 9b). Early Holocene glacial advances are
noted in the Himalaya (Owen et al., 2009) and in the Qinghai region of the plateau (Yi et al.,
2008). These are possibly associated with increased precipitation, but as the majority of Tibetan
glaciers are most sensitive to ablation, reduced melting due to higher cloud cover is also likely
(Rupper and Roe, 2008; Rupper et al., 2009).
Holocene Lake Regression (8.6-2.4 ka)
The NRT record indicates lake decline due to rainfall decrease after 8.6 ka. This overall
declining trend is punctuated by three abrupt decreases, centered at 8.5, 5.9, and 3.6 ka. The
general trend of declining lake level is similar to other shoreline-based chronologies from Sumxi
Tso (Kong et al., 2007), Tso Kar (Wünneman et al., 2010), Chabyer Caka/Lagkor Tso (Lee et al.,
2009; Hudson and Quade, 2013), Baqan Tso (Huth et al., 2015), and Linggo Tso (Pan et al.,
2012) (Fig. B1). The majority of paleoclimate records from the central (Herzschuh et al., 2006;
Morrill et al., 2006; Wu et al., 2006; Li et al., 2011) and eastern Tibetan Plateau (Kramer et al.,
75
2010; Zhang et al., 2009) also show gradual weakening of monsoon climate throughout the
Holocene.
Abrupt precipitation decreases in the NRT record correlate closely with inferred weak
monsoon events in speleothem records and the Oman upwelling record (Fig. B9b, c, d, e, f).
Abrupt weak monsoon events are also implicated in other plateau paleoclimate records (Van
Campo and Gasse, 1993; Gasse et al., 1996), and are coeval with glacial advances in the southern
Himalaya (Owen et al., 2009) and Muztag Ata region of the Pamir (Seong et al., 2009).
Late Holocene (2.3 ka to present)
After 2.3 ka NRT lake level indicates similar or lower-than-modern precipitation.
Geomorphically, the majority of shoreline benches were constructed during lake regression, so
drowned shorelines indicate NRT lake level was below modern and recently transgressed.
Monsoon strength rise toward the present is suggested in the Bangong Tso record (Gasse et al.,
1996) between ~2.1-1.3 ka, in the Tso Kar pollen record ~3.1-1.3 ka (Demske et al., 2009) and
consistent with a trend of increasing monsoon intensity inferred in the Oman upwelling record
(Gupta et al., 2003), and speleothems in the monsoon region (Fleitmann et al., 2003; Wang et al.,
2005; Dykoski et al., 2005)(Fig. B9).
2.6.3 Implications for Holocene ISM Climate in Tibet
Hydrologic budget modeling by Huth et al. (this issue) for Baqan Tso, directly north of
the NRT system, indicates the rainfall increase required to produce the early Holocene lake
highstand was ~55% greater than modern, and generally declined towards the present to ~25%
greater by 5.0 ka. The close agreement at orbital and millennial scales between other western
Tibet proxy records and the Baqan Tso and NRT shoreline records indicates annual rainfall
likely varied between <100-150% of modern, in phase with the observed changes in monsoon
climate. Precipitation in the western Tibetan Plateau is highly correlated with modern ISM
precipitation (Conroy and Overpeck, 2011) and summer moisture is derived from south of the
Himalaya today (Tian et al., 2007). The NRT record shows this pattern is reflected consistently
in climate variation on longer timescales. However, the west receives a larger proportion of
Westerly-derived winter moisture than the eastern plateau today (Bookhagen et al., 2010).
Because the NRT and Baqan Tso hydrologic budgets respond to annual precipitation, we cannot
76
resolve how rainfall seasonality varied in the past. The Tso Kar pollen record indicates monsoon
and Westerly circulation intensity are generally inversely correlated throughout the Late Glacial
and Holocene (Fig. B1; Demske et al., 2009).
Therefore, it is likely that the increased
precipitation in western Tibet was summer monsoon-derived during the Holocene lake
maximum, with a smaller-than-modern proportion of winter precipitation.
The NRT record shows precipitation in western Tibet was reduced during widely
identified weak monsoon intervals throughout the monsoon region (Fig. B9; Cai et al., 2012;
Gupta et al., 2003; Fleitmann et al., 2003; Wang et al., 2005). Unlike changes in other records,
the NRT lake level drops are step changes with little lake level recovery following weak
monsoon events. This may reflect a hydrologic budget deficit during the weak event from which
lake level did not recover under overall decreasing Holocene summer insolation.
Millenial-scale weak monsoon events have been correlated to Holocene ice-rafted debris
(IRD) events in the North Atlantic (Fig. B9a; Bond et al., 2001), and the ~1500 yr periodicity of
occurrence has been attributed to solar forcing for both the North Atlantic and the ISM region
(Gupta et al., 2005). However, changes in solar insolation over the Holocene, including during
the weak intervals, are small: on the order of 1 W/m2 (Steinhilber et al., 2009). Such small
changes in insolation would likely require some amplification within the global climate system.
Analysis of modern climate datasets and climate modeling indicate solar forcing may act by
directly modulating Indian Ocean sea surface temperatures (Kodera et al., 2004); or indirectly,
via atmospheric teleconnection most likely associated with changed Westerly circulation during
weak monsoon intervals, not necessarily associated with solar forcing (Goswami et al., 2006;
Luo et al., 2011; Zhu et al., 2011). Recent study of North Atlantic IRD events found a 1500 yr
cycle, but found no similar cyclicity in reconstructed solar insolation (Darby et al., 2012). The
authors instead suggest the IRD records reflect millennial-scale variation in the Arctic
Oscillation with a dubious relationship to solar forcing. This finding is compatible with modern
observations of a North Atlantic-ISM atmospheric teleconnection, and is a viable hypothesis for
the coincidence between the IRD records and ISM weak monsoon events during the Holocene.
While the cause of the 1500 yr cycle remains in question, the NRT record clearly shows that it
coincides with substantial changes in rainfall in southwestern Tibet.
77
2.7 CONCLUSIONS
Dating lake shoreline tufas in the NRT system encountered many known problems
inherent with dating carbonates in closed-basin lakes. The lake record highlights the need for
careful assessment of
14
C reservoir effects and reproducibility of shoreline ages throughout the
paleolake basin. Based on shoreline dating and stratigraphic evidence, NRT lake level and Aw
varied coincident with other records of ISM precipitation on the Tibetan Plateau and elsewhere
in the monsoon region at both orbital and millennial timescales. This indicates that monsoon
circulation changes, driven by direct insolation forcing or influenced by regional climate
teleconnections, resulted in substantial changes in monsoon precipitation amount, which are
reflected in the vegetation, hydrologic budget of lakes, and speleothem δ18O records in the
region.
2.8 ACKNOWLEDGEMENTS
This research was funded by grants from the Comer Science and Education Foundation to
JQ and AMH and the Henry Luce Foundation to JWO, NSF Grants 1211299, 1103403, and
0908792 to RLE and HC, and a student grant to AMH from the Geological Society of America.
GL was supported by NSF-China grant 41101189. The authors thank the editor and anonymous
reviewers for comments that improved the manuscript.
2.9 REFERENCES
Benson, L.V., and Paillet, F.L., 1989, The use of total lake-surface area as an indicator of
climatic change: examples from the Lahontan Basin: Quaternary Research, v. 32, p. 262-275.
Benson, L., Kashgarian, M., Rubin, M., 1995. Carbonate deposition, Pyramid Lake subbasin,
Nevada: 2. Lake levels and polar jet stream positions reconstructed from radiocarbon ages
and elevations of carbonates (tufas) deposited in the Lahontan basin. Palaeogeography,
Palaeoclimatology, Palaeoecology 117, 1-30.
78
Berger, A., Loutre, M.F., 1991. Insolation values for the climate of the last 10 million years.
Quaternary Science Reviews 10, 297-317.
Beukema, S.P., Krishnamurthy, R.V., Juyal, N., Basavaiah, N., Singhvi, A.K., 2011. Monsoon
variability and chemical weathering during the late Pleistocene in the Goriganga basin,
higher central Himalaya, India. Quaternary Research 75, 597-604.
Blard, P.-H., Sylvestere, F., Tripati, A.K., Claude, C., Causse, C., Coudrain, A., Condom, T.,
Seidel, J.-L., Vimeux, F., Moreau, C., Dumoulin, J.-P., Lavé, J., 2011. Lake highstands on
the Altiplano (Tropical Andes) contemporaneous with Heinrich 1 and the Younger Dryas:
new insights from 14C, U-Th dating and δ18O of carbonates. Quaternary Science Reviews 30,
3973-3989.
Bond, G., Kromer, B., Beer, J., Muscheler, R., Evans, M.N., Showers, W., Hoffman, S., LottiBond, R., Hajdas, I., Bonani, G., 2001. Persistent solar influence on North Atlantic climate
during the Holocene. Science 294 2130-2136.
Bookhagen, B., 2010. Appearance of extreme monsoonal rainfall events and their impact on
erosion in the Himalaya. Geomatics, Natural Hazards and Risk 1, 37-50.
Boos, W.R., Kuang, Z., 2010. Dominant control of the South Asian monsoon by orographic
insulation versus plateau heating. Nature 463, 218-222.
Cai, Y., Zhang H., Cheng H., An Z., Edwards, R.L., Wang X., Tan L., Liang F., Wang J., Kelly,
M., 2012. The Holocene Indian monsoon variability over the southern Tibetan Plateau and its
teleconnections. Earth and Planetary Science Letters 335, 135-144.
Cheng H., Edwards, R.L., Shen C., Polyak, V.J., Asmerom, Y., 2000. The half-lives of uranium234 and thorium-230. Chemical Geology 169, 17-33.
79
Cheng H., Edwards, R.L., Broecker, W.S., Denton, G.H., Kong X.G., Wang Y.J., Zhang R.,
Wang X.F., 2009. Ice Age terminations. Science 326, 248-252.
Cheng H., Sinha, A., Wang X., Cruz, F.W., Edwards, R.L., 2012. The Global Paleomonsoon as
seen through speleothem records from Asia and the Americas. Climate Dynamics 39, 10451062.
Cheng H., Edwards, R.L., Shen C.C., Polyak, V.J., Asmerom, Y., Woodhead, J., Hellstrom, J.,
Wang Y.J., Kong X.G., Spotl, C., Wang X.F., Alexander, E.C., 2013. Improvements in Th230, Th-230 and U-234 half-life values, and U-Th isotopic measurements by multi-collector
inductively coupled plasma mass spectrometry. Earth and Planetary Science Letters 371, 8291.
Conroy, J.L., and Overpeck, J.T., 2011. Regionalization of present-day precipitation in the
greater monsoon region of. Journal of Climate 24, 4073–4095.
Darby, D.A., Ortiz, J.D., Grosch, C.E., Lung, S.P., 2012. 1,500-year cycle in the Arctic
Oscillation identified in Holocene Arctic sea-ice drift. Nature Geoscience 5, 897-900.
Demske, D., Tarasov, P.E., Wünneman, B., Riedel, F., 2009. Late glacial and Holocene
vegetation, Indian monsoon and westerly circulation in the Trans-Himalaya recorded in the
lacustrine pollen sequence from Tso Kar, Ladakh, NW. Palaeogeography, Palaeoclimatology,
Palaeoecology, 279, 172–185.
Dykoski, C.A., Edwards, R.L., Cheng H., Yuan D., Cai Y., Zhang M., Lin Y., Qing J., An Z.,
Revenaugh, J., 2005. A high-resolution absolute-dated Holocene and deglacial Asian
monsoon record from Dongge Cave, China. Earth and Planetary Science Letters 233, 71-86.
Fleitmann, D., Burns, S.J., Mudelsee, M., Neff, U., Kramers, J., Mangini, A., Matter, A., 2003.
Holocene forcing of the Indian Monsoon recorded in a stalagmite from southern Oman.
Science 300, 1737-1739.
80
Gasse, F., Fontes, J.C., Van Campo, E., Wei, K., 1996. Holocene environmental changes in
Banggong Co basin (western Tibet). 4. Discussion and conclusions. Palaeogeography,
Palaeoclimatology, Palaeoecology 120, 79–92.
Goswami, B.N., Madhusoodanan, M.S., Neema, C.P., Sengupta, D., 2006. A physical
mechanism for North Atlantic SST influence on the Indian summer monsoon. Geophysical
Research Letters 33, L02706, doi:10.1029/2005GL024803.
Gupta, A., Anderson, D.M., Overpeck, J.T., 2003. Abrupt changes in the Asian southwest
monsoon during the Holocene and their links to the North Atlantic Ocean. Nature 421, 254–
357.
Gupta, A.K., Das, M., Anderson, D.M., 2005. Solar influence on the Indian summer monsoon
during
the
Holocene.
Geophysical
Research
Letters
32,
L17703,
doi:10.1029/2005GL022685.
Herzschuh, U., Winter, K., Wünneman, B., Li, S., 2006. A general cooling trend on the central
Tibetan Plateau throughout the Holocene recorded by the Lake Zigetang pollen spectra.
Quaternary International 154–155, 113–121.
Hou J., D’Andrea, W.J., Liu Z., 2012. The influence of
14
C reservoir age on interpretation of
paleolimnological records from the Tibetan Plateau. Quaternary Science Reviews 48, 67-79.
Hudson, A.M., Quade, J., 2013. Long term east-west asymmetry in monsoon climate on the
Tibetan Plateau. Geology 41, 351-354.
Jaffey, A.H., Flynn, K.F., Glendeni, L.E., Bentley, W.C., Essling, A.M., 1971. Precision
measurement of half-lives and specific activities of 235U and 238U. Physics Review C 4, 18891906.
81
Juyal, N., Pant, R.K., Basavaiah, N., Yadava, M.G., Saini, N.K., Singhvi, A.K., 2004. Climate
and seismicity in the higher Central Himalaya during 20-10 ka: evidence from the Bargayang
basin, Uttaranchal, India. Palaeogeography, Palaeoclimatology, Palaeoecology 213, 315-330.
Kapp, P., Taylor, M., Stockli, D., Ding L., 2008. Development of active low-angle normal fault
systems during orogenic collapse: insight from Tibet. Geology 36, 7-10.
Kashiwaya K., Masuzawa T., Morinaga H., Yaskawa K., Yuan B., Liu J., Gu, Z., 1995. Changes
in hydrological conditions in the central Qing-Zang (Tibetan) Plateau inferred from lake
bottom sediments. Earth and Planetary Science Letters 135, 31-39.
Kodera K., 2004. Solar influence on the Indian Ocean Monsoon through dynamical processes.
Geophysical Research Letters 31, L24209, doi:10.1029/2004GL020928.
Kong, P., Na, C., Fink, D., Huang, F., Ding, L., 2007. Cosmogenic
10
Be inferred lake-level
changes in Sumxi Co basin, western Tibet. Journal of Asian Earth Sciences 29, 698–703.
Kramer, A., Herzschuh, U., Mischke, S., and Zhang, C., 2010. Late glacial vegetation and
climate oscillations on the southeastern Tibetan Plateau inferred from the Lake Naleng
pollen. Quaternary Research 73, 324–335.
Kummerow, C., Barnes, W., Kozu, T., Shiue, J., and Simpson, J., 1998, The Tropical Rainfall
Measuring Mission (TRMM) sensor package. Journal of Atmospheric and Oceanic
Technology 15, 809–817.
Lao, Y., Benson, L., 1988, Uranium-series age estimates and paleoclimatic significance of
Pleistocene tufas from the Lahontan Basin, California and Nevada. Quaternary Research 30,
165-176.
Lei G.L., Zhang H.C., Li Z.Z., Hudson, A.M., Zhu Y., Jiang X.Y., Chen X.L., Chang F.Q., Li
H.Y., 2013. Geochemical features and significance of shoreline tufas from a closed-basin
82
lake Ngangla Ring Tso in the western Tibetan Plateau. Quaternary Science 33, 839-847 (in
Chinese).
Lee, J., Li, S., Aitchison, J.C., 2009. OSL dating of paleoshorelines at Lagkor Tso, western
Tibet. Quaternary Geochronology 4, 335–343.
Levin, I., Kromer, B., Hammer, S., 2013. Atmospheric Δ14CO2 trend in Western European
background air from 2000 to 2012. Tellus B 65, 20092.
Li, L., Yang, S., Wang, Z., Zhu, X., Tang, H., 2010. Evidence of warming and wetting climate
over the Qinghai-Tibet Plateau. Arctic, Antarctic, and Alpine Research 42, 449–457.
Li, Q., Lu, H., Zhu, L., Wu, N., Wang, J., Lu, X., 2011. Pollen-inferred climate changes and
vertical shifts of alpine vegetation belts on the northern slope of the Nyainqentanglha
Mountains (central Tibetan Plateau) since 8.4 kyr BP. The Holocene 21, 939–950.
Liu K.B., Yao Z., Thompson, L. G. 1998. A pollen record of Holocene climatic changes from the
Dunde ice cap, Qinghai-Tibetan Plateau. Geology 26, 135-138.
Long H., Lai Z.P., Frenzel, P., Fuch, M., Haberzettl, T., 2012. Holocene moist period recorded
by the chronostratigraphy of a lake sedimentary sequence from Lake Tangra Yumco on the
south Tibetan Plateau. Quaternary Geochronology 10, 136-142.
Lu X., Zhu L., Nishimura M., Morita M., Watanabe T., Nakamura T., Wang Y., 2011. A highresolution environmental change record since 19 cal ka BP in Pumoyum Co, southern Tibet.
Chinese Science Bulletin: Geology 56, 2931-2940.
Luo F., Li S., Furevik, T., 2011. The connection between the Atlantic Multidecadal Oscillation
and the Indian Summer Monsoon in Bergen Climate Model Version 2.0. Journal of
Geophysical Research 116, D19117, doi:10.1029/2011JD015848.
83
McGee, D., Quade, J., Edwards, R.L., Broecker, W.S., Cheng H., Reiners, P.W., Evenson, N.,
2012. Lacustrine cave carbonates: Novel archives of paleohydrologic change in the
Bonneville Basin (Utah, USA). Earth and Planetary Science Letters 351, 182-194.
Mifflin, M.D., Wheat, M.M., 1979. Pluvial lakes and estimated pluvial climates of Nevada.
Nevada Bureau of Mines and Geology, Reno, Nevada, 57 p.
Mischke, S., Kramer, M., Zhang C., Shang H., Herzschuh, U., Erzinger, J., 2008, Reduced early
Holocene moisture availability in the Bayan Har Mountains, northeastern Tibetan Plateau,
inferred from a multi-proxy lake record. Palaeogeography, Palaeoclimatology, Palaeoecology
267, 59-76.
Morrill, C., Overpeck, J.T., Cole, J.E., Liu, K.-B., Shen, C.M., Tang, L.Y., 2006. Holocene
variations in the Asian monsoon inferred from the geochemistry of lake sediments in central
Tibet. Quaternary Research 65, 232–243.
Mügler, I., Gleixner, G., Gunther, F., Mäusbacher, R., Daut, G., Schütt, B., Berking, J., Schwalb,
A., Schwark, L., Xu B., Yao T., Zhu L., Yi C., 2010. A multi-proxy approach to reconstruct
hydrological changes and Holocene climate development of Nam Co, central Tibet. Journal
of Paleolimnology 43, 625–648.
Owen, L.A., 2009. Latest Pleistocene and Holocene glacier fluctuations in the Himalaya and
Tibet. Quaternary Science Reviews 28, 2150-2164.
Pan B., Yi C., Jiang T., Dong G., Hu G., Jin Y. 2012. Holocene lake-level changes of Linggo Co
in central Tibet. Quaternary Geochronology 10, 117-122.
Pausata, F.S.R., Battisti, D.S., Nisancioglu, K.H., Bitz, C.M., 2011. Chinese stalagmite δ18O
controlled by changes in the Indian monsoon during a simulated Heinrich event. Nature
Geoscience 4, 474–480.
84
Placzek, C., Quade, J., and Patchett, P.J., 2006. Geochronology and stratigraphy of late
Pleistocene lake cycles on the southern Bolivian Altiplano: Implications for causes of
tropical climate change. Geological Society of America Bulletin 118, 515–532.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey,
C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P.,
Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac,
F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney,
C.S.M., van der Plicht, J., Weyhenmeyer, C.E., 2009. INTCAL09 and MARINE09
radiocarbon age calibration curves, 0-50,000 years cal BP. Radiocarbon 51, 1111-1150.
Rupper, S., Roe, G., 2008. Glacier changes and regional climate: A mass and energy balance
approach. Journal of Climate 21, 5384-5401.
Rupper, S., Roe, G., Gillespie, A., 2009. Spatial patterns of Holocene glacier advance and retreat
in Central Asia. Quaternary Research 72, 337-346.
Seki, O., Meyers, P.A., Yamamoto, S., Kawamura, K., Nakatsuka, T., Zhou, W., Zheng, Y.,
2011. Plant-wax hydrogen isotopic evidence for postglacial variations in delivery of
precipitation in the monsoon domain of China. Geology 39, 875–878.
Seong, Y.B., Owen, L.A., Yi C., Finkel, R.C., 2009. Quaternary glaciation of Muztag Ata and
Kongur Shan: Evidence for glacier response to rapid climate changes throughout the Late
Glacial and Holocene in westernmost Tibet. Geological Society of America Bulletin 121,
348-365.
Severinghaus, J.P., Beaudette, R., Heandly, M.A., Taylor, K., Brook, E.J., 2009. Oxygen-18 of
O2 records the impact of abrupt change on terrestrial biosphere. Science 324, 1431-1434.
Shen C.C., Edwards, R.L., Cheng H., Dorale, J.A., Thomas, R.B., Moran, S.B., Weinstein, S.E.,
Edmonds, H.N., 2002. Uranium and thorium isotopic and concentration measurements by
85
magnetic sector inductively coupled plasma mass spectrometry. Chemical Geology 185, 165178.
Shen, J., Liu, X., Wang, S., Ryo, M., 2005. Palaeoclimatic changes in the Qinghai Lake area
during the last 18,000 years. Quaternary International 136, 131–140.
Sinha, A., Cannariato, K.G., Stott, L.D., Li H.C., You C.F., Cheng H., Edwards, R.L., Singh,
I.B., 2005. Variability of Southwest Indian summer monsoon precipitation during the
Bølling-Ållerød. Geology 33, 813-816.
Stuiver, M., Grootes, P.M., 2000. GISP2 oxygen isotope ratios. Quaternary Research 53, 277284.
Styron, R.H., Taylor, M.H., Sundell, K.E., Stockli, D.F., Oalmann, J.A.G., McAllister, A.T., Liu
D., Ding L. 2013. Miocene initiation and acceleration of extension in the South Lunggar rift,
western Tibet: Evolution of an active detachment system from structural mapping and (UTh)/He thermochronology. Tectonics 32, doi:10.1002/tect.20053.
Taft, L., Wiechert, U., Riedel, F., Weynell, M., & Zhang H., 2012. Sub-seasonal oxygen and
carbon isotope variations in shells of modern Radix sp. (Gastropoda) from the Tibetan
Plateau: potential of a new archive for palaeoclimatic studies. Quaternary Science Reviews
34, 44-56.
Tian, L., Yao, T., MacClune, K., White, J.W.C., Schilla, A., Vaughn, B., Vachon, R., and
Ichiyanagi, K., 2007. Stable isotopic variations in west China: a consideration of moisture
sources. Journal of Geophysical Research 112, D10112, doi:10.1029/2006JD007718.
Thompson, L.G., Yao T., Davis, M.E., Henderson, K.A., Mosley-Thompson, E., Lin, P.N., Beer,
J., Synal, H.-A., Cole-Dai, J., Bolzan, J. F. 1997. Tropical climate instability: The last glacial
cycle from a Qinghai-Tibetan ice core. Science 276, 1821-1825.
86
Van Campo, E., Gasse, F., 1993. Pollen- and diatom-inferred climatic and hydrological changes
in Sumxi Co basin (western Tibet) since 13,000 yr B.P. Quaternary Research 39, 300–313.
Wang, Y., Cheng H., Edwards, R.L., Kong X., An Z., Wu J., Kelly, M.J., Dykoski, C.A., Li X.,
2005. The Holocene Asian Monsoon: Links to solar changes and North Atlantic climate.
Science 308, 854-857.
Wischnewski, J., Mischke, S., Wang, Y., Herzschuh, U., 2011. Reconstructing climate variability
on the northeastern Tibetan Plateau since the last Lateglacial—a multi-proxy, dual-site
approach comparing terrestrial and aquatic signals. Quaternary Science Reviews, 30, 82–97.
Wu Y., Lucke, A., Jin Z., Wang S., Schleser, G.H., Battarbee, R.W., Xia W., 2006. Holocene
climate development on the central Tibetan Plateau: a sedimentary record from Cuoe Lake.
Palaeogeography, Palaeoclimatology, Palaeoecology 234, 328-340.
Wünnemann, B., Demske, D., Tarasov, P., Kotlia, B.S., Reinhardt, C., Bloemendal, J.,
Diekmann, B., Hartmann, K., Krois, J., Riedel, F., Arya, N., 2010. Hydrological evolution
during the last 15 kyr in the Tso Kar lake basin (Ladakh, India), derived from
geomorphological, sedimentological and palynological records. Quaternary Science Reviews
29, 1138-1155.
Xu J., Haginoya S., Masuda K., Suzuki K., 2005. Heat and water balance estimates over the
Tibetan Plateau in 1997-1998. Journal of the Meteorological Society of Japan 83, 577-593.
Yi C., Chen H., Yang J., Liu B., Fu P., Liu K., Li S., 2008. Review of Holocene glacial
chronologies based on radiocarbon dating in Tibet and its surrounding mountains. Journal of
Quaternary Science 23, 533-543.
Zhang C., Mischke, S., 2009. A Lateglacial and Holocene lake record from the Nianbaoyeze
Mountains and inferences of lake, glacier and climate evolution on the eastern Tibetan
Plateau. Quaternary Science Reviews 28, 1970-1983.
87
Zhu X., Bothe, O., Fraedrich, K., 2011. Summer atmospheric bridging between Europe and East
Asia: Influences on drought and wetness on the Tibetan Plateau. Quaternary International
236, 151-157.
Zimmerman, S.R.H., Steponaitis, E., Hemming, S.R., Zermeño, P., 2012. Potential for accurate
and precise radiocarbon ages in deglacial-age lacustrine carbonates. Quaternary
Geochronology 13, 81-91.
88
92
89
Laboratory
ID
AA-91440
AA-96572
AA-96573
AA-91437
AA-91439
AA-91434
AA-96571
AA-96570
AA-96578
AA-96568
AA-96569
AA-96587
AA-96586
AA-91438
AA-96577
AA-92045
AA-96574
AA-82149
Kailas 21ab
Kailas 21c
AA-82150
NRC10-89b top
AA-91432
NRC10-89b bottom
AA-91436
NRC10-89a
AA-91430
NRC10-88 bottom
AA-92044
NRC10-87 bottom
AA-91433
NRC10-87 top
AA-91431
NRCJQ-1 top
AA-96580
NRCJQ-1 bottom
AA-96579
NRCJQ-2 top
AA-96372
NRCJQ-2 bottom
AA-96581
NRC10-86-1b
AA-91429
Kailas 20a
AA-82148
Kailas 20d
AA-82151
NRC11-6-1
AA-100105
NRC10-100-1 bulk
AA-92047
NRC10-100-1 bottom AA-96575
NRC10-101
AA-92048
NRC10-101 top
AA-96576
NRC10-98-1
AA-92046
NRC11-3-1
AA-100104
NRC11-1-1a
AA-96370
NRC10-93-1
NRC10-44-1
NRC10-44-2
NRC10-92a
NRC10-92b top
NRC10-92b bottom
NRC10-4-1 top
NRC10-4-1 bottom
NRC10-43-1
NRC10-5-1 top
NRC10-5-1 bottom
NRC10-41-1b
NRC10-41-1a
NRC10-91
NRC10-40-1
NRC10-90 top
NRC10-39-1
Sample Name
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
beach cement
tufa
tufa
beach cement
beach cement
tufa
soil carbonate
tufa
soil carbonate
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
tufa
Sample
Material
NRT
NRT
RS
NRT
NRT
NRT
NRT
NRT
RS
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
RS
NRT
RS
RS
NRT
NRT
RS
RS
NRT
NRT
NRT
RS
RS
RS
RS
RS
Lake Basin
14
Shoreline elevation Fraction Modern
Uncertainty
C Age
Uncertainty
Carbon
14
(m asl)
( C yrs BP) (14C yrs BP)
4877
0.0298
0.0017
28210
460
4864
0.3453
0.0014
8540
30
4864
0.3362
0.0012
8760
30
4862
0.3272
0.0024
8970
60
4862
0.3618
0.0028
8170
60
4862
0.3501
0.0028
8430
60
4862
0.3769
0.0012
7840
20
4862
0.3493
0.0017
8450
40
4859
0.3118
0.0011
9360
30
4853
0.3584
0.0012
8240
30
4853
0.2912
0.0012
9910
30
4846
0.3348
0.0020
8790
50
4846
0.3321
0.0011
8850
30
4839
0.3585
0.0025
60
8240
4839
0.7691
0.0006
2110
10
4829
0.4862
0.0026
6050
40
4824
0.7664
0.0009
2140
10
4823
0.4122
0.0022
7100
40
4823
0.4004
0.0021
7340
40
4817
0.3906
0.0028
7550
60
4817
0.2950
0.0033
9810
90
4817
0.3800
0.0027
7770
60
4805
0.3770
0.0024
7780
40
4800
0.3237
0.0025
9060
60
4800
0.4679
0.0028
6100
50
4796
0.5078
0.0025
5440
40
4796
0.2912
0.0019
9910
50
4790
0.4148
0.0018
7070
40
4790
0.3929
0.0025
7500
50
4775
0.4242
0.0035
6890
70
4775
0.4808
0.0021
5880
40
4775
0.4722
0.0024
6030
40
4771
0.4936
0.0035
5670
60
4762
0.5601
0.0028
4660
40
4762
0.3963
0.0029
7440
60
4752
0.3096
0.0022
9420
60
4752
0.3961
0.0029
7440
60
4736
0.5169
0.0039
5300
60
4728
0.4663
0.0024
6130
40
4727
0.2629
0.0017
10730
50
TABLE B1. RADIOCARBON RESULTS
( C yrs BP)
-8375
8595
8805
8005
8265
7675
8285
9195
8075
9745
8625
8685
8075
-5885
-6935
7175
7385
9645
7605
7615
8895
5935
5275
9745
6905
7335
6725
5715
5865
-4495
7275
9255
7275
5135
---
14
14
C Age
Corrected
rangea
2-sigma calibrated
14
Median age
( C yrs BP) (calendar yrs BP) (calendar yrs BP)
---30
9480-9550
9510
30
9630-9900
9770
60
9920-10,180
10,050
60
8990-9130
9060
60
9310-9470
9390
20
8580-8650
8610
40
9420-9530
9480
30
10,250-10,480
10,370
30
8790-9090
8940
30
11,140-11,230
11,190
50
9520-9700
9610
30
9550-9700
9620
60
8700-9030
8870
---40
6630-6800
6710
---40
7850-8000
7920
40
8020-8300
8160
60
8050-8180
8120
90
10,720-11,110
10,910
60
8310-8420
8370
40
8360-8480
8420
60
9690-10160
9930
50
6660-6800
6730
40
5940-6180
6060
50
10,880-11,250
11,060
40
7660-7830
7750
50
8020-8310
8160
70
7480-7620
7550
40
6410-6630
6520
40
6570-6780
6680
---40
4980-5300
5140
60
7970-8190
8080
60
10,260-10,570
10,420
60
7970-8190
8080
60
5720-6000
5860
-------
Uncertainty
93
90
AA-100102
AA-96584
AA-96585
AA-96374
AA-98444
AA-98445
AA-100213
AA-100251
AA-96583
AA-96373
AA-96375
AA-100214
AA-100215
AA-100108
AA-98442
AA-100211
AA-100212
AA-100106
AA-100107
NRC10-107-1
NRC10-25-1
NRC10-25-2
NRC10-25-3
NRC10-124-1
NRC10-124-2
NRCJQ 5-1
NRCJQ 5-3
NRC10-21-1
NRC10-21-2
NRC10-21-3
NRC10-7b-1
NRC10-7b-2
NRC10-105-1
NRC10-105-2
NRC10-82-1
NRC10-82-3
NRC10-96-3
NRC10-96-4
0.0022
0.0016
0.0017
0.0019
0.0018
0.0020
0.0020
0.0024
0.0017
0.0019
0.0019
0.0025
0.0024
0.0023
0.0020
0.0029
0.0038
0.0029
0.0031
0.0024
0.0031
0.0035
0.0026
8510
41640
41920
5680
10620
6680
10980
5310
52190
5490
5380
5650
5580
5990
5490
5490
3840
3300
2530
1410
2600
post-bomb
5600
50
2340
2530
30
50
40
60
40
8870
30
30
40
40
40
30
50
50
40
30
20
30
n/a
40
8345
--5505
10455
6515
10815
5145
-5325
5215
5485
5415
5825
5325
5325
3675
3135
2365
-----
Samples shown in bold were used for determining the 14C reservoir effect for tufa in the Ngangla Ring Tso lake system
0.3467
0.0056
0.0054
0.4933
0.2665
0.4355
0.2549
0.5161
0.0015
0.5048
0.5115
0.4949
0.4995
0.4771
0.5046
0.5047
0.6201
0.6630
0.7300
0.8395
0.7237
1.0287
0.4981
Calibration was performed using the Calib 6.0 software and the IntCal09 calibration curve (Reimer et al., 2009), median age is the median of the 2-sigma range
4831
4819
4819
4819
4818
4819
4806
4806
4788
4788
4788
4784
4784
4759
4759
4754
4754
4732
4732
4727
4727
n/a
n/a
b
TN
RS
RS
RS
NRT
NRT
NRT
NRT
RS
RS
RS
RS
RS
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
NRT
TN
a
Radix shell
Radix shell
Radix shell
Radix shell
Potamogeton
Potamogeton
Potamogeton
marl
Radix shell
Radix shell
Radix shell
Radix shell
Radix shell
Potamogeton
Potamogeton
Potamogeton
Potamogeton
Potamogeton
Potamogeton
AA-96370
tufa
AA-100103
tufa
AA-92042 aquatic plant
AA-92043
Radix shell
NRC11-1-1b
NRC11-2-1
NRC10-56-5
NRC10-108-2
TABLE B1. RADIOCARBON RESULTS (contd.)
50
--30
50
40
60
40
-30
30
40
40
40
30
50
50
40
30
----9150-9480
--6220-6400
12,130-12,560
7320-7510
12,580-12,870
5750-5990
-6000-6200
5910-6170
6210-6400
6030-6300
6500-6740
6000-6200
5950-6270
3870-4150
3260-3450
2340-2480
-----
9310
--6310
12,350
7420
12,730
5870
-6100
6040
6300
6160
6620
6100
6110
4010
3350
2410
-----
94
91
4877
4862
4829
4823
4823
4805
4800
4775
4775
4752
4727
4727
NRC10-93-1
NRC10-92b bottom
NRC10-90-1
Kailas 21a
Kailas 21c
NRC10-88 bottom
NRC10-87 bottom
Kailas 20a
Kailas 20d
NRC10-101-1
NRC11-1-1a
NRC-11-1-1b
±4.5
±9.6
±772.0
±89.0
±80.9
±175.2
±2.4
1599.6 ±2.9
17214.0 ±69.6
7602.3 ±23.1
18457 ±105
18230 ±138
2908.0
1852.8
45091.5
19255.9
15696.7
17261.8
1194.1
U
(ppb)
238
20.9
787.9
854.2
328.9
279.5
268.1
42.3
116
53
283.7
61.5
128.5
±0.4
±16.2
±22.7
±6.6
±5.6
±6.0
±0.8
±2
±1
±5.7
±1.2
±2.6
Th
(ppb)
232
476.89
1.37
19.04
19.80
18.59
23.30
13.28
38.98
86.36
3.99
1.69
1.24
±16.60
±0.05
±0.94
±0.69
±0.65
±0.93
±0.48
±1.36
±3.01
±0.14
±0.06
±0.04
1.246
1.460
1.484
1.514
1.502
1.515
1.441
1.784
1.763
1.478
1.372
1.302
±0.002
±0.005
±0.011
±0.003
±0.003
±0.006
±0.004
±0.005
±0.004
±0.002
±0.003
±0.002
234
U/238U
(activity)
1.1183
0.1898
0.1178
0.1104
0.1081
0.1182
0.1535
0.0807
0.0807
0.2310
0.0020
0.0068
±0.0022
±0.0014
±0.0029
±0.0006
±0.0006
±0.0017
±0.0015
±0.0006
±0.0005
±0.0015
±0.0000
±0.0000
Th / 238U
(activity)
230
TABLE B2. U-Th SERIES RESULTS
Th / 232Th
(activity)
230
211396
15077
8986
8224
8112
8822
12211
5030
5094
18369
157
573
±1548
±130
±239
±51
±52
±136
±132
±44
±33
±134
±1
±3
230
Th Age (yr)
(uncorrected)
211251
6344
8617
7898
7769
8525
11501
4927
5047
14876
81
195
±1549
±6220
±350
±236
±248
±248
±519
±84
±47
±2480
±54
±268
Th Age (yr)*
(corrected)
230
211191
6284
8557
7839
7710
8465
11441
4868
4988
14816
21
135
±1549
±6220
±350
±236
±248
±248
±519
±84
±47
±2480
±54
±268
Th Age (cal yr BP)
(Before AD 1950)
230
Samples in bold were used to determine the 14C reservoir effect for the Ngangla Ring Tso lake system
*Corrected 230Th ages assume the initial 230Th/232Th atomic ratio of 4.4±2.2 x10-6. Those are the values for a material at secular equilibrium, with the bulk earth 232Th/238U value of 3.8. The errors
are arbitrarily assumed to be 50%.
Shoreline Elevation
(m asl)
Sample
Number
2-sigma 14C age range
(cal yrs BP)
-9306-9470
-7848-8000
8022-8299
8361-8480
9693-10156
6410-6632
6565-6784
10257-10573
12557-12741
1292-1344
80°E
90°E
100°E
110°E
Winter Seaso
n
25°N
7 Gerze
10
2 17
21
3
23 1
4 18
Ngangla Ring Tso
5
19
20
Lhasa
14
15
9
22
8
12
6
20°N
Summer Season
13
16
10°N
15°N
70°E
11
30°N
35°N
40°N
60°E
Figure B1. Indian Monsoon Region Overview. Study area location is shown by open star.
Settlement and meteorological station of Gerze and capitol, Lhasa, shown by filled circles.
Schematic paths of summer (gray) and winter (white) precipitation onto the Tibetan Plateau are
shown by the bold arrows. Paleoclimate records referenced in the text are designated by symbols
according to type: speleothem δ18O records are open diamonds; ocean core records are open
circles; glacial records are open hexagons, lake core records are open squares; shoreline and
lacustrine exposure records are filled squares; lake systems with both core and shoreline records
are half-filled squares. Record references are: 1 – Ahung Tso, Morrill et al., 2006; 2 – Banggong
Tso, Gasse et al., 1996; 3 – Baqan Tso, Huth et al., this volume; 4 – Chabyer Tsaka/Lagkor Tso,
Lee et al., 2009, Hudson and Quade, 2013; 5 – Cuoe Lake, Wu et al., 2006; 6 – Dongge Cave,
Dykoski et al., 2005, Wang et al., 2005; 7 – Guliya Ice Core, Thompson et al., 1997; 8 –
Hongyuan peat sequence, Seki et al., 2011; 9 – Kuhai Lake, Wischnewski et al., 2011; 10 –
Linggo Tso, Pan et al., 2012; 11 – Muztag Ata, Seong et al., 2009; 12 – Naleng Lake, Kramer
et al., 2010; 13 – ODP 723, Gupta et al., 2003; 14 – Pumo Yumtso, Lu et al., 2011; 15 - Qinghai
Lake, Shen et al., 2005; 16 - Qunf Cave, Fleitmann et al., 2003; 17 - Sumxi-Longmu Tso, Van
Campo and Gasse, 1993; Kong et al., 2007; 18 – Tangra Yumtso, Long et al., 2012; 19 –
Tianmen Cave, Cai et al., 2012; 20 – Timta Cave, Sinha et al., 2005; 21 – Tso Kar, Demske
et al., 2009, Wünneman et al., 2010; 22 – Ximencuo Lake, Zhang et al., 2009; 23 – Zigetang Tso,
Herzschuh et al., 2006.
95
92
82°E
82.5°E
31.5°N
Bachan
Tso
an
112
Kailas 20
Kailas 21
JQ1,JQ2
85
98,100
JQ5
101
105 96
86-93
JQ3
Amo Tsangpo
Sh
Tso
Nag
107,108
114
83.5°E
32°N
117,120
121
83°E
Lun
gga
r
81.5°E
Ngangla Ring Tso
124
81,82
3,4,5
7,11-6
84
125
31.5°N
56,11-1,2,3
21,25
Rinchen
Shubtso
30,33,36
41,43,44
Sh
Ga
an
31°N
9,10
ngd
ise
31°N
Lun
g
N
gar
Sh
an
50 km
82°E
82.5°E
83°E
83.5°E
Figure B2. Ngangla Ring Tso and Baqan Tso (Huth et al., this volume) basins and study sites.
Major mountain ranges are labeled. Modern lake extents are shown in black. Holocene
highstand lake levels are shown by white outline. Perennial river systems are traced in black,
alpine glaciers by white polygons. Study sites are shown by open circles and labeled according
to sample name (Table B1).
96
93
W
Holocene Highstand
4825 m
Tso Nag
4806 m
4862-64 m
+135 m
E
4805 m
Rinqen Shub Tso
4771 m
Ngangla Ring Tso
4727 m
Figure B3. Schematic cross section west to east showing the hydrography of the NRT basin.
Modern lake elevations are indicated below each lake. Spilling threshold elevations are labeled
above. The early Holocene maximum lake level highstand is shown by the dashed line.
97
94
25
20
Gerze Mean
Monthly Precipitation
Gerze Mean
Monthly Temperature
60
50
Temperature (°C)
40
10
5
30
0
20
-5
10
-10
-15
Precipitation (mm/month)
15
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
Figure B4. Gerze meteorological station (see Figure B1 for location) climatology for western
Tibetan Plateau. Monthly averaged temperature and precipitation for the period 1973-2006 are
shown in red and blue, respectively.
98
95
a.)
b.)
Shoreline Tufa
CM
d.)
c.)
Spring Mound Tufa
1m
Figure B5. Study area and sample photos. a.) background: Ngangla Ring Tso shorelines with
distinct bands of tufa, foreground: bedrock encrustations of tufa, Trimble GPS unit for scale is
30 cm long; b.) slabbed tufa sample NRC10-87. Lower laminar tufa phase and upper porous
phase tufa are separated by black line; c.) background: modern Ngangla Ring Tso, foreground:
large tufa spring mounds; d.) example Radix shell (top) and Potamogeton fragment (bottom)
used for radiocarbon dating.
99
96
Shoreline Sections
0
100
10-107-1:
9146-9483
cal yr BP
10-124-1:
12,134-12,557
cal yr BP
10-124-2:
7324-7505
cal yr BP
NRC 10-124
4,819 m
Shallow
Deltaic
200
300
400
500
600
0
100
200
clay
fine silt
s
med and
. san
d
c
o
a
r
se s
and
pebb
le
cobb
le
Section Height (cm)
300
River
Loess Gravel
0
100
200
300
400
500
600
700
800
900
0
100
200
300
400
10-25-2:
>41920
14
C yr BP
10-25-1:
>41640
C yr BP
14
10-25-3:
6218-6396
cal yr BP
NRC 10-25
4,819 m
JQ-5-1:
12,625-13,071
cal yr BP
JQ-5-3:
5948-6207
cal yr BP
NRC JQ-5
4,806 m
1300
1200
1100
1000
900
800
700
600
500
400
300
200
100
0
0
100
200
300
10-21-1:
>52190±8870
14
C yr BP
10-21-2:
5997-6195
cal yr BP
10-105-1:
6504-6735
cal yr BP
10-105-2:
5997-6195
cal yr BP
NRC 10-105
4,757 m
0
100
200
300
400
10-21-3: 59116170 cal yr BP
NRC 10-21
4,788 m
10-7b-1:
6317-6532
cal yr BP
10-7b-2:
6296-6437
cal yr BP
10-82-1:
6208-6396
cal yr BP
10-82-3:
4094-4414
cal yr BP
NRC 10-82
4,754 m
0
100
200
300
400
NRC 10-7b
4,784 m
0
100
200
300
400
10-96-3:
3263-3445
cal yr BP
10-96-4:
2335-2484
cal yr BP
NRC 10-96
4,731 m
Gravel
Sand
Potamogeton remains
Shell fossils
Crossbedding
Planar bedding
Sharp contact
Gradational contact
Erosional contact
Modern soil
Legend
Deep
Lake
Shallow
Lake
Shallow
Lake
Alluvium/
Lake Colluvium
Shallow Shallow
Lake
Deltaic
Figure B6. Radiocarbon-dated stratigraphic sections. Sections are divided by geomorphic context and then ordered by elevation from
highest to lowest. Dating results are labeled by sample name (Table B1). Depositional environment interpretation for each
stratigraphic unit is shown to the right of each stratigraphic section.
Streamcut Sections
Shallow
Lake
Section Height (cm)
clay
fine silt
s
med and
.s
coar and
se s
and
pebb
le
cobb
le
NRC10-107
4,831 m
River
Beach
Gravel
Deltaic
Deltaic
Eolian
clay
fine silt
s
med and
. san
d
c
o
a
r
se s
and
pebb
le
cobb
le
Alluvium/
Alluvium/
Beach Lake Colluvium
Lake Colluvium
Shallow
Deltaic
clay
silt
fi
n
e
s
med and
. san
d
coa
r
s
e
sand
pebb
le
cobb
le
clay
fine silt
s
med and
. san
d
coa
r
s
e
sand
pebb
le
cobb
le
Eolian Shallow
Lake
Shallow
Deltaic
clay
fine silt
s
med and
.s
coar and
se s
and
pebb
le
cobb
le
Alluvium/
Shallow
Lake Colluvium Lake Lake
clay
Lake
Beach/
Alluvial
fine silt
s
med and
.s
coar and
s
e
s
a d
n
pebb
le
cobb
le
Alluvium/
Colluvium
clay
fine silt
s
med and
.s
coar and
s
e
s
a d
n
pebb
le
cobb
le
Beach
Shallow Alluvium
Lake
Shallow
Lake
clay
fine silt
s
med and
. san
d
co
a
r
s
e sa
nd
pebb
le
cobb
le
Lake
Loess Deltaic
Sands
100
97
A 300
NRC10-21-3:
5911-6170
cal yr BP
Section Height (cm)
Beach
200
Lake
NRC10-21-2:
5997-6195
cal yr BP
Colluvium
Lake
Nearshore
Lake
0
NRC10-21-1:
>52190±8870
14
C yr BP
Shallow
Lake
Eolian
1200
Section Height (cm)
Backbar
100
B
Beach
1000
800
Shallow
Deltaic
600
NRC10-105-2:
5997-6195
cal yr BP
400
200
0
Shallow
Lake
NRC10-105-1:
6405-6739
cal yr BP
Figure B7. Photo examples of dated shoreline and streamcut sediment exposures and stratigraphic
interpretations. a.) section NRC10-21, 4788 m asl; b.) section NRC10-105, 4757 m asl.
101
98
4860
4820
0.20
0.15
4780
0.10
4740
0.05
2010 Lake Level
0
2,000
4,000
6,000
8,000
10,000
Aw (Lake Area/Basin Area)
Shoreline elevation (m asl)
U-Th Series Tufa
C Aquatic Plant
14
C Marl
14
C Tufa
14
C Shell
14
C Tufa uncertain
14
12,000
Shoreline age (cal yrs BP)
Figure B8. Ngangla Ring Tso lake level chronology. Tufa 14C dates with uncertain relationship
to the lake chronology are shown in gray. Lake level interpretation curve is dashed where
uncertain.
102
99
δ18O PDB
δ18O PDB
G. bulloides (%)
Hematite (%)
20
0
25
-9.0
0.075
2
30 B. Gupta et al., 2003
25
4,000
5
-7.0
E. Cai et al., 2012
-2.5
F. Fleitmann et al., 2003
-2.0
-1.5
-1.0
-0.5
0.0
0.100
0.05
6,000
A. Bond et al., 2001
10
3
8,000
7
-20.0
-18.0
-16.0
10,000
8
560
20
540
520
15
500
10
480
C. Wang et al., 2005
-8.0
D. Dykoski et al., 2005
-22.0
-14.0
G. Huth et al., this issue
14
14
C Tufa
C Tufa replicate
510
500
0.050
490
2011 Lake Level
H. This study
U-Th Series Tufa
14
C Aquatic Plant
0.15
14
C Marl
14
C Tufa
14
C Shell
14
C Tufa uncertain
0.10
2010 Lake Level
12,000
480
0.20
510
500
490
480
Figure B9. Comparison of Monsoon region paleoclimate records for the period 0-13 ka (see Figure 1 for record locations). a.) Percent hematite-stained grains
(ice-rafted debris) from North Atlantic Ocean core MC52-VM29-191 (Bond et al., 2001). b.) Percent abundance of planktic foraminifer Globigerina bulloides
from Oman margin core ODP 723, a proxy for monsoon wind-driven ocean upwelling, and high latitude Northern Hemisphere June insolation (Gupta et al.,
2003). c.) Dongge Cave speleothem ‘DA’ calcite δ18O record. Black vertical bars denote weak monsoon events interpreted in this record (Wang et al., 2005). d.)
Dongge Cave speleothem D4 calcite δ18O record (Dykoski et al., 2005). e.) Tianmen Cave speleothem TM-18 calcite δ18O record (Cai et al., 2012). f.) Qunf
Cave speleothem Q5 calcite δ18O record (Fleitmann et al., 2003). g.) Baqan Tso Aw record and June 30°N insolation (Huth et al., this issue). h.) Ngangla Ring
Tso Aw record, and June 30°N insolation (Berger and Loutre, 1991). Percent hematite-stained grains in a. and all δ18O axes are reversed so that lower values
(less ice-rafted debris and wetter conditions) are up. Long, vertical gray bars denote millennial-scale weak monsoon events (modified from Gupta et al., 2003).
Age (yr BP)
6
Insolation 30° N (W/m2)
2,000
4
Insolation 30° N (W/m2)
0
1
Insolation 60°N (W/m2)
BT Aw
(Lake Area/Basin Area)
15
δ18O PDB
NRT Aw
(Lake Area/Basin Area)
5
103
100
APPENDIX C.
14
APPLICATION OF C DATING OF INTERDUNE PALEOWETLAND DEPOSITS TO
CONSTRAIN THE AGE OF MID- TO LATE HOLOCENE MICROLITHIC ARTIFACTS
FROM THE ZHONGBA SITE, SOUTHWESTERN QINGHAI-TIBET PLATEAU
101
APPLICATION OF 14C DATING OF INTERDUNE PALEOWETLAND DEPOSITS TO
CONSTRAIN THE AGE OF MID- TO LATE HOLOCENE MICROLITHIC ARTIFACTS
FROM THE ZHONGBA SITE, SOUTHWESTERN QINGHAI-TIBET PLATEAU
Adam M. Hudsona
John W. Olsenb
Jay Quadea
a
Department of Geosciences, University of Arizona, Tucson, Arizona, 85721-0077, USA
b
School of Anthropology, University of Arizona, Tucson, Arizona, 85721-0030, USA
Corresponding Author:
Adam M. Hudson
Department of Geosciences
University of Arizona
Tucson, AZ 85721-0077, USA
Phone: 001-520-626-1853
Email correspondence: amhudson@email.arizona.edu
Keywords: Qinghai-Tibet Plateau; microblade technology; Holocene; reservoir effect
This manuscript was published in Geoarchaeology in June 2014.
3.1 ABSTRACT
Microlithic artifacts, some found in situ, are abundant in the Zhongba archaeological site
in southwestern Tibet. The site environment consists of extant wetlands and paleowetland
deposits found in depressions between sand dunes derived from Yarlung Tsangpo floodplain.
Constraining 14C dates from wetland vegetation and shell from one site fall between ca. 6600-
102
2600 cal. yr B.P., while a second site is dated 3400-1200 cal. yr B.P. A significant and variable
radiocarbon reservoir effect – up to 1400 14C yrs – limits these ranges to terminus post quem
constraints. The in situ artifacts are supplemented by surface collections fully characterizing raw
material and typological variability for each site. Raw material found at Zhongba is chert and
chalcedony likely sourced from Cretaceous bedrock near the site. Typologically, microblades
are non-geometric and are derived from conical and wedge-shaped cores similar to those
identified in the Qinghai Lake basin and the Chang Tang Nature Reserve of similar or greater
age. The later occupation period at Zhongba is broadly contemporaneous with sites on the
Qinghai-Tibet Plateau containing bronze and iron artifacts, indicating microlithic technology
remained an important tool-making strategy in western Tibet late into the protohistoric period.
3.2 INTRODUCTION
Human habitation of the Qinghai-Tibet Plateau showcases the human behavioral and
physiological adaptions that permitted colonization of some of the most extreme terrestrial
environments on Earth. There is considerable debate however, as to 1) how long ago humans
first ventured onto the plateau, and 2) where and how the behavioral and physiological
adaptations for permanent residence at high altitude were developed.
3.2.1 Previous Archaeological Research
Known high altitude archaeological sites of the Qinghai-Tibet Plateau yielding
assemblages of material culture are found in two main regions (Fig. C1): the northeastern plateau
spanning from Qinghai Lake to the eastern Chang Tang (e.g. An, 1982; Brantingham, Olsen, &
Schaller, 2001; Madsen et al., 2006; Rhode et al., 2006), and the southeastern plateau from the
Chamdo area to the Yarlung and upper Kyichu River valleys near Lhasa (Aldenderfer & Zhang,
2004; Aldenderfer, 2011). The earliest sites surrounding Qinghai Lake are assigned to the Late
Paleolithic (ca. 10-40 ka) and Epi-Paleolithic (ca. 6-10 ka) periods based on the absence of
ceramics, limited faunal remains, and the appearance of temporary to semi-permanent camps
with no structures (Madsen et al., 2006; Rhode et al., 2006). Only one site, Jiangxigou 2, shows
evidence of semi-permanent occupation, with multiple cultural horizons containing ceramics
(Rhode et al., 2006). All of the Qinghai sites contain abundant lithic artifacts, including
103
microliths. The sites found on the southeastern plateau are late Stone Age and much younger
than those from Qinghai (Aldenderfer & Zhang, 2004). The two most notable sites are Karou, a
fully developed Neolithic settlement site in the Chamdo region of eastern Tibet, and Qugong,
near Lhasa (Fig. C1). Karou contains evidence of domesticated pig, cultivated millet and
permanent dwellings dating to 4.2-5.9 ka (Aldenderfer and Zhang, 2004). Qugong has reported
radiocarbon dates spanning 3.1 ka-3.7 ka and also contains remains of domesticated pig and
sheep (Aldenderfer and Zhang, 2004).
3.2.2 Plateau Colonization
The prevailing archaeological hypothesis for plateau colonization is the ‘three-step
model’, which involves three phases of occupation spanning the last glacial termination
(Brantingham et al., 2003; Brantingham & Gao, 2006; Madsen et al., 2006). This model
suggests that seasonal forays by hunter/gatherer groups into Qinghai (up to 3000-4000 m asl)
began around the Last Glacial Maximum (LGM), and persisted through the early Holocene, in
continuous contact with lowland hunter/gatherers. Brantingham et al. (2003; 2006) suggest this
led to behavioral adaptations necessary to survive the high plateau conditions. The second step,
year-round occupation at intermediate elevations, is implicated as an essential transitional period
for selecting towards the high altitude physiological adaptations found in modern Tibetan
populations. The final step in the model is a move towards pastoralism, which allowed early
plateau inhabitants to subsist in the cold, and increasingly arid plateau during the late Holocene
(Madsen et al., 2006).
This colonization model is supported by a robust radiocarbon chronology for eastern
plateau archaeological sites. Radiocarbon dates from Qinghai sites indicate limited presence of
hunter/gatherer groups as early as 13.0±0.1 ka (Madsen et al., 2006). More extensive occupation
of the Qinghai Lake basin, at Jiangxigou 2, spans continuously between 9.0-5.0 ka, marking the
transition from model step one to step two. After 5.0 ka, occupation around Qinghai Lake was
greatly decreased. The mid-Holocene hypothetically corresponds to the move towards step
three: pastoralism. Despite scant direct archaeological evidence, the mid-Holocene is marked by
deforestation and ecosystem degradation on the northeastern and western fringes of the plateau,
and pervasive gullying along the Yarlung Tsangpo valley, perhaps driven by increased herd
grazing (Miehe et al., 2008, 2009; Schlütz & Lehmkuhl, 2009; Pelletier et al., 2011). In the
104
Qinghai region, ecosystem change is evident as early as 7.2 ka, within the range of occupation
for Jiangxigou 2. Pastoralist presence is likewise indicated by widespread charcoal horizons
related to selective burning, dated to as early as 6.5 ka, in the alpine valleys of northern Bhutan
(Meyer et al., 2009). In the western Hindu Kush (Fig. 1), evidence for deforestation is no older
than 5.5 ka, and is coincident with a major shift towards drier climate in the region, so human
influence is only tenuously indicated (Miehe et al., 2009).
In contrast to the east, archaeological evidence of human presence in the western plateau
is poorly documented, and it is difficult to assess the timing and path by which inhabitants
reached the region. Although there is clear evidence of pre-Buddhist occupation of western
Tibet during the late Holocene, culminating with the Zhang Zhung (or Xang Xung) polity by ca.
500 BCE (Aldenderfer & Zhang, 2004; Bellezza, 2008), the precise antiquity of the earliest sites
is unconstrained. We present a new radiocarbon chronology for the Zhongba microlithic site in
western Tibet that sheds light on the timing and nature of earlier occupation of the western
plateau.
3.3 ENVIRONMENTAL SETTING AND STUDY AREA
3.3.1 Qinghai-Tibet Plateau Climate and Paleoclimate
The average elevation of the Qinghai-Tibet Plateau is greater than 4000 m asl and
modern mean annual temperatures are at or below 0°C. More than 60% of annual precipitation
occurs during the summer from the Indian Ocean, and the rain shadow created by the Himalayan
massif makes the plateau quite arid, with annual precipitation averaging 100-400 mm/yr today
(Bookhagen, 2010). These temperature and precipitation conditions limit plant cover to grasses
and shrubs (Yu et al., 2001). Trees are restricted to wetter/warmer regions at lower elevation on
the eastern and southern edges of the plateau today, though this may in part be due to sustained
human influence since the mid-Holocene (Miehe et al., 2008, 2009; Meyer et al., 2009).
The Qinghai-Tibet Plateau underwent dramatic warming climate change following the
global LGM (Kramer et al., 2010). The earliest documented archaeological sites from the central
(Zhang & Li, 2002) and northeastern Qinghai-Tibet Plateau (Madsen et al., 2006) are of preHolocene age, so it is likely that early plateau inhabitants were affected by the global transition
from the LGM to the Holocene. During the last glacial termination, annual temperatures and
105
monsoon intensity on the Qinghai-Tibet Plateau increased, evidenced by higher lake levels
(Wünneman et al., 2010) and increased abundance of arboreal and alpine meadow pollen taxa
(Herzschuh et al., 2006; Demske et al., 2009; Li et al., 2011; Wischnewski et al., 2011).
Temperature and monsoon rainfall decreased continually following the peak monsoon period of
the early Holocene and the modern environment is likely similar to the climate and ecosystem of
the late Holocene, ca. 4.0 cal ka to present.
3.3.2 Study Sites
The Zhongba site is located in the upper Yarlung Tsangpo valley, between the Himalayan
massif and the Gangdise Mountains (Fig. C1). It is located in a small system of eolian dunes
with an area of ~2 km2 deposited in a southwest-facing valley of the Gangdise range. Dune sand
is sourced locally from the floodplain of the Yarlung Tsangpo and tributaries directly southwest
of the site and transported by westerly to southwesterly winds, the dominant local wind direction
both during the winter and summer monsoon seasons (Tian et al., 2007). Dunes within the
region include linear and parabolic types, with large barchan dunes found occasionally. Locally
at Zhongba, most dunes are parabolic with blown-out central depressions. An extensive system
of small, freshwater wetlands and ponds form in these depressions where the shallow water table
intersects the low topography (Fig. C4b). The interdune wetlands support abundant algae and
aquatic plants, and freshwater semi- and fully aquatic gastropods (Pupillidae sp., Succineidae
sp., Radix sp., Gyraulus sp.). The wetlands site at Tsangpo 18-2 is in an identical geological
context, but is located in another tributary valley to the Yarlung Tsangpo ~90 km west of
Zhongba (Fig. C1).
Accurately characterizing the modern climate for Zhongba (29.67° N, 84.17° E, 4570 m
asl) and Tsangpo 18-2 (30.11° N, 83.39° E, 4601 m asl) is difficult because of the remote
location. The nearest meteorological station to Zhongba is at Gerze (32.15° N, 84.42° E, 4415 m
asl), 288 km away (Fig. C1). Despite the distance, mean annual temperature at Gerze, 0.1°C, is
probably comparable to Zhongba because it is near the same elevation.
Because of the extreme precipitation gradient across the Gangdise Mountains,
precipitation at Gerze is less than in the Yarlung Tsangpo Valley. We use the Tropical Rainfall
Measuring Mission (TRMM) gridded precipitation data to estimate Zhongba precipitation
instead (Kummerow et al., 1997). Mean annual precipitation for the grid square (0.25°x0.25°)
106
containing the Zhongba site for the period 1998-2010 was 304 mm/yr, ranging between 133
mm/yr and 504 mm/yr. Precipitation at Tsangpo 18-2 for the same period averaged 215 mm/yr
ranging between 103 mm/yr and 401 mm/yr.
Widespread in the interdune depressions are ubiquitous organic- and gastropod-rich
deposits 1-2 m above the modern water levels. Archaeological remains are found on the surface
and within these deposits. We focus here on two specific interdune depressions (sites Zhongba
10-1 and 10-9) where artifacts were found in situ and on the surface. We also focus on three
additional modern wetland locations (sites Zhongba 10-4, 10-6, and 10-10) that were
investigated to assess the 14C reservoir effect for the wetlands system (Fig. C1).
The Tsangpo 18-2 site has no in situ archaeological materials but provides additional 14C
dating control on wetter climate intervals, which formed the ubiquitous paleo-wetland deposits
found along the western Yarlung Tsangpo valley in contexts identical to Zhongba. The site
consists of one well-exposed section of wetland strata subsampled for dating (Fig. C2), and
surface lag concentrations of archaeological materials that are typologically similar to those at
Zhongba.
3.4 SEDIMENTOLOGY AND APPEARANCE OF DEPOSITS
The paleo-wetland sediments are clearly distinguished within the dune deposits at all
locations. They generally consist of well-sorted, medium to coarse sand that commonly shows
strongly contorted bedding (Fig. C2). The sands vary from pale green to orange colored due to
extensive and vivid Fe-oxidation, especially along modern root traces. Individual sands grains
are all frosted. The sediments are variably calcareous, grading from slightly consolidated sands
to distinct, interbedded horizons of highly calcareous silty marl. Dark gray to black organic-rich
silt and sand is also commonly interbedded, and tends to be contorted like the sands (Fig. C2).
These sedimentary features are characteristic of inundation of dune sands by an elevated
water table. Groundwater with dissolved carbonate precipitates due to evaporative enrichment to
form calcareous sediments. Organic-rich deposits used for 14C dating are derived from
successive growth and burial of algae and other aquatic and vascular plant material with dune
migration. Contorted bedding most likely results from seasonal freezing (cryoturbation). In
contrast, typical eolian deposits are well-sorted, massive, non-calcareous medium to coarse
107
sands. They tend to have very little organic matter, except for modern root bioturbation, and few
shells or shell fragments.
Gastropods, also used for dating, are common in the wetland sands at all localities,
including tiny land snails (family Pupillidae), semi-aquatic snails (family Succineidae), and fully
aquatic snails (Radix sp., Gyraulus sp.). The fully aquatic species in the middle of the
paleowetland deposits give way to mostly terrestrial species in the top of the deposits.
3.5 METHODS
3.5.1 Site Identification And Artifact Sampling Methods
The Zhongba localities were initially identified by targeting likely areas for associated
wetlands deposits and exposed sedimentary sequences on remote sensing imagery. Preliminary
ground reconnaissance conducted in 2010 confirmed the presence of surface lag aggregates
including both stone and ceramic artifacts, the distribution and extent of which were investigated
by pedestrian survey and recording surface deposits of artifacts without excavation. Artifacts
already exposed in situ in vertical exposures or uncovered in the process of recording profiles
were studied in situ and left in place. Systematic collections of artifacts and débitage in lag
deposits on the sandy surfaces surrounding remnant yardang features were made at locations
throughout the wetlands and dune-field topography.
3.5.2 Dating Methods
All 14C samples were collected in situ from stratigraphic sections or from obvious
modern contexts in the case of the reservoir effect assessment samples. These included organic
and carbonate-rich sediment (marl), living aquatic plant material and plant macrofossils, and
living and fossil mollusk shells. Samples from stratigraphic sections were collected in the field
from measured depths. Modern samples were collected from within or directly adjacent to the
existing interdune ponds. Living mollusks were submerged in ethanol in 15 mL centrifuge tubes,
then rinsed, dried, and sealed for shipping. All sample locations and stratigraphic sections were
described and photographed in the field.
At the University of Arizona, all samples were initially dried overnight in a 70°C drying
oven. Organic 14C samples were subjected to a standard acid-base-acid (ABA) treatment with
108
HCl and NaOH. Shell samples were rinsed ultrasonically in deionized water to remove detritus
and placed in a 3% H2O2 solution to remove organic conchiolin. Carbonate sediment samples
were not pretreated chemically, but special care was taken when sampling to exclude possible
contamination from modern roots and other detritus.
For organic samples, each pretreated sample was combusted under vacuum with CuO
powder and Ag foil. For shell and carbonate sediment, each pretreated sample was completely
dissolved under vacuum in 100% H3PO4. Sample CO2 gas for all samples was extracted under
vacuum and cryogenically purified, and passed through a 600°C Cu/Ag furnace to remove
contaminant gases. Purified CO2 samples were then graphitized using Zn and Fe powder. AMS
measurements were conducted by the Arizona AMS Laboratory. Raw 14C dates were calibrated
using Calib 6.0 software with the IntCal09 calibration curve (Reimer et al., 2009).
3.6 RESULTS
The results of our study are divided into two sections: we first detail the 14C
determinations from modern materials. Against this backdrop, we present the stratigraphy and
dating of Zhongba 10-1, Zhongba 10-9, and Tsangpo 18-2.
3.6.1 14C Dating of the Modern System
We sampled three modern localities, measuring the apparent 14C ages of five samples
(Table C1). Two 14C dates each (four total) were performed on modern materials collected in
identical contexts from interdune ponds at both Zhongba 10-4 and 10-6. The first was on living
aquatic plant material collected from the margin of each pond. The second was on a single
carbonate shell of the aquatic molluscan taxon, Radix sp. indet. (Taft et al., 2012), collected from
the shore of the modern ponds and, though not living, were clearly modern. For Zhongba 10-4,
the 14C age for the aquatic plant was 1360±20 14C yrs BP (AA94393). The 14C age for the shell
was 1150±20 14C yrs BP (AA95307). For Zhongba 10-6, the aquatic plant material (AA99888),
and the Radix shell fragment (AA93968) had a fraction modern carbon (FMC) of 1.027 and
1.037, respectively; that is, post-nuclear testing, and are therefore post-AD 1950 in apparent age.
109
One final sample, a living Radix, was 14C dated from the Zhongba 10-10 site. This
sample was taken from the modern pond in the interdune depression containing the Zhongba 109 stratigraphic sections. The 14C age was determined to be 1370±30 14C yrs BP (AA99889).
Of the five samples dated, two have modern ages and three have apparent ages of >1000
14
C yrs. Within each site, dates for both organics and shell compare closely (Table C1).
Therefore, it appears that each interdune depression has a spatially specific, but consistent age
offset for modern materials.
3.6.2 Zhongba 10-1 Stratigraphy and Dating
The stratigraphic section measured at Zhongba 10-1 is broken into three related parts
around the yardang in the center of the depression (Fig. C3, Fig. C4a). Five 14C dates were
measured (Table C2) and six lithic artifacts were recovered in situ in the sections (Table C3).
Section Zhongba 10-1a exposes the base of the site stratigraphic sequence. The lowest exposed
organic-rich horizon was dated at 8,790±40 14C yrs BP (AA96594). Directly above lies 50 cm of
eolian sand topped by a second organic-rich wetland horizon. Bulk organic material from the top
of this unit was dated at 5800±40 14C yrs BP (AA95308). Above this is a calcareous sand unit
with dispersed gastropod shells; one Radix shell was dated at 7410±50 14C yrs BP (AA93967).
A bulk sediment date on marl at the top of the unit was 9,130±50 14C yrs BP (AA96592). A
wedge-shaped microblade core and an unmodified blade-like flake were recovered within the
upper 5 cm of this marl (Fig. C4d).
The base of section Zhongba 10-1b contains mostly gastropod shell fragments, but one
articulated bivalve shell (Pisidium sp.) was 14C dated at 2550±40 14C yrs BP (Fig. 3, AA95306).
Above this unit there is 35 cm of sandy marl topped by an organic wetland horizon. The top of
section 10-1b consists of >1 m of eolian sand. Though not dated, section Zhongba 10-1c can be
stratigraphically correlated with both 10-1a and 10-1b. Four lithic artifacts were found in situ at
the base of the marl layer in section 10-1c (Fig. C4c).
3.6.3 Zhongba 10-9 Stratigraphy and Dating
Zhongba 10-9 consists of two correlative stratigraphic sections. The full stratigraphy of
the interdune depression is described in Zhongba 10-9a, which can be correlated to the short
section at Zhongba 10-9b (Fig. C5a, b). The base of section 10-9a is composed of eolian sand
110
that is overlain by a 10 cm organic wetland horizon (Fig. C3). The bulk organic material was 14C
dated at 2050±30 14C yrs BP (AA95305). The upper section returns to eolian sand.
Zhongba 10-9b is visibly continuous with the 10 cm organic-rich horizon in the middle of
section 10-9a (Fig. C3). The lower 2 cm of organic-rich sand was 14C dated at 3180±40 14C yrs
BP (AA95304). The upper 8 cm of the section consists of contorted marl, where one unmodified
blade-like flake was collected in situ (Fig. C5b). The marl sediment was 14C dated at 1390±30
14
C yrs BP (AA96593). More lithic artifacts associated with occasional ceramic sherds are
dispersed on the paleo-wetland surface (Fig. C6; Table C3).
3.6.4 Tsangpo 18-2 Stratigraphy and Dating
The stratigraphy of Tsangpo 18-2 consists of contorted eolian sand, overlain by an
interbedded marl horizon (Fig. C3). This is overlain by an organic wetland horizon similar to
those at Zhongba, which is overlain by a final marl horizon at the top of the section. Two dates
in the lower marl and organic-rich parts of the section have ages of 7070±50 (AA82152) and
4710±40 14C yrs BP (AA82160), respectively (Fig. C3; Table C2). Two shell dates from the
upper marl horizon at Tsangpo 18-2 give ages of 2570±30 (AA82153) and 2700±30 14C yrs BP
(AA82154).
3.7 DISCUSSION
3.7.1 Reservoir Correction
The old ages for the modern material at Zhongba highlight the difficulty of dating
artifacts in wetlands environments indirectly using natural materials. Obtaining good age control
relies on understanding the materials being dated and the stratigraphic context of the dates with
respect to the artifacts, otherwise age control for human occupation may be inconclusive (e.g.
Haynes et al., 2013). In wetlands situations, reservoir effects are generally manageable as long
as the archaeological site is well downstream from the groundwater discharge point, giving
enough time for dissolved carbon to equilibrate with the atmosphere at the time of deposition
(e.g. Copeland et al., 2012), however at Zhongba, where the groundwater discharge is direct
from the water table in the ponds this problem is not surprising. We interpret apparent old ages
on modern material to be a result of the familiar 14C reservoir effect, an effect common in many
111
spring, lake, and marine settings (Deevey et al., 1954). In the case of Zhongba, dissolved carbon
species in groundwater systems can be isolated from exchange with the atmosphere for long
periods, or may contain 14C-deficient carbon from dissolution of bedrock, which is then
incorporated into the organic material and carbonate of aquatic plants and mollusks prior to
deposition (Pigati et al., 2004). These 14C ‘reservoir effects’ cause measured 14C ages to appear
older than the true age of deposition. In general, the modern sample age offset should form the
basis for a reservoir correction applied to stratigraphic dates, but the variable offsets measured
for modern Zhongba materials make this difficult.
Zhongba 10-10 is located in the same wetland as the Zhongba 10-9 archaeological site, so
it should provide an accurate reservoir age for that site. However, if we assume the dates from
Zhongba 10-9 (Table C2) are terminus post quem ages for the time of occupation, by up to 1300
14
C yrs, the ages obtained for the fossil materials in section 10-9b would indicate an essentially
modern minimum limiting age for the deposits, which is unlikely.
Assigning a reservoir correction to Zhongba 10-1 is also problematic because there is no
extant pond, and therefore no modern samples from the depression where it is located. Results
from Zhongba 10-4 and 10-6, both nearby, indicate a significantly different correction could be
applied. Thus, in light of the variable effects, for the purpose of this study we choose to report
the calibrated radiocarbon ages for the stratigraphic samples with no reservoir correction, but the
caveat that the ages must be treated as maximum limiting ranges on the period of occupation.
The 14C dates obtained from separate paleo-wetland systems at both Zhongba and Tsangpo 18-2
are very similar, providing additional confidence that the dates obtained at Zhongba do not suffer
from a large reservoir effect, and that our age interpretation is justified.
3.7.2 Site Occupation Intervals
For both Zhongba 10-1 and 10-9 we interpret the ages for the artifact-bearing horizons
based on the limiting dates in the deposits. For Zhongba 10-1, samples 10-1-1, 10-1-2 and 10-13 are below the lithic horizon, providing maximum limiting dates. 10-1-4 and 10-1-5 provide
minimum limiting dates (Fig. C3, Table C2). Three dates are in stratigraphic order, but samples
10-1-3 and 10-1-4 show large stratigraphic age inversions (Fig. C3). Radix shells are robust
enough to survive eolian transport, so for 10-1-3, the age inversion likely results from the
reworking of Radix from older deposits. However, a reservoir effect may also explain the
112
inversion. The marl dated in 10-1-4 formed in situ, so a significant reservoir effect must be
present, as there are three younger dates lower in the stratigraphic profile at this site. We
therefore exclude both samples 10-1-3 and 10-1-4 in constraining the site age and favor the age
of 10-1-2 of 6492-6678 cal. yr B.P. as the best terminus post quem for the artifacts at Zhongba
10-1. Sample 10-1-5, at 2489-2754 cal. yr B.P., provides the only minimum limiting age
constraint. It is unlikely that the bivalve shell dated from sample 10-1-5 was reworked because it
was fully articulated when sampled. We therefore suggest that the age of the in situ artifacts can
be bracketed between 6600-2600 cal. yr B.P., based on the calibrated ranges for the 14C dates
(Table C2).
The age of the in situ flake found at Zhongba 10-9 is constrained by three dates. Sample
10-9b-3, at 3357-3454 cal. yr B.P., is located below the artifact in the organic deposit, and is
therefore a maximum limiting date (Fig. C). Sample 10-9-1 is stratigraphically correlated with
the middle of section 10-9b and so the date of 1930-2071 cal. yr B.P. is closely contemporaneous
in age with the flake. The marl date of 1278-1347 cal. yr BP of 10-9b-4 is a minimum limiting
age (Table C2; Fig. C3). Here, the dating results show no stratigraphic inversions, and therefore
we suggest that the age of deposition for this artifact, and the associated surface collection, falls
in the interval 3400-1300 cal yr B.P. (Table C2).
The dates obtained for Tsangpo 18-2 fall into similar time intervals as those at Zhongba.
The calibrated ages for Tsangpo 18-2a at 7792-7981 cal. yr B.P., and 18-2c at 5321-5582 cal. yr
B.P., place the older sediments in the interval 8000-5400 cal. yr B.P. This time period is similar
to the reworked gastropod shell (Zhongba 10-1-3) and organic matter (Zhongba 10-1-2) dates
from the older Zhongba sediments, implying wetter than modern conditions prevailed in both
locations during the early and middle Holocene (Fig. C3, Table C2). The age of the upper
interval at Tsangpo 18-2 is dated to ~2700 cal. yr B.P. by the two indistinguishable shell dates
(samples 18-2d: 2511-2757 cal. yr B.P. and 18-2e: 2755-2853 cal. yr B.P.), which is similar in
age to the upper shell date at Zhongba 10-1, and falls within the dated interval bracketing the
wetland sediments at Zhongba 10-9. The agreement between wetland intervals at both separate
sites is consistent with the interpretation of regional wet periods during the Holocene, and
provides additional confidence in the ages for the Zhongba artifacts.
113
3.7.3 Zhongba Artifact Assemblages and Site Function
The Zhongba stone artifacts are made on several types of raw material including jasper,
chert, and chalcedony (Fig. C6). Bedrock with similar composition, included in highly folded
Cretaceous tectonic melange (Fig. C5c; Ding, Kapp, & Wan, 2005), crops out locally
approximately two kilometers south of the artifact-bearing localities.
The lithic artifact assemblages from Zhongba 10-1 and 10-9 are broadly similar. Both are
dominated by non-geometric microliths. At both localities, conical and wedge-shaped
microblade cores constitute the primary technological source of modified lithic elements (Fig.
C6; Table C2). Finished retouched tools comprise only a small fraction of the total assemblages
and most artifacts were recovered from the surface rather than in situ. Nonetheless, sample sizes
are large enough to allow some broad interpretations. The non-geometric character of the
microlithic assemblages from both Zhongba 10-1 and 10-9 is of interest since this fact, along
with the typologically consistent conical and wedge-shaped microblade cores from which the
majority of retouched pieces was derived, suggests that these southern Tibetan assemblages
share greater similarities with their counterparts to the north and east than with contemporaneous
technologies from south of the Himalayas.
Typologically distinct nuclei from which larger blade-like flakes were derived were not
recovered from either buried or surface contexts at the Zhongba localities, leaving open the
possibility that they were produced off-site, near raw material sources, and subsequently
transported to the immediate vicinity of the Zhongba localities for further reduction.
A small collection of only 26 ceramic sherds was recovered from the Zhongba 10-9
locality only (Fig. C6). These sherds derive exclusively from small containers rather than objets
d’art, beads, etc. and are all relatively thick (ca. 3-5 mm) with a gritty sand temper. Only two
rim-sherds were recovered, from two different vessels, both suggestive of small (ca. 15-20 cm
diameter) open bowls. The relatively well-preserved exterior surfaces of these sherds are
unadorned, either by painting or incision/impression, presumably indicative of a utilitarian ware.
Behaviorally, we interpret the Zhongba sites as workshops for the production of
microblades and finished lithic tools. The presence of conical and wedge-shaped microblade
cores, unaltered blades and blade-like flakes, and finished tools within the assemblages indicates
that most manufacturing took place on-site. The proximity of the raw material sources for all but
a tiny fraction of the Zhongba lithic assemblage argues against long-distance transport of raw
114
material to the locations where tools were recovered in situ. At the same time, the presence of
débitage – including micro-débitage – in the overall assemblage indicates on-site manufacturing
took place at or near the find-spots of the recovered artifacts. We do not use the term
“workshop” to indicate that only the fabrication of lithic artifacts took place at Zhongba; rather,
that the proximity of relatively high-quality raw material may have been sufficient inducement
for people to settle – or at least pause – in the area long enough to take advantage of that
resource. The lack of bone and other organics, including charcoal, as well as the weathered and
eroded character of the ceramic assemblage, may indicate taphonomic circumstances that
precluded the preservation of a more representative range of material culture. On the other hand,
the recovery of micro-débitage and both wedge-shaped and conical microblade cores at various
stages of preparation and reduction suggests that on-site manufacture of lithics sourced
principally from locally available raw materials was an important constituent part of activities
undertaken at Zhongba. The interdune depressions present an attractive location for occupation
because of access to water, and because the enclosed topography provides excellent protection
from the wind.
3.7.4 Regional Significance
In the absence of faunal remains and only limited ceramic remains, lithic artifact
typology and the timing of occupation are the primary characteristics available for comparison
with other nearby high altitude microlithic sites. We place the earliest occupation at Zhongba in
context with the Holocene microlithic sites to discuss the origins of plateau colonization, and
then discuss the implications of the later occupation at Zhongba in relation to contemporaneous
Neolithic and metal-bearing sites in Tibet.
Earlier Occupation
Sites on the Qinghai-Tibet Plateau with good age control are scarce, but several sites with
similar artifact assemblages have similar or older ages than the Zhongba time interval (Fig. C7).
Sites located in the Qinghai Lake basin, Heimahe and Jiangxigou, have the oldest periods of
occupation, attributed to Late Pleistocene hunter/gatherers (Madsen et al., 2006). Though a few
microliths are reported for this period, larger flakes, both modified and unmodified, comprise the
majority of the artifact assemblages. The younger occupation periods at these sites are
115
associated with early Neolithic inhabitants utilizing pottery and abundant microliths similar to
those recovered at Zhongba (Rhode et al., 2006). Both occupations are older than the mid to late
Holocene age of artifacts at our sites.
An additional locality located in the Burhan Budai sub-range of the Kunlun Mountains,
Xidatan, has yielded a similar buried artifact assemblage dated to ca. 6.5 ka (Brantingham &
Gao, 2006) and many more microlithic assemblages from the Chang Tang Nature Reserve
(Brantingham, Olsen, & Schaller, 2001) and as far south as Seling Tso (An, 1982) are assigned
to the Holocene (Fig. C7). Obsidian provenance of microliths recovered from the Chang Tang
sites, and in Neolithic contexts at Jiangxigou 2 and Xidatan, indicates a single geologic source
located near the lake Migriggyangzham Tso, which supports the Holocene age assignment
(Brantingham & Gao, 2006). Thus, all available evidence indicates that by the mid-Holocene
people employing microlithic technology achieved permanent and widespread occupation of the
high plateau. The nucleus typology described in the Chang Tang sites, namely Levallois-like
flake cores and conical and wedge-shaped microblade cores, is very similar to that found at
Zhongba, indicating the microlithic industry spanning the eastern plateau may extend as far west
as Zhongba. The typological similarities and later age of the Zhongba sites are compatible with
the published three-step colonization model of Brantingham et al. (2003, 2006). This link
requires further investigation, however, given the inconsistent and incomplete nature of the
surface sampling at Zhongba and elsewhere in Tibet, from a statistical viewpoint.
Osh Khona, located at >4000 m asl in the Markansu basin of the Pamir Mountains, is
another high altitude microlithic site of note (Fig. C7). The calibrated 14C dates for the site span
10.8-7.8 ka, putting the age similar to the oldest microlithic sites in Qinghai, and older than
Zhongba (Ranov et al., 1979). Although the site contains microliths, the majority of artifacts are
instead pebble tools and choppers, which is in stark contrast to the assemblages noted from
Tibet, including Zhongba. Pebble tools in inferred Holocene contexts are known in the high
Hindu Kush, indicating colonization of the plateau and its high periphery may not be restricted to
a single eastern vector (Malasse & Gaillard, 2011). The Zhongba assemblage bears more
similarity to sites to the east than the west, and so the geographic boundary of this pebble tool
technocomplex is likely west of Zhongba.
116
Later Occupation and Increasing Complexity
The age range limiting the time of deposition of artifacts at Zhongba 10-9, 3.4-1.3 ka, is
very young in comparison to the previously mentioned sites, and is broadly contemporary with
the Neolithic and Bronze Age occupation of the Lhasa area (e.g. Qugong), and the nearby village
and mortuary complex of Dindun (Fig. C7) west of Zhongba (Aldenderfer & Zhang, 2004).
Dindun contains both ground stone tools and bronze and iron artifacts, and is tenuously
connected in age by the authors to the protohistoric Zhang Zhung polity. The exclusively
microlithic and ceramic artifact assemblage of similar age at Zhongba indicates that microlithic
technology was still being utilized in the region despite clear evidence of metal in occupation
and mortuary contexts at the time. This circumstance indicates the pre-Buddhist history of
western Tibet is exceedingly complex and requires further investigation, particularly with respect
to the timing and directionality of initial and later human incursions onto the plateau.
3.8 CONCLUSIONS
The 14C dating of the Zhongba microlithic sites highlights the importance of
understanding the materials being dated in order to establish a realistic site chronology. In this
case, investigating the modern materials from the wetlands revealed variable reservoir effects,
indicating caution is required when interpreting the ages of in situ artifacts in similar
paleoenvironments. Good agreement between ages from Zhongba and Tsangpo 18-2 suggests
that reporting the calibrated 14C ages with no reservoir effect is valid, but the occupation
intervals we report for Zhongba still represent terminus post quem constraints. In any case, our
suggested occupation intervals at Zhongba 10-1 and 10-9, at 6600-2600 cal. yr B.P. and 34001300 cal. yr B.P., respectively, can be no older than the true ages, despite the reservoir effect
considerations. The occupation intervals for the site show that microlithic technology remained
important during the late Holocene in west Tibet despite increasing aridity and ecological
complexity. The typology of artifacts recovered at Zhongba is similar to that of known
assemblages from the Chang Tang eastward, and the earliest Zhongba occupation must postdate
the ages of those assemblages. The later period of occupation, coincident with metal-bearing
sites on the plateau and in the immediate vicinity, emphasizes the need for improved site
117
chronologies to understand the development of late prehistoric societies in western Tibet in
relation to lowland regions surrounding the plateau.
3.9 ACKNOWLEDGEMENTS
This research was funded by grants from the Henry Luce Foundation, the American
Council of Learned Societies, the Comer Science and Education Foundation, and the University
of Arizona Je Tsongkhapa Endowment for Central and Inner Asian Archaeology. We thank
Nyingcha-gyal, Lei Guoliang, John Pelletier, Tyler Huth and Erin Clair for their invaluable
assistance in the field. We also thank the editor and two anonymous reviewers whose comments
greatly improved this work.
3.10 REFERENCES
Aldenderfer, M. (2011). Peopling the Tibetan Plateau: insights from archaeology. High Altitude
Medicine & Biology, 12, 141-147.
Aldenderfer, M. & Zhang Y. (2004). The prehistory of the Tibetan Plateau to the seventh century
A.D.: perspectives and research from China and west since 1950. Journal of World
Prehistory, 18, 1-55.
An Z. (1982). Palaeoliths and microliths from Shena and Shuanghu, northern Tibet. Current
Anthropology, 23, 493-499.
Bellezza, J.V. (2008). Zhang Zhung, Foundations of Civilization in Tibet, A Historical and
Ethnoarchaeological Study of the Monuments, Rock Art, Texts, and Oral Tradition of the
Ancient Tibetan Upland. Denkschriften der phil.-hist. Klasse 368. Beitraege zur Kultur- und
Geistesgeschichte Asiens 61, Verlag der Oesterreichischen Akademie der Wissenschaften,
Wien.
118
Bookhagen, B. (2010). Appearance of extreme monsoonal rainfall events and their impact on
erosion in the Himalaya. Geomatics, Natural Hazards and Risk, 1, 37-50.
Brantingham, P.J., Olsen, J.W., & Schaller, G. (2001). Lithic assemblages from the Chang Tang
region, northern Tibet. Antiquity, 75, 319-327.
Brantingham, P.J., Ma H., Olsen, J.W., Gao X., Madsen, D.B., & Rhode, D.E. (2003).
Speculation on the timing and nature of Late Pleistocene hunter-gatherer colonization of the
Tibetan Plateau, Chinese Science Bulletin, 48, 1510-1516.
Brantingham, P.J., & Gao X. (2006). Peopling of the northern Tibetan Plateau. World
Archaeology, 38, 387-414.
Copeland, A., Quade, J., Watson, J.T., McLaurin, B.T., & Villalpando, E. (2012). Stratigraphy
and geochronology of La Playa archaeological site, Sonora, Mexico. Journal of
Archaeological Science, 39, 2934-2944.
Deevey, E.S., Gross, M.S., Hutchinson, G.E., & Kraybill, H.L. (1954). The natural 14C contents
of materials from hard-water lakes. Proceedings of the National Academy of Sciences, 40,
285-288.
Demske, D., Tarasov, P., Wünnemann, B., & Riedel, F. (2009). Late glacial and Holocene
vegetation, Indian monsoon and westerly circulation in the Trans-Himalaya recorded in the
lacustrine pollen sequence from Tso Kar, Ladakh, NW India. Palaeogeography,
Palaeoclimatology, Palaeoecology, 279, 172-185.
Ding L., Kapp, P., & Wan X. (2005). Paleocene-Eocene record of ophiolite obduction and initial
India-Asia collision, south central Tibet. Tectonics, 24, doi:10.1029/2004TC001729.
Haynes, C.V., Surovell, T.A., & Hodgins, G.W.L. (2013). The U.P. Mammoth Site, Carbon
County, Wyoming, USA: More questions than answers. Geoarchaeology, 28, 99-111.
119
Herzschuh, U., Winter, K., Wunneman, B., & Li S. (2006). A general cooling trend on the
central Tibetan Plateau throughout the Holocene recorded by the Lake Zigetang pollen
spectra. Quaternary International, 154-155, 113-121.
Hudson, A.M., & Quade, J. (2013). Long term east-west asymmetry in monsoon climate on the
Tibetan Plateau. Geology, 41, 351-354.
Li Q., Lu H., Zhu L., Wu N., Wang J., & Lu X. (2011). Pollen-inferred climate changes and
vertical shifts of alpine vegetation belts on the northern slope of the Nyainqentanglha
Mountains (central Tibetan Plateau) since 8.4 kyr BP. The Holocene, 21, 939-950.
Kramer, A., Herzschuh, U., Mischke, S., & Zhang, C. (2010). Late glacial vegetation and climate
oscillations on the southeastern Tibetan Plateau inferred from the Lake Naleng pollen profile.
Quaternary Research, 73, 324-335.
Kummerow, C., Barnes, W., Kozu, T., Shiue, J., & Simpson, J. (1997). The Tropical Rainfall
Measuring Mission (TRMM) sensor package. Journal of Atmospheric and Oceanic
Technology, 15, 809-817.
Madsen, D.B., Ma H., Brantingham, P.J., Gao X., Rhode, D., Zhang H., & Olsen, J.W. (2006).
The late Upper Paleolithic occupation of the northern Tibetan Plateau margin. Journal of
Archaeological Science, 33, 1433-1444.
Malasse, J.W., & Gaillard, C. (2011). Relations between climatic changes and prehistoric human
migrations during Holocene between Gissar Range, Pamir, Hindu Kush and Kashmir: The
archaeological and ecological data. Quaternary International, 229, 123-131.
Meyer, M.C., Hofmann, Ch.-Ch, Gemmell, A.M.D., Haslinger, E., Hausler, H., & Wangda,
D. (2009). Holocene glacier fluctuations and migration of Neolithic yak pastoralists into the high
valleys of northwest Bhutan. Quaternary Science Reviews, 28, 1217-1237.
120
Miehe, G., Kaiser, K., Co, S., Zhao X., Liu J. (2008). Geo-ecological transect studies in northeast
Tibet (Qinghai, China) reveal human-made mid-Holocene environmental changes in the
upper Yellow River catchment changing forest to grassland. Erkunde, 62, 187-199.
Miehe, G., Miehe, S., & Schlütz, F. (2009). Early human impact in the forest ecotone of southern
High Asia (Hindu Kush, Himalaya). Quaternary Research, 71, 255-265.
Pelletier, J.D., Quade, J., Goble, R.J., & Aldenderfer, M.S. (2011). Widespread hillslope gullying
on the southeastern Tibetan Plateau: human or climate-change induced?. Geological Society
of America Bulletin, 123, 1926-1938.
Pigati, J.S., Quade, J., Shanahan, T.M., & Haynes Jr., C.V. (2004). Radiocarbon dating of minute
gastropods and new constraints on the timing of late Quaternary spring-discharge deposits in
southern Arizona, USA. Palaeogeography, Palaeoclimatology, Palaeoecology, 204, 33-45.
Ranov, V.A., Davis, R.S., Aigner, J.S., Gabori, M., Gallus, A., Marks, A.E., Mohapatra, G.C.,
Movius, H.L., & Zeiler, I.S. (1979). Toward a new outline of the Soviet Central Asian
Paleolithic. Current Anthropology, 20, 249-270.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey,
C., Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P.,
Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac,
F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney,
C.S.M., van der Plicht, J., & Weyhenmeyer, C.E. (2009). INTCAL09 and MARINE09
radiocarbon age calibration curves, 0-50,000 years cal BP. Radiocarbon, 51, 1111-1150.
Rhode, D., Zhang H., Madsen, D.B., Gao X., Brantingham, P.J., Ma H., & Olsen, J.W. (2006).
Epipaleolithic/early Neolithic settlements at Qinghai Lake, western China. Journal of
Archaeological Science, 34, 600-612.
121
Schlütz, F., & Lehmkuhl, F. (2009). Holocene climatic change and the nomadic Anthropocene in
Eastern Tibet: palynological and geomorphological results from the Nianbaoyeze Mountains.
Quaternary Science Reviews, 28, 1449-1471.
Taft, L., Wiechert, U., Riedel, F., Weynell, M., & Zhang H. (2012). Sub-seasonal oxygen and
carbon isotope variations in shells of modern Radix sp. (Gastropoda) from the Tibetan
Plateau: potential of a new archive for palaeoclimatic studies. Quaternary Science Reviews,
34, 44-56.
Tian, L., Yao, T., MacClune, K., White, J.W.C., Schilla, A., Vaughn, B., Vachon, R., and
Ichiyanagi, K. (2007). Stable isotopic variations in west China: a consideration of moisture
sources. Journal of Geophysical Research, 112, D10112, doi:10.1029/2006JD007718.
Wischnewski, J., Mischke, S., Wang Y., & Herzschuh, U. (2011). Reconstructing climate
variability on the northeastern Tibetan Plateau since the last Lateglacial – a multi-proxy,
dual-site approach comparing terrestrial and aquatic signals. Quaternary Science Reviews,
30, 82-97.
Wünnemann, B., Demske, D., Tarasov, P., Kotlia, B.S., Reinhardt, C., Bloemendal, J.,
Diekmann, B., Hartmann, K., Krois, J., Riedel, F., & Arya, N. (2010). Hydrological
evolution during the last 15 kyr in the Tso Kar lake basin (Ladakh, India), derived from
geomorphological, sedimentological and palynological records. Quaternary Science Reviews,
29, 1138-1155.
Yu G., Tang L., Yang X., & Ke X. (2001). Modern pollen samples from alpine vegetation on the
Tibetan Plateau. Global Ecology & Biogeography, 10, 503-519.
Zhang, D.D., & Li S.H. (2002). Optical dating of Tibetan human hand- and footprints: an
implication for the palaeoenvironment of the last glaciation of the Tibetan Plateau.
Geophysical Research Letters, 29.
122
Table C1. AMS radiocarbon dates from modern Zhongba sites
14
Sample Site
Zhongba 10-4
Zhongba 10-6
Zhongba 10-10
C age ±uncert.
Sample Material
Sample ID
Lab Number
FMC*
10-4-1
AA94393
0.844±0.002
1360±20
aquatic plant
10-4-2
AA95307
0.867±0.003
1150±20
Radix sp. shell
10-6a
AA99888
1.027±0.004
modern
aquatic plant
10-6b
AA93968
1.037±0.003
modern
Radix sp. shell
10-10a
AA99889
0.843±0.003
1370±30
Radix sp. shell
(14C yr BP)
*Fraction Modern Carbon, where 'modern' refers to the 14C content of the atmosphere in AD 1950
128
123
129
124
AA82152
AA82160
AA82153
AA82154
18-2a
18-2c
18-2d
18-2e
14
2700±30
2570±30
4710±40
7070±50
1390±30
3180±30
2050±30
2550±40
9130±50
7410±50
5800±40
8790±50
( C yr BP)
14
C age ±uncert.
2755-2853
2511-2757
5321-5582
7792-7981
1278-1347
3357-3454
1930-2071
2489-2754
10204-10422
8158-8364
6492-6678
9601-10147
range (cal yr BP)
a
95% Calibrated age
1
1
1
1
0.99
0.99
0.99
1
0.98
0.97
0.97
1
areab
Relative
2800±50
2630±120
5450±130
7890±100
1310±40
3410±50
2000±70
2620±110
10310±110
8260±100
6590±90
9870±270
Median age
(cal yr BP)
This age range represents the 95% confidence interval for the IntCal09 radiocarbon curve (Reimer et al., 2009)
AA96593
AA95305
10-9-1
10-9b-4
AA95306
10-1-5
AA95304
AA96592
10-9b-3
AA93967
AA95308
10-1-2
10-1-3
AA96594
10-1-1
10-1-4
Lab Number
Sample ID
Succineidae sp. shell
Puppilae sp. shell
bulk organic sediment
Radix sp. shell
carbonate sediment (marl)
bulk organic sediment
bulk organic sediment
articulated bivalve shell
carbonate sediment (marl)
Radix sp. shell
bulk organic sediment
fibrous organic material
Sample Material
This represents the area under the probability density function for each calibrated date within the age range. This emphasizes the age range of
maximum probability and eliminates tails resulting from calibration uncertainty. Maximum is 1.0. No areas less than 0.97 are reported.
b
a
Tsangpo 18-2
Zhongba 10-9
Zhongba 10-1
Sample Site
Table C2. AMS radiocarbon dates from Zhongba and Tsangpo stratigraphic sections
Table C3. Number and type of artifacts from the Zhongba archaeological site
Site Name
Collection Context
In Situ
Zhongba 10-1
Surface
Total
In Situ
Zhongba 10-9
Surface
Artifact Type
Number (n)
microblade core, wedge-shaped
1
blade-like flake, unmodified
5
microblade core, conical
7
microblade core, wedge-shaped
9
perçoir/borer
2
thumbnail scraper
3
biface
1
blade-like flake, unmodified
12
microblade, unmodified
6
--
46
blade-like flake, unmodified
1
microblade core, conical
3
microblade core, wedge-shaped
9
perçoir/borer
1
thumbnail scraper
1
waisted blade
3
burin
2
blade-like flake, unmodified
24
microblade, unmodified
33
ceramic, body sherd
24
ceramic, rim sherd
2
Total
--
103
Grand Total
--
149
130
125
80°E
90°E
100°E
36°N
70°E
Heimahe
Qinghai Lake
Jiangxigou
36°N
Hindu
Kush
Xidatan
Chang Tang
Region
30°N
Gerze
Ga
Karou
ng
dis
Chamdo
s.
Lhasa
u
Kyich
Yarlung Tsangpo
Hima
laya
s
Qugong
24°N
Zhongba
tn
30°N
Tsangpo 18-2
eM
80°E
90°E
100°E
Figure C1. Overview map of the Qinghai-Tibet Plateau. Study sites and key localities referenced
in the text are labeled. Yarlung Tsangpo and Kyichu rivers are approximated by the thick dotted
line and Qinghai Lake by the gray polygon. Modern towns are indicated by open circles.
131
126
Eolian dune sand
Organic-rich sand
Fe-stained sand
Contorted beds
Marl
Figure C2. Tsangpo 18-2 example stratigraphy showing paleo-wetland deposit characteristics.
Trowel for scale is 25 cm in length.
132
127
Legend
Zhongba10-1 sections
Massive sand
Weak planar bedding
Cryoturbated sand
Marl (carbonate sed.)
10-1b
10-1a
Organic-rich horizon
200
200
Modern root bioturb.
Lithic artifact
Shell
150
150
Height (cm)
10-1c
150
10,060±73
cal yrs BP
100
100
100
8261±103
cal yrs BP 50
6581±95
cal yrs BP
9504±123
50
clay
0
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
clay
0
clay
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
0
2622±132
cal yrs BP
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
50
10-9a
200
Zhongba10-9
sections
Tsangpo 18-2
150
150
2634±123,
2804±49
cal yrs BP
100
5452±130
cal yrs BP
7887±95
cal yrs BP
50
10-9b
100
10
1154±58
cal yrs BP
2022±100
cal yrs BP
50
3401±69
cal yrs BP
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
0
clay
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
clay
0
si
fine lt
sand
med
. san
d
coar
se s
and
pebb
le
cobb
le
clay
0
Figure C3. Zhongba and Tsangpo stratigraphic sections.
133
128
1m
5 cm
5 cm
Figure C4. Zhongba 10-1 site. a) Yardang (see text) where stratigraphic sections were measured.
b) Typical modern analog interdune wetland. c) Wedge-shaped microblade core in situ in section
10-1c, photo location shown in a). d) Large unmodified anthropogenic blade-like flake in situ
near section 10-1a, photo location shown in a).
134
129
Figure C5. Zhongba 10-9 site. a) Cryoturbated upper wetland deposits, modern wetland pond at
right. b) Unmodified anthropogenic blade-like flake in situ in section 10-9b. c) Unweathered
Cretaceous bedrock outcrop with weathered raw material.
135
130
1
2
6
3
4
7
9
5
8
11
10
14
12
13
Figure C6. Representative lithic and ceramic artifacts from Zhongba. 1-4: flake; 5: blade-like
flake; 6, 9, 12: ceramic sherd; 7, 10, 11, 14: wedge-shaped microblade core; 8, 13: conical
microblade core. Pencil for scale is 14 cm in length.
136
131
70°E
80°E
90°E
100°E
36°N
Osh Khona
7.8-10.8 cal ka
Xidatan
8.2-6.4 cal ka
Central Chang
Thang Sites
Holocene
Dindun
2.5-2.1 cal ka
30°N
Jiangxigou
14.1-14.8 cal ka
9.5-5.0 cal ka
36°N
Heimahe
12.5-13.5 cal ka
8.5 cal ka
Qugong
4.0-2.0 cal ka
30°N
Zhongba
6.6-2.6 cal ka
3.4-1.3 cal ka
Karou
5.0-4.0 cal ka
24°N
Lunana Charcoal
6.7-4.5 cal ka
80°E
90°E
100°E
Figure C7. High altitude microlithic archaeological sites on the Qinghai-Tibet Plateau. Site age
ranges are derived from published 14C ages interpreted for each site. Reference for each site as
follows: Osh Khona – Ranov et al., 1979; Dindun, Qugong, Karou – Aldenderfer & Zhang, 2004;
Lunana Charcoal – Meyer et al., 2009; Central Chang Tang sites – Brantingham, Olsen, &
Schaller, 2001; An, 1982; Xidatan – Brantingham & Gao, 2006; Heimahe, Jiangxigou –
Madsen et al., 2006; Rhode et al., 2006; Zhongba – this study.
137
132