Late Holocene Glacier Change in the Eastern Nyainqêntanglha Range, SE Tibet

Late Holocene glacier change in the eastern Nyainqêntanglha Range, SE Tibet

Von der Fakultät für Georessourcen und Materialtechnik der Rheinisch‐Westfälischen Technischen Hochschule Aachen

zur Erlangung des akademischen Grades eines

Doktors der Naturwissenschaften

genehmigte Dissertation

vorgelegt von M.A.

David Loibl

aus Aachen

Berichter: Univ.‐Prof. Dr. rer. nat. Frank Lehmkuhl Univ.‐Prof. Dr. rer. nat. Achim Bräuning

Tag der mündlichen Prüfung: 06. Juli 2015

Diese Dissertation ist auf den Internetseiten der Hochschulbibliothek online verfügbar.

Table of contents

List of figures ...... i

List of tables ...... ii

List of abbreviations ...... iii

1 Introduction ...... 1 1.1 Motivation ...... 1 1.2 Objectives and approach ...... 2 1.3 Thesis outline...... 3

2 Current state of research ...... 5 2.1 Atmospheric circulation patterns ...... 5 2.2 Paleoclimate and glacier fluctuations ...... 7 2.2.1 Paleoclimate reconstruction ...... 7 2.2.2 Holocene paleoclimate on the Tibetan Plateau ...... 10 2.2.3 Glaciers and climate change ...... 13 2.2.4 Late Holocene glacier fluctuations and climatic forcing ...... 14 2.2.5 Human perspective: risks and challenges ...... 15 2.3 Methodological background ...... 15 2.3.1 Remote sensing of glaciers ...... 15 2.3.2 Equilibrium line altitude calculation ...... 17 2.3.3 Dating of Holocene glacier deposits ...... 19 3 Regional setting ...... 22 3.1 Geology ...... 22 3.2 Climate ...... 23 3.3 Geomorphology ...... 25 3.4 Glaciers ...... 26 3.5 Vegetation ...... 27

4 Reconstructing glacier retreat since the Little Ice Age in SE Tibet by glacier mapping and equilibrium line altitude calculation ...... 28 4.1 Introduction ...... 28 4.2 Regional setting and glacier distribution ...... 30 4.3 Methods and materials ...... 33 4.3.1 Data selection ...... 33 4.3.2 Glacier mapping ...... 35 4.3.3 Glacier classification ...... 36 4.3.4 Calculation of ELA estimates ...... 38 4.3.5 Fieldwork ...... 39 4.4 Results ...... 40 4.4.1 Glacier case studies ...... 40 4.4.2 Glacier types and spatial distribution ...... 41 4.4.3 Length change analysis ...... 44 4.4.4 ELA calculations ...... 46 4.5 Discussion ...... 50 4.5.1 Quality assessment ...... 50 4.5.2 Influences of topography and climate ...... 51 4.5.3 Glacier change ...... 52 4.5.4 Monsoon and glaciers ...... 53 4.6 Conclusions ...... 55

5 Glaciers and equilibrium line altitudes of the eastern Nyainqêntanglha Range, SE Tibet ...... 58 5.1 Introduction ...... 58 5.2 Study area ...... 59 5.3 Methods ...... 62 5.3.1 Data and glacier mapping ...... 62 5.3.2 Glacier classification ...... 64 5.3.3 ELA calculation ...... 64 5.3.4 Post‐processing ...... 66 5.4 Map Design ...... 66 5.5 Interpretation ...... 67 5.6 Conclusions ...... 69 5.7 Software ...... 69

6 Toward a late Holocene glacial chronology for the eastern Nyainqêntanglha Range, southeastern Tibet ...... 70 6.1 Introduction ...... 70 6.2 Regional setting ...... 72 6.3 Methods ...... 74 6.3.1 Geomorphological mapping ...... 75 6.3.2 Sedimentological analysis ...... 75 6.3.3 OSL analysis ...... 76 6.3.4 14C analysis ...... 77 6.3.5 Tree sample analysis ...... 77 6.4 Results ...... 78 6.4.1 Proglacial geomorphology ...... 78 6.4.2 Sedimentology ...... 82 6.4.3 Tree ring dating ...... 86 6.4.4 OSL and 14C dating results ...... 86 6.5 Discussion ...... 89 6.5.1 Dating methods ...... 89 6.5.2 Comparison of different glaciers ...... 91 6.5.3 Glacial chronology ...... 92 6.5.4 Glaciers and paleoclimate ...... 95 6.6 Conclusions ...... 97

7 Synthesis ...... 98 7.1 Challenges and opportunities ...... 98 7.2 Glaciers and ELAs: patterns and driving forces of distribution and post‐LIA changes ...... 98 7.3 Proglacial settings and late Holocene moraine sequences ...... 100 7.4 Synoptic conclusions, desiderata and outlook ...... 102

Overall references ...... 105

Appendices ...... 150 Appendix A: Tables ...... 150 Appendix B: Figures ...... 151

Acknowledgements ...... 165

Abstract ...... 167

Kurzfassung ...... 168

List of figures

Fig. 2.1: Circulation patterns over southern and eastern Asia, focusing on TP and bordering mountains.

Fig. 3.1: Topographic overview map of the study area.

Fig. 3.2: Average annual air temperature at 2 m above ground.

Fig. 3.3: Spatial distribution of average annual precipitation.

Fig. 4.1: Study area in southeastern Tibet. Elevations are based on SRTM data.

Fig. 4.2: Climate diagrams for the two stations in the study area, Bomi and Nyingchi.

Fig. 4.3: Late summer precipitation and temperature averages of Bomi and Nyingchi meteorological stations for the period 1960–2004.

Fig. 4.4: Example of glacier outline quality assessment for the Glacier Inventory of China (GIC) data by comparison to a Landsat ETM+ scene (channels 5‐4‐3).

Fig. 4.5: Data sources, interpretation and errors.

Fig. 4.6: Typology of glaciers used in this study.

Fig. 4.7: Map of three selected glaciers and selected glacial and geomorphological features.

Fig. 4.8: Spatial distribution of glacier types for 1999 and LIA.

Fig. 4.9: Average glacier length reduction plotted against different parameters.

Fig. 4.10: Plot of average ridge and terminus altitudes.

Fig. 4.11: Evaluation of ELAs calculated by (a) TSAM and (b) TRAM by comparison to TSL alti‐ tudes mapped from the Landsat ETM+ scene from 23 September 1999.

Fig. 4.12: ELAs calculated for the 1999 stage.

Fig. 4.13: Interpolated differences between local ELA calculations (ΔELA) for the LIA and 1999 stages.

Fig. 4.14: Rose diagrams for the calculated ELAs in different ranges of the eastern Nyainqêntanglha.

Fig. 5.1: Overview of the study area in southeastern Tibet.

Fig. 5.2: The study area in the Landsat ETM+ scene from 23 September 1999 which was used for mapping (channels 5‐4‐3).

Fig. 5.3: Mapping concepts for glaciological and geomorphological features as applied in this study.

Fig. 5.4: Typology of glaciers used in this study. i

Fig. 5.5: Schematic diagram of the three morphometric approaches of ELA calculation.

Fig. 5.6: Interpretation of the spatial distribution patterns of ELAs.

Fig. 6.1: Regional setting of the study sites at Mt. Pulongu and Mt. Zepu Kangri in southeast‐ ern Tibet.

Fig. 6.2: Sites of case studies.

Fig. 6.3: Panorama photos of the studied glacier forelands.

Fig. 6.4: Geomorphological maps of the proglacial areas of the glaciers Xinpu and Gongpu.

Fig. 6.5: Different sample sites at Xinpu and Gongpu.

Fig. 6.6: Grain size distributions for different sediment types at Xinpu and Gongpu.

Fig. 6.7: Geochemical indices plotted for different archives.

Fig. 6.8: Conceptual morphochronology for Xinpu.

Fig. 6.9: Compilation of dating results from this and selected other studies and paleoclimate reconstructions.

Fig. 7.1: Comparison of CGI glacier outlines and glacier outlines mapped in this study

List of tables

Table 1.1: Publications.

Table 4.1: Comparison of numeric glacier type distribution for the LIA and 1999 stages.

Table 6.1: Geochemical indices used in this study and their interpretation.

Table 6.2: OSL dating results.

Table 6.3: Calibrated 14C dating result of tree trunk from major moraine at Xinpu glacier.

List of abbreviations

AAR: Accumulation area ratio AABR: Area‐altitude balance ratio ASTER: Advanced spaceborne thermal emission and reflection radiometer BR: Balance ratio CGI: Chinese glacier inventory DFG: Detusche Forschungsgesellschaft EASM: East Asian summer monsoon ELA: Equilibrium line altitude ENSO: El Niño‐Southern Oscillation GDEM: Global digital elevation model GIC: see CGI GLIMS: Global Land Ice Measurements from Space GLOF: Glacier lake outburst flood gLGM: Global Last Glacial Maximum IDW: Inverse distance weighted ISM: Indian summer monsoon ITCZ: Innertropical convergence zone LIA: Little Ice Age MELM: Maximum elevation of lateral moraines NAO: North Atlantic Oscillation OSL: Optically stimulated luminescence RGI: Randolph Glacier Inventory SRTM: Shuttle radar topography mission THAR: Toe‐to‐headwall altitude ratio TL: Thermoluminescence TP: Tibetan Plateau TRAM: Toe‐to‐ridge altitude method TSAM: Toe‐to‐summit altitude method TSL: Transient snowline

iii

1 Introduction

1.1 Motivation The Tibetan Plateau (TP) is a singular topographic feature, with an average altitude > 5000 m asl, framed by some of the world’s largest mountain belts which include the world’s highest summits (Fielding et al., 1994). Originating from the collision of India and Eurasia (Allégre et al., 1984; Harrison et al., 1992; Tapponier et al., 2001; Kapp and Guyn, 2004), the TP is recognized as a key element of Asia’s climate system (Domrös and Peng, 1988; Anders et al., 2006; Molnar et al., 2010). The TP’s relevance as a controlling factor of the two branches of the Asian summer monsoon, namely the East Asian summer monsoon (EASM) and the Indian summer monsoon (ISM), is considered to be particu‐ larly high (e.g., Zhisheng et al., 2001), even though the physical forcing mechanisms are still under debate (e.g., Flohn, 1953; Li and Yanai, 1996; Boos and Kuang, 2010, 2013). Glaciers on the TP act as long‐term water storage and their meltwater feeds some of the largest rivers in the world, providing water to more than 1.3 billion people (Yao et al., 2007; Immerzeel et al., 2010).

Located at the southeastern fringe of the TP, the eastern Nyainqêntanglha Range is prone to the moisture‐bearing monsoonal air masses. Owing to abundant precipitation and pronounced high mountain topography, this mountain range exhibits c. 8000 km² of glaciated area. The glaciers belong to the ‘monsoonal temperate’ type which is often reported to be highly sensitive to climate change (e.g., Zhou et al., 1991; Yang, B., et al., 2008, Yang, W., 2011, 2013). Recent studies have shown that glacier mass loss in this area distinctly exceeds the average in High Asia (Gardelle et al., 2013; Neckel et al., 2014). A better understanding of these changes is of utmost importance, i.a. because of the glaciers’ relevance for Southeast Asia’s water supply and potential melt‐induced natural hazards such as landslides (e.g., Korup et al., 2010), glacier lake outburst floods (e.g., Benn et al., 2012), rock ava‐ lanches (e.g., Hewitt, 1999, 2009), debris flows, flash floods, etc. (e.g., Milly et al., 2002; Fort, 2014).

Knowledge on late Holocene glacier fluctuations in southeastern Tibet and its forcing mechanisms is, however, still widely lacking. Studies from the wider region found heterogeneous patterns regarding the adaption of glaciers to climate change (e.g., Scherler et al., 2011a; Yao et al., 2012; Owen and Dortch, 2014; Murari et al., 2014). Previous attempts to obtain a late Holocene glacial chronology for the eastern Nyainqêntanglha Range led to variable results which are contradictory in some parts (e.g., Yang, B., et al., 2008; Yi et al., 2008; Xu and Yi, 2014). A time of particular interest regarding the investigation of Holocene glacier fluctuations is the so‐called ‘little ice age’ (LIA), referring to cold phases of near‐global impact within the last millennium (Grove, 2004). Lasting approximately from the 16th until the end of the 19th century, its maximum represents the last major turning point in glacial development, i.e., switching from an advancing to a retreating regime.

The research synthesized in this dissertation aims to improve the understanding of spatial and tem‐ poral patterns of late Holocene glacier fluctuations in the eastern Nyainqêntanglha Range and to identify relevant climatic forcing mechanisms. The studies have been conducted within the scope of the project ‘Monsoonal variations and climate change during the late Holocene derived from tree rings and glacier fluctuations’ which was established by the German Research Foundation (Deutsche 1

Forschungsgemeinschaft, DFG) as part of the priority program 1372 entitled ‘Tibetan Plateau: For‐ mation – Climate – Ecosystems (TiP)’. The project contributed to two TiP‐topics, ‘Late Cenozoic evo‐ lution and environmental response’ and ‘Phase of human impact and global change’. Within this pro‐ ject, a multi‐proxy approach including glaciological, geomorphological, sedimentological and dendro‐ chronological methods, aided by optically stimulated luminescence (OSL) and radiocarbon (14C) da‐ ting, was applied. This was facilitated by cooperation between the workgroups of Prof. F. Lehmkuhl (RWTH Aachen University), Prof. A. Bräuning (FAU Erlangen University) and the Institute of Tibetan Plateau Research at the Chinese Academy of Sciences.

1.2 Objectives and approach The studies presented here focus on glaciological, geomorphological and sedimentological aspects, aiming to contribute knowledge regarding the following key research questions:

‐ How did glaciers in the eastern Nyainqêntanglha Range change since the Little Ice Age? ‐ Which, if any, regional spatial and temporal patterns are evident? ‐ Which factors influence the heterogeneous dynamics displayed by different glaciers? ‐ Which forcing mechanisms drive the changes?

The multi‐proxy approach consisted of three main phases. In the first step, remote sensing‐based analyses were conducted to obtain insight into the glacier‐topography‐climate interactions on a re‐ gional scale. This involved:

‐ the creation of a glacier inventory from multi‐spectral satellite imagery; ‐ the identification and detailed analysis of key settings with high explanatory value; ‐ the calculation of equilibrium line altitudes (ELAs) by a geomorphometrical approach; ‐ the statistical analysis of topographical constraints regarding glacier features; ‐ the analysis of spatial patterns to identify possible climatic forcing mechanisms; and ‐ the creation of detailed maps depicting the findings.

This also allowed identification of promising sites for fieldwork which constituted the second phase, which encompassed:

‐ detailed geomorphological mapping of selected glacier forelands; ‐ sediment sampling of different archives, including moraines, scree and glaciofluvial terraces; ‐ OSL and 14C sampling; and ‐ tree ring sampling (FAU Erlangen).

The third phase was comprised of laboratory analyses and the synthesis of results: ‐ Grain size analysis ‐ Geochemical analysis and application of element ratios ‐ Dendrochronological analysis and oldest‐tree moraine dating (FAU Erlangen) ‐ OSL and 14C dating ‐ Creation of a moraine morphosequence ‐ Evaluation of dating results from the different approaches

‐ Creation of a conceptual chronosequence in a synopsis of all findings ‐ Assessment of possible paleoclimatic forcing mechanisms of glacier fluctuations

1.3 Thesis outline Serving as background information for the following, Chapter 2 outlines the current state of knowledge regarding the relevant scientific aspects and related methods. The presentation is first focused on the overall atmospheric circulation patterns of Southeast Asia, particularly on the mon‐ soonal systems. Subsequently, the state of research regarding paleoclimate reconstruction, glaciers and their late Holocene fluctuations, and the relevance of these changes toward humanity is out‐ lined. Then, methods for the analysis of past and present glacier changes are reviewed, including remote sensing, ELA calculation, glacial geomorphology, glacial sedimentology, and dating of glacier deposits.

Chapter 3 introduces the study area in detail, including geology, climate, geomorphology and vegeta‐ tion. Thereafter, Chapters 4 through 6 present the results in the same order they were submitted to international, peer‐reviewed journals (Table 1.1).

Chapter 4 focuses on remote sensing‐based glacier mapping. Particular focus was put on the delinea‐ tion of debris‐covered glacier termini, the identification of relevant topographic and topoclimatic forcing mechanisms, the characterization of different glacier types, and the deciphering of spatial distribution patterns. Moraines of the LIA stage were identified by their morphostratigraphic posi‐ tions and from interpretation of the vegetation succession. Different approaches for calculating ELA estimations were evaluated, resulting in the application of an optimized toe‐to‐summit method which allowed further insights into the study areas’ glacier state and late Holocene dynamics.

In Chapter 5, a detailed map showing the key findings of Chapter 4 is presented. Here, the focus is on the cartographical challenges of depicting complex, high‐resolution data in a way that fosters intui‐ tive investigation of spatio‐temporal relationships. However, the detailed examination of the data and optimization of both interpolation and visual representation also led to insights going beyond those of the prior study. In particular, the spatial distribution of ΔELA values, i.e., the quantitative change of ELAs since the LIA, could thus be evaluated in considerably higher detail.

Results from fieldwork and subsequent laboratory analyses are presented in Chapter 6. Here, the application of the multi‐proxy approach is demonstrated at selected study sites at Mount Pulongu and compared to other settings in the Zepu catchment. Applied methods include detailed geomor‐ phological mapping, sedimentological analysis, and numerical as well as relative dating methods, i.e., OSL, tree ring, 14C and weathering indices. Results allow an assessment of the applicability of differ‐ ent dating methods for LIA deposits in this high‐energy environment. Consideration of local topogra‐ phies and topoclimates yields insights regarding the local glacier–topography–climate interactions, fostering the creation of a conceptual glacier chronology for the eastern Nyainqêntanglha Range. Finally, paleoclimatic datasets are reviewed to evaluate possible large‐scale climatic forcing mecha‐ nisms.

Table 1.1: Publications

Chapter Publication Status (May 2015) 4 Reconstructing glacier retreat since the Little Ice Age in SE Tibet by Geomorphology glacier mapping and equilibrium line altitude calculation Submitted Loibl, D., Lehmkuhl, F., Grießinger, J. 23 Oct 2013 Accepted DOI: 10.1016/j.geomorph.2014.03.018 12 Mar 2014 Published online 20 Mar 2014

5 Glaciers and equilibrium line altitudes of the eastern Nyainqêntan‐ Journal of Maps glha Range, SE Tibet Submitted Loibl, D., Lehmkuhl, F. 18 Feb 2014 Ac‐ cepted DOI: 10.1080/17445647.2014.933451 7 Jun 2014 Published online 24 Jun 2014

6 Toward a late Holocene glacial chronology for the eastern Quaternary Sci‐ Nyainqêntanglha Range, southeastern Tibet ence Reviews Loibl, D., Hochreuther, P., Schulte, P., Hülle, D., Zhu, H., Bräuning, A., Submitted Lehmkuhl, F. 19 Aug 2014 Accepted DOI: 10.1016/j.quascirev.2014.10.034 31 Oct 2014 Published online 25 Nov 2014

Chapter 7 provides a synthesis in which the major findings are joined in a synoptic conclusion. In addition, challenges and opportunities which altered the direction of the initial project concept are outlined. Finally, desiderata discovered during the course of the investigations and resulting perspec‐ tives for further research are described.

2 Current state of research

2.1 Atmospheric circulation patterns The TP’s high plateau topography strongly influences High Asia’s climate system, particularly the jet‐ stream‐induced mid‐latitude westerlies, the Asian summer monsoon, and the so‐called winter mon‐ soon (Domrös and Peng, 1988). This situation leads to complex interactions and the various domi‐ nances of polar, continental, mediterranean, subtropical and tropical air masses, often associated with extreme climatic contrasts (Böhner, 1994). Distinct differences are evident between the sum‐ mer and winter situations, arising from reversed large‐scale longitudinal barometric pressure gradi‐ ents (Fig. 2.1; Barry and Chorley, 2009). During summer, the northward shift of the innertropical con‐ vergence zone (ITCZ) results in a division of upper‐tropospheric wind directions around the TP, with prevailing westerlies north of the Kunlun Range and easterlies south of the Himalayas (Strahler and Strahler, 2004). Closer to the surface, advection of air masses from the Indian Ocean and the eastern Pacific Ocean transports moisture and heat toward the TP, representing the ISM and EASM, respec‐ tively (Böhner, 1994). During winter the ITCZ is south of the equator, constraining the westerlies to split into a northern and a southern branch at the western fringe of the TP (Prell and Kutzbach, 1992; Raymo and Ruddiman, 1992). These two branches are converging in the lee of the TP over north‐ eastern China (Strahler and Strahler, 2004).

Fig. 2.1: Circulation patterns over southern and eastern Asia, focusing on TP and bordering moun‐ tains. Solid lines indicate airflow at c. 3000 m, dashed lines at c. 600 m. The names refer to the air systems. Land surfaces above 4000 m are shaded. The left detail shows summer conditions, right win‐ ter conditions (adapted from Benn and Owen, 1998; Lehmkuhl and Haselein, 2000; Strahler and Strahler, 2004; Barry and Chorley, 2009).

Observations indicate that the ISM’s anomalous vorticity is primarily induced by Rossby wave dynam‐ ics whereas the EASM also encompasses multicellular meridional circulation (Lau et al., 2000). Con‐ sequently, the ISM is considered a ‘classical’, and the EASM a ‘hybrid’ monsoonal system. The rele‐ vant forcing mechanisms –particularly the role of the TP– have, however, been under debate for decades and are still not finally clarified (e.g., Flohn, 1953; Manabe and Terpstra, 1974; Ye, 1981, Tao and Ding, 1981; Ding, 1992). The traditional explanatory approaches focus on the ‘heat plate’ effect of the TP which causes a mid‐tropospheric thermal low, triggering the advection of tropical air mass‐ es (e.g., Flohn, 1968; Manabe and Terpstra, 1974; Hahn and Manabe, 1975). Subsequently, these ideas have been extended to better integrate sea surface temperatures and other relevant low pres‐ sure systems, e.g., the monsoonal depression over India, resulting in a broader concept which con‐ siders land–sea heating contrasts as the dominant forcing mechanisms (Webster 1987; Young 1987; Fu and Fletcher, 1985; Meehl, 1994; Li and Yanai, 1996). Recent modeling results indicate that ele‐ vated heating over the TP may be of minor importance for controlling the monsoonal systems; in‐ stead, the role of the Himalayas as a “thermal insulator” is being stressed (Boos and Kuang, 2010, 2013).

As a macro‐scale phenomenon, the Asian summer monsoon is intertwined and interactive with other global‐scale circulations (Kutzbach, 1987). The most extensively studied teleconnection concerns the El Niño‐Southern Oscillation (ENSO) which is found to be negatively correlated to Asian monsoon intensity (Shukla and Paolino, 1983; Webster and Yang, 1992; Ju and Slingo, 1995; Chou, 2003). Sev‐ eral studies found significant relationships between the North Atlantic Oscillation (NAO) and the ISM (e.g., Dugam et al., 1997; Gupta et al., 2003; Goswami et al., 2006), the EASM (e.g., Linderholm et al., 2011), and the Asian Winter Monsoon (Wu and Huang, 1999; Gong et al., 2001). Additionally, obser‐ vations show a positive correlation of ISM intensity to the mean tropospheric temperature at 200– 700 hPa in western Eurasia (e.g., Liu and Yanai, 2001) and Eurasian snow cover (e.g., Hahn and Shukla, 1976; Dey and Bhanu Kumar, 1982; Bamzai and Shukla, 1999). Particularly for the northwest‐ ern part of the TP and the Tien Shan region, a substantial influence of the mid‐latitude westerlies on local climates has been identified (e.g., Scherler et al., 2011a; Yao et al., 2012). Mölg et al. (2013) have shown that this influence penetrates far into the plateau by identifying a strong positive corre‐ lation between central Tibetan wind speeds and glacier mass balance. Considering millennial time‐ scales based on a statistical evaluation of Asian speleothem and ice core records from both hemi‐ spheres (cf. chapter 2.2.1), Rohling et al. (2009) suggest “that the East Asian monsoon is influenced by a combination of northern hemisphere ‘pull’ (which is more intense during boreal warm periods), and southern hemisphere ‘push’ (which is more intense monsoon during austral cold periods).” They argue that dominant control of millennial‐scale monsoon variability differs between glacial and inter‐ glacial times, i.e., stronger influence of southern hemisphere climate changes during glacial times (weak monsoon) and stronger influence of northern hemisphere climate changes during deglacial and interglacial times (strong monsoon), respectively.

Observations show that temperatures on the TP have increased at higher rates than the Northern Hemispheric mean, particularly during winter (Wang et al. 2001; Liang et al. 2009: Liu et al. 2009). This leads to more winter snowfall and the retreat of permafrost (Zhao et al. 2004; Li et al., 2008). The relevance of these observations in the context of atmospheric circulation patterns arises from findings that suggest a coupling of snow cover/depth on the TP with monsoon intensity in the follow‐

6 ing summer (Zhao and Moore, 2004; Zhang et al., 2004; Mölg et al., 2012). Modelling scenarios pre‐ dict an intensification of the Asian monsoons to be associated with rising temperatures, and indeed, precipitation trends in most regions of China are positive (Zhao et al. 2004; Wu et al. 2006) with an increasing trend of heavy rainfall events (Liu et al. 2005). Climate stations on the TP are, however, sparse. Consequently, considerable effort is necessary to extrapolate the few direct measurements to the vast extent of the TP. This represents an ongoing challenge for the climate modelling commu‐ nity, particularly regarding the integration of the abundant complex local topoclimatic influences. Results from such work are provided by Böhner et al. (2006), Maussion et al. (2014), and Gerlitz et al. (2015), for example.

2.2 Paleoclimate and glacier fluctuations The following chapters provide a brief summary of the broad field of paleoclimate reconstruction research, glacier fluctuations and their relationships. To keep this section short and focused, priority was placed on the temporal and spatial domains relevant for the subsequent investigations, i.e., the Holocene Epoch, in particular the last millennium, and the TP, in particular the eastern Nyainqêntan‐ glha Range. Studies from other regions or with focus on different time spans were only added where considered relevant, e.g., through teleconnections, for a broader perspective on certain phenomena, or to demonstrate methods, techniques, concepts, etc. For the sake of simplicity and comparability, all dating results will be noted in ‘ka’ or ‘CE’ despite different dating methods.

2.2.1 Paleoclimate reconstruction To reconstruct climate conditions prior to recorded measurements, proxy data has to be analyzed. Numerous different archives may be used as paleoclimatic proxy datasets, each with different prop‐ erties regarding temporal resolution, covered timespan, reaction time, proneness to data gaps, etc. Furthermore, many proxies are not reacting solely to climate change and variability in the datasets may often be induced by multiple climatic parameters, such as temperature, precipitation, and cloud cover. Typically, archives used for paleoclimate reconstruction are continuously accumulating mate‐ rial with time. Prominent paleoclimate archive types comprise lake sediments, bog sediments, loess– paleosol sequences, speleothems, tree rings, ice cores, marine sediments, and corals (Lowe and Walker, 1997). Except for the latter two, data from all these proxies are available for the TP and its surroundings. However, in comparison to other regions, such as the European Alps or the Rocky Mountains, relatively few proxy datasets are available and they were gathered over a huge and cli‐ matologically heterogeneous area.

Lake sediments have become one of the best studied archives of Quaternary paleoclimate since De Geer (1912) introduced the concept of varve chronology. During recent years, a wealth of studies investigating paleoclimate on the TP based on lake sediments have been undertaken, particularly within the scope of the Sino‐German TiP project (cf. chapter 1.1). Typically, corers are used to re‐ trieve sediment samples from lake bottoms (cf. Brinkhurst et al., 1969; Wright Jr., 1990). The analyti‐ cal techniques employed to examine lake sediment vary widely, but include geochemistry (e.g., Mor‐ rill et al., 2006; Herzschuh et al., 2010; Mischke and Zhang, 2010; Aichner et al., 2012), palynology (e.g., Herzschuh et al., 2009, 2014; Wang, Y., et al., 2014), magnetics (e.g., Herb et al., 2013), stable isotope measurements on different organic compounds (e.g., Günther et al., 2011, 2013; Zech et al.,

2014), analysis of remnants of benthic/limnic life, such as diatoms (e.g., Wischnewski et al., 2011b), ostracods (e.g., Mischke et al., 2010; Börner et al., 2013; Yan and Wünnemann, 2014) and gastropods (e.g., Taft et al., 2012, 2013), as well as various combinations of these. A comparative study by Wischnewski et al. (2011c) found, however, that inference from lake sediment core proxy data to paleoclimate is exacerbated by different in‐lake process regimes, dating uncertainties, pronounced heterogeneity of local climates, and changing atmospheric CO2 concentrations during the Holocene. Consequently, they conclude that combining lacustrine and terrestrial archives is indispensable when drawing regional scale conclusions. The importance of in‐lake processes is also highlighted by Her‐ zschuh et al. (2010) in the context of geochemical and stable isotope analysis of aquatic macro‐ phytes, challenging the application of these proxies for paleoclimatic inferences. Additionally, the nature of lakes as an integrative archive (Wünnemann, 2014) and obstacles to robust dating, such as the lake reservoir effect (Mischke et al., 2013, and references therein), have to be considered. Re‐ garding the last millennium, paleoclimatic constraints from limnolgical proxy data are in many cases far from conclusive owing to limited temporal resolution, dating uncertainties (cf. chapter 2.3.3) and disturbance of the uppermost sediment layers by the coring procedure.

Another archive widely used for inferring paleoclimatic insights is peat bog sediments. Analytical methods typically include palynology (e.g., Schlütz and Lehmkuhl, 2009; Cao et al., 2015), geochemis‐ try (e.g., Ferrat et al., 2012a) and isotope analysis (e.g., Wang et al., 2010; Ferrat et al., 2012b). As a consequence of slow accumulation, limited temporal resolution, and dating uncertainties, the appli‐ cation of peat‐bog‐derived paleoclimate reconstructions focuses on millennial to orbital timescales. Consequently, relatively few studies present results regarding late Holocene climate on the TP based on analyses of peat bog sediments.

Loess and loess–paleosol sequences are frequently studied archives in the semiarid regions of the central and northern TP as well as bordering regions, particularly the Chinese Loess Plateau (e.g., Liu and Ding, 1998; Zan et al., 2013). Grain size (e.g., IJmker et al., 2012a; Stauch et al., 2012), magnetic susceptibility (e.g., Liu and Ding, 1998; Maher et al., 2003) and geochemical (e.g., Gallet et al., 1996; Chen et al., 1999; IJmker et al., 2012b; Lehmkuhl et al., 2014) analyses of loess sediments allow the drawing of general conclusions about paleotemperature and –precipitation regimes.

To date, no speleothem records are available from the TP. Nevertheless, several datasets from mon‐ soonal Asia (e.g., Wang et al., 2001, 2008; Dykoski et al., 2005; Duan et al., 2014) allow inferences on Holocene monsoon intensity variability based on δ18O ratios (cf. Hendy and Wilson, 1968; Thompson et al., 1974; McDermott, 2004). Ideally, incremental dating is possible when speleothems show an‐ nual layers; otherwise numerical dating is typically provided by mass spectrometry of 230Th (Lawrence Edwards et al., 1987). Owing to the varying but generally slow accumulation of speleothem calcite, most proxy records from this archive only provide trends for the last millennium.

During the accumulation of glacier ice, the annual increments are preserved as a sequence of annual layers (Walker and Lowe, 1997; cf. Jouzel, 2013, and references therein). Contained in the glacier ice are numerous paleoenviromental signals which may be utilized as proxy records, including atmos‐ pheric dust (e.g., Petit et al., 1990; Ram and Koenig, 1997; Lambert et al., 2008), chemical composi‐ tion/trace element content (e.g., Mayewski et al., 1990, 1993; Palais et al., 1992; Taylor et al., 1993; Crowley and Unterman, 2013), and stable isotope composition (e.g., Jouzel et al., 1987; Leuenberger 8 et al., 1992; Johnsen et al., 1995). Five ice core datasets from glaciers on the TP have been published to date, originating from the Dunde ice cap in the Qilian Range (Thompson et al., 1989; Yao et al., 1991), the Guliya ice cap in the western Kunlun Range (Thompson et al., 1995), the Purogangri ice cap in the Tanggula Range (Thompson et al., 2006), and two Himalayan glaciers, Dasopu (Thompson et al., 2000) and East Rongbuk (Qin et al., 2002; Kang et al., 2002).

A paleoclimatic proxy archive of particularly high relevance for the late Holocene is tree rings (cf. Stokes and Smiley, 1968; Fritts, 2001). Tree ring chronologies are typically developed from multiple trees, often including multiple samples per tree, using statistical methods (cf. Holmes, 1983; Cook and Kairiukstis, 1990; Fritts, 2001). Owing to the sensitivity of trees to a variety of ecological and climatological factors, the properties of individual tree stands, e.g., upper tree line or arid lower for‐ est border, are of high importance for paleoclimatic inferences; typically, dendrochronoligcal records are most sensitive to the greatest limiting factor at a particular stand (LaMarche, 1974; Kienast et al., 1987). Tree‐ring‐related analytical methods comprise ring width (e.g., Esper et al., 2002; Fritts et al., 1970; LaMarche, 1974), maximum latewood density (e.g., Briffa et al., 1988; D’Arrigo, 1992; Davi et al., 2003), latewood lignin content (e.g., Gindl et al., 2000), and stable isotopes (e.g., Lipp et al., 1991; Anderson et al., 1998).

As seen in the previous paragraphs, the application of stable isotope ratios is abundant in proxy‐ based paleoclimate research. Drawing inferences from these ratios about a certain paleoclimatic parameter, such as precipitation or temperature, is, however, far from straightforward in many cas‐ es, thus demanding a closer look. The method is based on observations showing that certain states of the atmosphere lead to a relative enrichment/depletion of heavy isotopes, such as D, 13C or 18O (Daansgard, 1964). The processes behind these ratios are, however, manifold –often dependent on atmospheric, biologic, pedogenic, chemical and other system components– and intertwined in com‐ plex relationships. Regarding the 18O/16O ratio, a study by Dansgaard et al. (1969) found a good glob‐ al scale correlation between annual means of δ18O and temperature (∼ 0.7‰ per degree Celsius). This interrelationship, often referred to as ‘temperature effect’, could be “explained semiquantita‐ tively by considering the formation of precipitation as a Rayleigh process, whereby with continuing formation of precipitation from an air mass the remaining water vapor is progressively being deplet‐ 18 18 16 ed in O because H2 O condenses preferably to H2 O” (Siegenthaler and Oeschger, 1980; cf. McCrea, 1950; Dansgaard, 1964; Eichner and Siegenthalter, 1976). However, δ18O in precipitation is also influenced by numerous other factors and processes, such as the isotopic composition of water vapor in the source region (e.g., Yamanaka et al., 2002), admixture of isotopically different water vapor (e.g., Friedman et al., 1962), 18O enrichment due to evaporation (e.g., Ehhalt et al., 1968), etc. Additionally, δ18O in precipitation is not only correlated to temperature but is also influenced by the so‐called ‘amount effect’, i.e., decreasing 18O in meteoric water is dependent on the humidity and amount of precipitation (Daansgard, 1964). Further fractionation and isotopic exchange may occur in the terrestrial systems, for example by selective processes in vegetation, soils and different water bodies, before the isotope contents are ultimately fixed in a proxy record (e.g., Luo and Wang, 2009; Luo et al., 2013; Kanner et al., 2014). Regarding the interpretation of stable isotope measurements from trees, complex processes of intra‐plant fractionation have to be considered (e.g., Roden et al., 2000). Similar atmospheric, terrestric and biologic effects apply to the δD and δ13C ratios but a de‐ tailed discussion of those would be beyond the scope of this brief review.

As a consequence of the variety of possible climatic and non‐climatic factors influencing stable iso‐ tope fractionations, the determination of the factor predominantly controlling the isotopic composi‐ tion in the proxy record is crucial to paleoclimatic inferences. For high‐resolution archives accumulat‐ ing material to the present, living trees for instance, this can be achieved by a comparative regression analysis or bootstrapped response function with instrumental data from nearby meteorological sta‐ tions (cf. Guiot, 1991; Fritts, 2001). With other archives, e.g., speleothems or ice cores, such calibra‐ tions are much more complex, exacerbating the inference from proxy to climate signal (cf. Krinner et al., 1997; Fisher and Treble, 2007; Fairchild and Baker, 2012). Additional challenges arise from local climate heterogeneity, as in situations when the meteorological station is relatively far away or lo‐ cated in a different topographical position than the archive. Furthermore, anthropogenic modifica‐ tions may add other forcings during the recent period, thus ‘corrupting’ the climate signal and ‘violat‐ ing the fundamental assumption of uniformitarianism that underpins proxy climate research (Briffa et al., 1996).

2.2.2 Holocene paleoclimate on the Tibetan Plateau The last major peak of worldwide glaciation, the global Last Glacial Maximum (gLGM), was in the late Pleistocene, lasting approximately from 26.5 ka to 19–20 ka (Clark et al., 2009; cf. Ehlers and Gib‐ bard, 2004a, 2004b, 2004c). The subsequent warming during the Bølling and Allerød interstadials was followed by a sudden cooling that occurred at ∼ 12.8 ka and led to a return to near‐glacial condi‐ tions for > 1000 years, referred to as the Younger Dryas period (Mangerud et al., 1974). After 11.7 ka, abrupt warming marked the beginning of the Holocene which was, however, far from stable regard‐ ing climatic conditions (e.g., Bond et al., 1997). Several distinct climate change events have been observed in proxy data, typically characterized by polar cooling, tropical aridity, and major atmos‐ pheric circulation changes (Mayewski et al., 2004; Wanner et al., 2008, 2011). Two outstanding events at 8.2 ka and 4.2 ka, both of near‐global impact, are widely used to subdivide the Holocene Epoch into ‘early’, ‘mid’ and ‘late’ phases (Walker et al., 2012). During the late Holocene, climatic shifts became increasingly pronounced in what is commonly termed the ‘Neoglacial’ (e.g., Nesje and Kvamme, 1991), resulting in climatic optima such as the ‘Medieval Warm Period’ (∼ 1000–1300 CE; MWP; cf. Hughes and Diaz, 1994; Grove and Switsur, 1994; Crowley and Lowery, 2000; Hormes et al., 2001) and pessima such as the ‘Little Ice Age’ (∼ 1300‐1890 CE; LIA; cf. Mann, 2002; Grove, 2004).

A first review study of Holocene climate on the TP by Wang and Fang (1987) suggested a regional subdivision into five climatic periods: (1) the slightly cold and dry Wumadung interval, 10–7.5 ka; (2) the warm and moist Qilongduo interval, 7.5–3 ka; (3) the cold mid‐Neoglacial period, 3–1.5 ka with an intermittent warm phase 2.5–2 ka; (4) the mild Dawelong interval, 1.5–0.3 ka; and (5) the cold Little Ice Age, 0.3–0 ka. Since then, a wealth of studies –briefly summarized in the following– added to the subject, supporting, challenging, or adding detail to this early work.

During the last decade, numerous studies from different archives, including peat bog (e.g., Schlütz and Lehmkuhl, 2009), speleothems (e.g., Shao et al., 2006), lake sediments (e.g., Morill et al., 2006; Colman et al., 2007; Herzschuh et al., 2009; Aichner et al., 2012; Opitz et al., 2012; Herzschuh et al., 2014; Wang, Y., et al., 2014) and loess‐paleosol sequences (e.g., Porter, 2001; Stauch et al., 2012; Lehmkuhl et al., 2014) found that increasingly warm and moist conditions characterized the early Holocene on the TP. Dates for the beginning of this phase vary between ∼ 13 ka (e.g., Wischnewski

10 et al., 2011c; Wang, Y., et al., 2014) and ∼ 10 ka (e.g., Colman et al., 2007; Herzschuh et al., 2014). Some archives indicate an abrupt onset of warmer and moister conditions (e.g., An et al., 2012; Her‐ zschuh et al., 2014; Wang, Y., et al., 2014); others show gradual changes (e.g., Stauch et al., 2012; Lehmkuhl et al., 2014). Strong heterogeneity is evident between different studies regarding the end of the warm and moist early Holocene in both timing and climatic characteristics. The results can be grouped into studies that suggest (i) a two‐phase Holocene, (ii) a three‐or‐more‐phase Holocene, and (iii) a relatively gradual increase/decrease pattern throughout the Holocene. Type (i) studies found ongoing warm‐wet climate until ∼ 7.5–7 ka (Wang et al., 2005; Morrill et al., 2006), ∼ 6.5–6 ka (Her‐ zschuh et al., 2009; Schlütz and Lehmkuhl, 2009), ∼ 5 ka (Herzschuh et al., 2014), ∼ 4 ka (Shao et al., 2006; Colman et al., 2007) or even ∼ 3.5 ka (Dykoski et al., 2005). Most studies of type (ii) indentified intercalated cold/dry phases, e.g., at 9.5–7.3 ka (Wang, Y., et al., 2014), 8 ka (Aichner et al., 2012), 6.1–5.4 ka (Mischke et al., 2010) and 8.7–8.3 ka and ∼ 7.4 ka (Herzschuh et al., 2006). A gradual in‐ crease to a Holocene optimum and subsequent gradual decrease, i.e., type (iii), is suggested by Wang et al. (2005), Schlütz and Lehmkuhl (2009), and Dong et al. (2010), for example. Despite considerable differences regarding the timing, most studies indicate a return to colder and dryer conditions during the late Holocene (e.g., Porter, 2001; Colman et al., 2007; An et al., 2012). In addition to the long‐ term variability, many studies identified distinct events on decadal to centennial time scales, such as droughts/monsoon failures (e.g., Hu et al., 2008; Cai et al., 2010; Duan et al., 2014) and cold events (e.g., Colman et al., 2007; Zheng et al., 2007; Large et al., 2009; Mischke and Zhang, 2010).

Several review studies aimed to identify large‐scale patterns behind the heterogeneous results from different archives and regions in terms of both timing and paleoclimatic characteristics. Analyzing 92 proxy records from 72 sites, Wang et al. (2010) found that maximum wet conditions occurred during the early Holocene in the ISM‐dominated regions of Central Asia. In contrast, the EASM‐ and particu‐ larly the Westerly‐dominated parts of northern China show a delayed peak in maximum moisture, often occurring during the mid‐Holocene. The authors explain these observations by a strengthened Hadley circulation during the early Holocene, leading to subsidence in the EASM and Westerly re‐ gions. Herzschuh et al. (2010) found similar patterns and additionally points out, that a widespread trend to reduced effective moisture can be observed in the late Holocene. The out‐of‐phase relation‐ ship between monsoonal and continental climates in Asia was observed clearly in studies focusing on arid central Asia (Chen et al., 2008). Even though the aforementioned reviews mostly see monsoonal air masses as a key driver of Holocene climate change in Central Asia, implying that higher moisture is typically coupled with higher temperatures, a review with a clear focus on temperature changes in Central Asia for the whole Holocene is still lacking.

Regarding the late Holocene and particularly the LIA, which is of the highest interest for the study at hand, paleoclimatic datasets from the TP are still relatively scarce. Consistently, studies on late Holo‐ cene precipitation and monsoon intensity on the TP and in monsoonal China found a decrease in comparison to the Holocene optimum (e.g., Porter, 2001; Cai et al., 2010; Dong et al., 2010; Stauch et al., 2012). However, overall conditions were still relatively wet (e.g., Mischke et al., 2010; Wischnew‐ ski et al., 2011c), exhibiting dry phases/weak monsoon in events with significant centennial to deca‐ dal oscillations (Cai et al., 2010; Chen et al., 2015). Speleothem records from monsoonal China indi‐ cate widespread dry phases at ∼ 3.7–3.1 ka, ∼ 500–1000 CE and during the LIA (Hu et al., 2008; Duan et al., 2014). At Linhua cave in central China, Yin et al. (2014) found that severe droughts in mon‐

11 soonal China occurred during the Maunder, Dalton and Damon sunspot minima, highlighting the coupling of EASM intensity and solar irradiation. A speleothem record from the northeastern fore‐ land of the TP indicates relatively weak monsoon between the mid‐14th and late‐19th century, exhibit‐ ing substantial fluctuations which correlate well to historical records (Zhang et al., 2008). Similar pat‐ terns apply for ISM influenced regions (e.g., von Rad et al., 1999; Agnihotri et al., 2002; Anderson et al., 2002; Gupta et al., 2003, 2005; Kasper et al., 2012). In contrast, studies from the Himalayas often found above‐average precipitation during the LIA, indicating a substantial portion of the moisture may have been conveyed by stronger‐than‐today Westerlies (e.g., Davis et al., 2005; Kotlia et al., 2012). Higher‐than‐today summer season moisture during the LIA was also reconstructed using tree‐ ring‐based studies for sites near Reting (Grießinger et al., 2011) and Lhamcoka (Wernicke et al., 2015). Thompson et al. (2006) found indications for slightly warm and very wet climate 1600–1730 CE, cool and wet conditions 1730–1780 CE, and intermittent cool/warm and mostly dry climate 1780–1915 CE, followed by warmer and wetter conditions prevailing until present at the Purogangri ice cap. In the eastern Nyainqêntanglha Range, Liu et al. (2013) found indications for weakened ISM from 1600–1650 CE, continuous ISM strengthening until 1740 CE, and a slight weakening from 1740 CE to present. According to their results, the wettest period of the last 400 years occurred from 1700–1750 CE. Shi et al. (2012) report 1807–1817 CE to be “an exceptional decade marked by persis‐ tent dry conditions, possibly related to the impact of repeated volcanic eruptions on regional mon‐ soon flow”. A recent review by Chen et al. (2015) found different “hydroclimatic dipole modes” dur‐ ing the MWP and LIA; i.e., a “West–East” and a “North–South” mode for mid‐latitude Asia and mon‐ soonal eastern China, respectively.

In comparison to paleoprecipitation and monsoon intensity, studies on late Holocene paleotempera‐ tures in central Asia are relatively scarce. A review paper by Yang et al. (2003), including ice core, tree ring and lake sediment records, identified warm intervals during 800–1000 CE and 1150‐1400 CE and cold periods during 300–500 CE and 1400–1900 CE, representing the LIA. Furthermore, they found partially reversed patterns of warming and cooling on the northern and southern TP, respectively, possibly indicating that the aforementioned interactions between Westerlies and monsoonal air masses also apply on centennial to decadal timescales. Only a few paleotemperature reconstructions are available for the southern and eastern TP. Schlütz and Lehmkuhl (2009) investigated peat records from the Nianbaoyeze region on the eastern TP, identifying cool and dry climates prevailing from 2.75 ka, climaxing during the LIA. Zheng et al. (2007) and Large et al. (2009) conducted similar studies on peat records in the neighboring Hongyuan region; the former found cold periods occurring at 3.1 ka, 2.1 ka and 1000 CE, the latter at 3.1 ka and 200–500 CE, notably both lacking a LIA signal. Large et al. (2009) point out that their records indicate rapid transitions to cold and dry climates whereas the subsequent recovery takes considerably longer. In comparison to central and eastern China, cold and dry phases seem to be of longer duration on the eastern TP (ib.). A LIA cold event (∼ 1400–1900 CE) is, however, evident in a record from Lake Ximencou in the Nianbaoyeze Mountains. Additionally, this record also supports a cold phase at ∼ 300–700 CE. Analyzing tree ring data covering approxi‐ mately the last 400 years, Bräuning and Mantwill (2004) found phases of cool summers in southeast‐ ern Tibet at around 1700 CE, 1730–1750 CE, 1810–1820 CE and in the 1970s; these results are gener‐ ally in accordance with earlier findings by Wu (1992) and Cook et al. (2003) from other regions on the TP and in the Himalayas, implying broad‐scale climate anomalies. Wang, J., et al. (2014) found cold

12 phases in the Hengduan Range during the last millennium at 1140–1350 CE and 1600–1800 CE. A tree ring study from the eastern Nyainqêntanglha Range found warmer than average August tem‐ peratures during the periods 1446–1494 CE, 1509–1522 CE, 1553–1567 CE, 1797–1812 CE, 1845– 1905 CE and 1918–2002 CE (Zhu et al., 2011). Conversely, cooler than average Augusts occurred at 1385–1416 CE, 1426–1445 CE, 1495–1508 CE, 1523–1552 CE, 1568–1686 CE, 1695–1718 CE, 1725– 1796 CE, 1813–1844 CE and 1906–1917 CE (ib.).

2.2.3 Glaciers and climate change Glaciers are sensitive to climate change owing to the direct linkage of temperature and precipitation to glacier mass balance; simplistically, falling/rising temperature and increasing/decreasing precipita‐ tion lead to positive/negative glacier mass balance (Benn and Evans, 2010). Regarding glaciers in low latitudes, Thompson et al. (2005) suggest that precipitation is the primary driver of glaciation, argu‐ ing that “this is consistent with the time‐transgressive nature of precession‐driven changes in insola‐ tion (and hence precipitation) such that glaciers advance/retreat in the tropics north of the equator while glaciers retreat/advance in the tropics south of the equator”. Net annual glacier mass balance is defined as the sum of accumulation and ablation over one year (Benn and Evans, 2010). As a con‐ sequence of hypsographic temperature gradients, typical mountain glaciers accumulate ice in their upper parts and loose mass in their lower parts. Mass is transferred from the accumulation area to the ablation area by ice flow. A glacier’s ice flow velocity depends on various parameters, particularly basal ice temperature, glacier bed topography and mass budget (Hutter, 1983).

The equilibrium line marks the border between annual accumulation and ablation areas on the glaci‐ er surface, joining points where the net annual mass balance is zero (Benn and Evans, 2010). Consid‐ ering a conceptual longitudinal cross‐section of a mountain glacier, the equilibrium line is located at a certain altitude which is termed ELA, thus being widely synonymous to the terms ‘firn line’ (Menzies, 1951) and ‘niveau 365’ (Gross et al., 1977). The ELA concept is, however, widely theoretical and sim‐ plistic because local effects such as windblown snow and shadowing induce strong local effects (e.g., Oerlemans, 2001; Dahl et al., 2003). Consequently, real world equilibrium lines often vary in altitude across a glacier. Furthermore, considering the dynamic nature of glaciers, ELAs reflect solely the situ‐ ation in the year of observation and may be highly variable over longer periods. Therefore, the con‐ cept of the steady‐state ELA has been introduced, i.e., the ELA for a glacier which has zero net mass balance as a whole (cf. Benn and Lehmkuhl, 2000). Since most glaciers are, however, subject to con‐ stant change, a steady state can only be assumed for large‐scale transition phases from advancing to retreating or retreating to advancing regimes, i.e., glacial maxima and minima (Maisch et al., 1981). For all other glacier change observations, glacier response times must be considered carefully be‐ cause changes in terminus position are integrated reactions to past climate changes (e.g., Jóhannes‐ son et al., 1989; Harrison et al., 2001; Leysinger Vieli and Gudmundsson, 2004; Raper and Braithwaite, 2009). Beside the abovementioned parameters directly controlling the ice flow velocity, glacier response time is also affected by glacier size (e.g., Bahr et al., 1998) as well as the abundance and properties of supraglacial debris cover (e.g., Anderson and Mackintosh, 2012). Consequently, climatic inferences from length and area change are more challenging than from mass changes, but the latter are harder to measure (Bolch et al., 2012).

2.2.4 Late Holocene glacier fluctuations and climatic forcing The LIA glacier advance was the most extensive in most northern hemisphere alpine settings during the Neoglacial period, compromising the reconstruction of earlier advances from glacial landforms (Koch, 2014, and references therein). Reasons for the outstanding intensity of the LIA advance are still under debate. In general, the lack of observational data and the scarcity of high‐resolution proxy data covering the whole Neoglacial period (cf. chapter 2.2.1) exacerbate a comparative investigation. Additionally, the available proxy data records show strong spatial and temporal heterogeneity (e.g., Bradley and Jones, 1993; Wanner et al., 2008, 2011; Mann et al., 2009).

Since anthropogenic emissions are widely negligible for the LIA, two natural forcing mechanisms are typically held responsible for the majority of the observed/reconstructed climate change: (i) varying solar insolation and (ii) major volcanic eruptions (Crowley, 2000). (i) Of particular relevance in the context of solar climate forcing are the last millennium’s grand solar minima, such as the ‘Wolf’, ‘Spörer’, ‘Maunder’ and ‘Dalton’ minima (cf. Beer et al., 1991). During these periods, sunspots were absent, resulting in reduced solar radiation output (e.g., Suess, 1979; Stuiver and Quay, 1980; Usoskin et al., 2003; Vonmoos et al., 2006; Muscheler et al., 2007). Several studies conclude that LIA climate change was predominantly caused by this variability of solar irradiance (e.g., Lean and Rind, 1999; Büntgen et al., 2006). The difference in total solar irradiance caused by the grand minima events is, however, still under vivid debate (e.g., Weber, 2010; Feulner, 2011a; Shapiro et al., 2011; Schrijver et al., 2011) and several recent studies suggest it is rather small (e.g., Feulner, 2011b; Hind and Moberg, 2012; Moberg, 2013). (ii) Large volcanic eruptions inject considerable amounts of aero‐ sols into the atmosphere, resulting in decreasing surface temperatures (Rampino and Self, 1982). Documented volcanic events during the late LIA comprise the Tambora (1815 CE; Stothers, 1984) and the Krakatau (1883 CE; Self and Rampino, 1981) eruptions, both larger than any event during the last century. Ice core‐based aerosol optical depth reconstructions yield, however, evidence for many more major eruptions (Crowley et al., 2008; Crowley and Untermann, 2013). An early modelling study by Robock (1979) found that volcanic dust forcing provides much better simulation results for reconstructions of past climate than sunspot numbers. Miller et al. (2012) observed that cool sum‐ mers and ice growth in Arctic Canada and Iceland temporarily coincide with phases of high volcanic activity. The dataset of Crowley et al. (2008) also shows good fits regarding the timing of major vol‐ canic events and cold excursions in a tree ring‐based temperature reconstruction.

In contrast to earlier Neoglacial cold phases, increased precipitation during the LIA was observed in several proxy records (Mayewski et al., 2004; Rasmussen et al., 2010; Wilhelm et al., 2012; Grießinger et al., 2011; Wang et al., 2013). Lamoureux et al. (2001) found that major summer rainfall events during the LIA are strongly clustered in years immediately following major volcanic events in Arctic Canada. However, there is also evidence for extreme drought events, for example, in eastern China (Shen et al, 2007).

Even though the number of local dating results is increasing quickly –particularly since cosmogenic nuclide dating became widely available– the construction of a unifying glacier chronology for the TP remains a challenging task. During recent years, summaries have been published by Yang, B., et al. (2008), Yi et al. (2008), Owen (2009), Murari et al. (2014), Owen and Dortch (2014), and Xu and Yi (2014). Out of these, Yi et al. (2008) focus on the Holocene, Yang, B., et al. (2008) on the last 2000

14 years and Xu and Yi (2014) on the LIA. All three studies rely, however, substantially on radiocarbon ages from pioneer studies which are subject to major uncertainties (Lehmkuhl, 1997). Yi et al. (2008) derive three phases of Holocene glaciation on the TP, 9.4–8.8 ka, 3.5–1.4 and 1.0–0.13 ka. The review by Yang, B., et al. (2008) suggest three main periods of glacier advances in Southeastern Tibet during the last two millennia, i.e., 200–600 CE, 800–1150 CE, and 1400–1920 CE. Xu and Yi (2014) found that most glaciers on the southern and northwestern TP advanced to their maximum LIA position during the early 14th century; conversely, they identify phases of glacier retreat during the late‐14th to early‐ 15th and 16th to early‐18th centuries. Based on reconstructions of regional ELAs for the maximum LIA glacier advance, Smiraglia (1997) and Su and Shi (2002) deduced a temperature depression of rough‐ ly 1K, not taking into account changes in monsoon precipitation.

2.2.5 Human perspective: risks and challenges In the context of global change, the impacts of glacier retreat and changing mountain environments on human societies are attracting increasing scientific attention. Particular focus has been put on estimations of glacier melt contributions to sea level rise (e.g., Pfeffer et al., 2008; Jacob et al., 2012; Gardner et al., 2013) owing to its far‐reaching consequences (cf. Nicholls and Cazenave, 2010). Even though the absolute ice mass incorporated in mountain glaciers and ice caps is small in comparison to the large ice sheets of Greenland and Antarctica, their heterogeneity and high sensitivity to cli‐ mate change has fostered ongoing scientific endeavors (e.g., Meier, 1984; Meier et al., 2007; Raper and Braithwaite, 2009). Glaciers in High Asia are of particular importance due to their role as a ‘water tower’ in the hydrological system supplying more than a billion people with water (Immerzeel et al., 2010). Within this system, glaciers act as temporal water storage, buffering water during times of peak precipitation (e.g., Hagg et al., 2006). A substantial retreat of High Asia’s mountain glaciers will impair this buffer function, potentially leading to increased flood risk in basins and lowlands during precipitation peaks (e.g., Milly et al., 2002; Fort, 2014). Simultaneously, drought risk during warm and dry seasons may increase as a consequence of declining melt water discharge (e.g., Hagg et al., 2006; Mall et al., 2006). At numerous glaciers in High Asia the retreat facilitated the emergence of moraine‐ dammed lakes, increasing the risk of glacial lake outburst floods (GLOF; e.g., Richardson and Reyn‐ olds, 2000; Qunicey et al., 2007; Benn et al., 2012). GLOFs pose a substantial threat not only to villag‐ es, farmland and basic infrastructure in the vicinity of drainageways but also to hydroelectric power plants (Kattlemann, 2003).

Additionally, climate warming may lead to landslides at slopes formerly stabilized by permafrost (e.g., Dadson and Church, 2005; Gruber and Haeberli, 2007; Kääb, 2008). The Himalayas and the eastern Nyainqêntanglha Range are reported as particularly prone to large landslide events (e.g., Shang et al., 2003; Hasegawa et al., 2008; Korup et al., 2010). Growing evidence indicates that rock‐ slides from adjacent slopes trigger large‐scale glacier surges (Hewitt, 2005, 2009) and vice versa (Gardner and Hewitt, 1990).

2.3 Methodological background

2.3.1 Remote sensing of glaciers Since most mountain glaciers are located in remote areas and hard to access owing to rugged topog‐ raphy, field mapping is often challenging. Therefore, remote sensing is the widely preferred method 15 to map glaciers, particularly in the context of creating comprehensive glacier inventories. In addition to the advantages of delineating glacier extents from an aerial perspective, multichannel satellite imagery facilitates automatization by identifying glaciated areas through spectral properties.

During the past decade, glaciers in considerable portions of the TP and bordering mountains have been mapped by remote sensing, particularly within the scope of the “Global Land Ice Measurements from Space” (GLIMS, Kargel et al., 2014) and “Randolph Glacier Inventory” (RGI, Pfeffer et al., 2014) projects. For High Asia, the quality of the resulting freely available datasets is widely superior to pre‐ vious inventories based on topographic maps and aerial imagery, such as the “Chinese Glacier Inven‐ tory” (CGI; Shi et al., 2010).

Beyond the mere mapping of glacier outlines, glacier dynamics and their adaption to changing cli‐ mates are of high interest. Several case studies conducted in the past decades have shown that High Asia’s glaciers are reacting with great heterogeneity (cf. Yao et al., 2012), ranging from considerable advances, as in the Karakorum region (Hewitt, 2005; Scherler et al., 2011a), to strong retreat, as in southeastern Tibet (Yang, W., et al., 2008). Gardelle et al. (2013) investigated surface elevation changes by measuring the difference in digital elevation models (DEMs) obtained at different times, i.e., SRTM and SPOT5 stereo imagery. Neckel et al. (2014) applied a similar approach based on multi‐ ple IceSat data takes. The results of both studies underline the spatial heterogeneity of glacier changes, indicating a general tendency to glacier retreat but high variability of absolute rates in the different regions under consideration. Furthermore, Kääb et al., (2012) measured glacier mass bal‐ ances along the Himalayas using remote sensing approaches. Their results show generally negative mass balances in the monsoon‐dominated south‐facing mountain ranges of the Himalaya and Kara‐ korum. Conversely, the Westerly‐prone northern portion of the Karakorum exhibits zero‐to‐positive mass balance values, such that the overall picture is consistent with the abovementioned case study results.

The observed heterogeneity is suggested to be closely linked to different climatological regimes, i.e., varying influences of the ASM, westerlies and winter monsoon across High Asia (Yao et al., 2012). Mölg et al. (2013) and Maussion et al. (2014) derived different accumulation type regimes of glaciers throughout the TP and bordering mountain ranges from climatological data. Their findings show dis‐ tinct regional patterns, ranging from predominant winter accumulation in the Karakorum and north‐ ern mountains ranges to summer and spring accumulation in the southeastern part of the TP. These patterns are generally consistent with the frequently observed linkage between glacier sensitivity to climate change and continentality of the climate regime (Klinge et al., 2003; Klingbjer et al., 2005; Hoelzle et al., 2007; Lehmkuhl et al., 2011; Bolch, 2007; Rowan et al., 2014).

The climatic conditions in High Asia and the abundant rugged topography also pose great challenges for remote sensing of glaciers. Steep accumulation areas promote glacier ice accumulation by ava‐ lanches which typically supply considerable amounts of debris (Benn and Lehmkuhl, 2000). At most such glaciers, this debris forms a continuous cover in the ablation area, obscuring the ice in the glaci‐ er tongue (Scherler et al., 2011a, 2011b; Sorg et al., 2012). Beside the distinct effects this debris cov‐ er has on the glaciers’ mass balance regimes, it also exacerbates the delineation of glacier outlines by remote sensing. Specifically, the spectral properties of debris cover are widely similar to those of neighboring bedrock and debris originating from neighboring slopes in many cases (Nicholson and 16

Benn, 2006; Racoviteanu and Williams, 2012). Additionally, glacier outlines may be obscured by cloud cover in optical remote sensing data. For the study area in southeastern Tibet, this is particularly problematic due to the coincidence of the monsoon season and the end of the ablation phase (cf. chapter 3.2), which should be used for glacier mapping (Raup et al., 2007a). After the monsoon phase, the switch to westerly‐dominated climate facilitates snow precipitation down to the valley floors, leaving only a very limited time frame to acquire adequate satellite imagery.

2.3.2 Equilibrium line altitude calculation Glacier mass balance data is typically obtained by ablation stakes drilled into the glacier along a lon‐ gitudinal transect (Østrem and Stanley, 1969). Naturally, the challenges for glacier mapping as de‐ scribed above apply for field‐work‐based mass balance measurements as well. Consequently, very few reliable mass balance datasets are available for the eastern Nyainqêntanglha Range (Yang, W., et al., 2008, 2010a, 2011, 2013). Additionally, in‐field mass balance measurements are always case studies of individual glaciers; their transferability to other glaciers for regional integration has to be evaluated carefully. An alternative approach to obtain information on glacier mass balances is meas‐ uring the ELA, the altitude were annual net mass loss is equal to annual net mass gain (cf. chapter 2.2.2). Since this border of accumulation and ablation areas is closely coupled to many aspects of glacier dynamics, many observable phenomena may be used as indicators for its modern position as well as to infer ELAs of former glacier stands. In general, methods to derive ELAs from glacier proper‐ ties are classified by basic approaches; ‘glaciological’ approaches comprise all concepts based on phenoma of the ice body, whereas ‘morphometrical’ approaches focus on hypsographical and geo‐ morphological properties. Further, it is important to differentiate between ELAs of individual glaciers and regional ELAs. Individual ELAs are strongly influenced by local topography and coupled topocli‐ matic effects, such as the aspect and its influence on insolation, orographic rainfall and shielding effects, valley systems channeling precipitation, etc. Conversely, regional ELAs integrate averaged data from multiple glaciers, providing a more general measure which is less prone to local effects. In the following, different ELA‐related phenomena and relevant methods for ELA measurement and calculation will be outlined briefly, including (i) the transient snowline, (ii) the convex–concave sur‐ face transition, (iii) the maximum elevation of lateral moraines, (iv) cirque floor altitudes, (v) the bal‐ ance ratio, (vi) the accumulation‐area ratio, (vii) the area‐altitude balance ratio, and (viii) different versions of the toe‐to‐summit ratio method.

(i) The term ‘transient snowline’ (TSL) refers to the edge between wet and dry snow on the glacier surface (Llibourty, 1964; Østrem, 1973, 1975). At the end of the ablation season, the TSL coincides with the individual glacier’s ELA for the particular observed year. Since wet snow and dry snow have distinctly different spectral characteristics, the boundary marked by the TSL can be identified in optical remote sensing data (e.g., Klein and Isacks, 1999; Pelto, 2011). (ii) In the accumulation area, convergent ice flow results in a concave plan curvature of the glac‐ ier surface (Benn and Evans, 2010). Conversely, divergent ice flow in the ablation area fosters convex plan curvatures. The ELA is therefore assumed to be at the transition between convex and concave plan curvatures at the glacier surface. This transition –also referred to as ‘kine‐ matic ELA’– can be identified by the contour inflection in topographic maps or by remote sensing when a DEM of sufficient resolution and geometrical precision is available (Leonard

and Fountain, 2003). Since the kinematic ELA is subject to substantial time lags it is providing an indication of the multi‐annual individual ELA of the investigated glacier. (iii) The transition from convergent to divergent ice flow at the ELA also has a substantial impact on the transport processes within the glacier. In the accumulation area, debris is being trans‐ ported predominantly toward the center of the glacier. Below the ELA, divergent flow facili‐ tates deposition of debris at the glacier margins, possibly resulting in the formation of lateral moraines (Nesje, 1992). Since the formation of lateral moraines is thus limited to the abla‐ tion area, the ‘maximum elevation of lateral moraines’ (MELM) provides an approximation of the individual glacier’s maximum ELA during the time the moraines were deposited (Lichtenecker, 1938; Visser, 1938; Andrews, 1975; Meierding, 1982). (iv) Another implication of these glacier flow characteristics is that glacial erosion is most effec‐ tive in the accumulation area, peaking slightly above the ELA. Provided sufficient time, these erosional processes tend to create cirque landforms with steep cirque walls and near‐planar cirque floors. Considering the related process‐response system, the altitude of cirque floors indicates the average ELA level during the time of cirque formation (Benn and Lehmkuhl, 2000). The possibility of paleorelief inheritance has, however, to be considered carefully in this context (Haman and Embleton, 1988; Richardson and Holmlund, 1996) (v) The balance ratio (BR) method requires in‐field mass balance measurements to establish mass balance gradients –i.e., the rate with which mass balance increases with elevation– for both accumulation and ablation areas (Furbish and Andrews, 1984). Provided that both gra‐ dients are linear and sufficiently good hypsometric data is available, the ratio between both gradients can be applied to obtain ELA estimates (cf. Benn and Gemmell, 1997). (vi) The basic assumption behind the ‘accumulation‐area ratio’ (AAR) method is that a linear relationship exists between the sizes of the accumulation and ablation areas of glaciers in an investigated region (Porter, 1975). Provided sufficiently detailed glacier outlines and eleva‐ tion data, this method determines an ELA approximation by dividing the glacier shape into an upper and a lower part using a ratio of their respective portions of the glacier’s spatial ex‐ tent. (vii) Based on similar assumptions but incorporating more glaciological knowledge, the ‘area‐ altitude balance ratio’ (AABR) is probably the most sophisticated method to calculate ELA approximations available so far (Osmaston, 2005). As a consequence, the AABR also requires most base data of all ELA calculation approaches: glacier outlines, glacier contours, detailed hypsometry, and balance ratio data (ib.). (viii) In contrast, the morphometrical ‘toe‐to‐summit’ (TSAM) approaches are highly simplified, being based solely on the assumption that the ELA represents a regional ratio between the highest and lowest points of a glacier. The key challenge is the adequate definition of the highest point of the glacier, resulting in three concepts differing in this point: the TSAM (Lou‐ is, 1955) uses the highest summit in the glacier catchment; the ‘toe‐to‐headwall ratio’ (THAR) approach defines it as the glacier headwall; and the ‘toe‐to‐ridge method’ (TRAM; Höfer, 1879) specifies it as the average ridge altitude. Höfer’s (1879) original definition of ridge as the ‘average altitude of the ridge above the ELA’ introduces, however, a circular argument, demanding clarification before application.

All of the abovementioned approaches except for the TSL are based on the observation of features that are subject to time lags regarding the glacier’s adaption to climatic forcing. Therefore, the con‐ cept of a ‘steady state ELA’ has to be introduced (Benn and Lehmkuhl, 2000), assuming that the glac‐ ier’s mass balance is zero during the observed timeframe. In the context of deriving ELAs from land‐ form evidence of former glaciations, this assumption is widely valid because the maximum glacier advance typically coincides with the transition from an advancing to a retreating regime, implying transient zero‐mass‐balance conditions (Maisch, 1987). Conversely, modern glaciers are subject to ongoing change, which is contradictory to the steady state hypothesis. Therefore, adaption lag times have to be considered carefully in this context.

As a consequence of the distinct coupling of glacier dynamics and local climate, absolute changes of ELAs over time (ΔELA) allow inferences on the relevant climate forcing mechanisms. Since the rela‐ tionships between temperature and precipitation at the ELA have been established using statistical and analytical methods (Ohmura et al., 1992; Shi, 1992; Shi et al., 2008), it is possible to derive pale‐ otemperatures when paleoprecipitation is known and vice versa. Therefore, regional ΔELA distribu‐ tions potentially allow the unravelling of past precipitation gradients, moisture sources and atmos‐ pheric circulation patterns (e.g., Miller et al., 1975; Sissons and Sutherland, 1976; Williams, 1983; Burbank and Kang, 1991; Lehmkuhl, 1998; Lehmkuhl et al., 1998; Klinge, 2001; Klinge et al., 2003) and correlation of moraines (e.g., Lehmkuhl, 1997; Klinge, 2001; Böhner and Lehmkuhl, 2005). Addi‐ tionally, glacier reactions to future climate changes can be predicted if robust climate prediction models are available.

2.3.3 Dating of Holocene glacier deposits Until the end of the 19th century, estimates of rates of geological and geomorphological processes were the only available basis for relative dating, i.e., methods that establish the order of antiquity (cf. Flint, 1971). These approaches were refined in the beginning of the 20th century, resulting in detailed moraine morphosequences in the European Alps (e.g., Penck and Brückner, 1909). More recently, new and refined methods lead to a reintroduction of relative moraine dating in the field. Schmidt hammers were used to investigate rock surface strength, providing an indication of surface weather‐ ing as indication of exposure time (Day and Goudie, 1977; Day, 1980; Goudie, 2006). If applied to moraine boulders of similar lithology, Schmidt hammer rebound measurements yield indications of the relative exposure times at different moraine stages (e.g., Matthews and Shakesby, 1984; Winkler, 2005, 2011; Matthews and Winkler, 2010). Another approach is to measure the geochemical finger‐ prints of glacial deposits and to analyze the degree of weathering by ratios of sensitive vs. insensitive elements, so‐called ‘weathering indices’. Several studies demonstrated that this method is suitable to distinguish moraines of different major Pleistocene and Holocene glacier advances (Bäumler et al., 1991, 1996; Guggenberger et al., 1998; Bäumler, 2001, 2004).

With the discovery of radiocarbon dating, numerical dating techniques, or methods that provide age estimates, for Holocene deposits became available in the 1950s (Libby, 1955). Primary glacial depos‐ its are, however, widely free of organic carbon. Consequently, the 14C method can only be applied to material incorporated into glacial sediments, such as buried trees or soil relicts, or to under‐ or over‐ lying organic material, such as buried soils, younger topsoils and peat (Lowe and Walker, 1997; Solo‐ mina et al., 2015). Noticeably, each of these constellations requires different interpretational ap‐

19 proaches, including minimum and maximum ages for glacier advances as well as intercalated phases of glacier retreat. Furthermore, numerical radiocarbon ages are dependent on calibration based on reconstructions of past atmospheric 14C content (Stuiver and Polach, 1977), which were recognized to be inconsistent soon after the method’s development (de Vries, 1958, 1959). This is particularly challenging for the period beyond dateable tree ring chronologies (Bronk Ramsey et al., 2006; Mel‐ lars, 2006; Turney et al., 2006; Blockey and Housley, 2009). However, a plateau in the calibration curve also exacerbates 14C age estimates for the period after ∼ 1650 CE, introducing significant da‐ ting uncertainties for radiocarbon samples from much of the LIA (Bard, 1998; Reimer et al., 2011). Numerous studies have applied 14C dating in glacial environments (e.g., Sissons, 1967; Reger et al., 1993; Lowell, 1995); however, many of them also highlight substantial challenges and limitations (e.g., Porter and Carson III, 1971; Matthews, 1984).

Due to the limitation of 14C dating to archives holding fossil organic matter, which are typically sparse in glacial environments, demand for alternative numerical dating methods increased in the 1970s (e.g., Dreimanis et al., 1978). This led to the increased application of luminescence dating techniques in glacial settings. According to the definition of Lian and Roberts (2006), “luminescence refers to the light emitted by some materials in response to some external stimulus”. Different luminescence methods focus on different stimuli, out of which heat (thermoluminescence, TL; Daniels et al., 1953; Aitken, 1985) and ionizing radiation (photoluminescence), particularly visible light (optically stimulat‐ ed luminescence, OSL; Aitken, 1998), are most widely used in Quaternary sciences (Berger, 1986, 1988; Duller, 2004; Lian and Roberts, 2006; Rhodes, 2011). Numerous studies investigated glacier‐ related sediments (see Thrasher et al., 2009 for a review), such as glaciofluvial outwash or loess cov‐ ering moraines, by TL (e.g., Berger, 1984) and OSL (e.g., Klasen et al., 2007; Bickel et al., 2015). The low sensitivity of quartz grains is, however, a common challenge in the dating of glacial sediments by luminescence methods (Rhodes and Pownall, 1994; Richards, 2000; Spencer and Owen, 2004). The difficulty has previously been attributed to a young geological bedrock source, a lack of subaerial weathering of the grains, or insufficient bleaching due to the proximity of the depositional setting to the provenance, requiring detailed analysis of the local settings (Preusser et al., 2006; Lukas et al., 2007; Fuchs and Owen, 2008; King et al., 2014a, 2014b).

During recent years, cosmogenic nuclide dating gained increasing importance as a method for nu‐ merical dating of moraine deposits (cf. Balco et al., 2011, and references therein). Conceptually, the technology utilizes the production of rare nuclides in surface rocks by cosmic rays, which is approxi‐ mately constant over time at a given location (Gosse and Phillips, 2001; Balco et al., 2008; Dunai, 2010). Production rates vary, however, with location on the Earth’s surface, and are predominantly controlled by altitude and position in Earth’s magnetic field (Balco et al., 2008). In the context of mo‐ raine dating, the cosmogenic nuclide approach is particularly useful because both glacier advances and retreats create fresh rock surfaces; dating of these surfaces’ exposures thus provides a relatively direct measure of the timing of glacial events. However, close attention has to be paid to sampling sites and material because incomplete exposure and post‐depositional shielding may alter the results (Heyman et al., 2011).

An alternative approach to obtain numerical dating results is incremental dating, which relies on “measurements of regular accumulation of sediment or biological materials through time” (Lowe and

Walker, 1997). The most widely used and prominent incremental approach is dendrochronology. It is based on the growth characteristics of coniferous trees which add cells to the outer perimeter each growth season, resulting in distinct tree rings that provide a means of determining the age of a tree (ib.). In a glaciological context, trees and their growth characteristics have high potential for infer‐ ences on past dynamics, including minimum ages for moraine stages, dating of events that tilted trees, or identification of years with climatic pessima (cf., e.g., Lawrence, 1950; Sigafoos and Hen‐ dricks, 1961; LaMarche and Fritts, 1971; Matthews, 1977; Karlén, 1984; Holzhauser and Zumbuehl, 1996). Additionally, dead trees incorporated in glacial deposits may be dated by tree rings if a chro‐ nology is present or otherwise by 14C, providing maximum ages for the forming events (Fritts, 2001). Trees are, however, highly sensitive to local stand parameters, including topography, climate, soils, hydrology, nutrient supply, etc. Consequently, these parameters need to be considered carefully, particularly for correction of the time necessary for growth to breast height and in multi‐site compar‐ isons (Sigafoos and Hendricks, 1969; Villalba et al., 1990). In addition, trees suffering from cold stress are susceptible to missing rings (Wilmking et al., 2012), requiring a cross validation of multiple cores per tree (Koch, 2009).

Several approaches aim to obtain numerical dating results from relative‐dating archives by calibrat‐ ing them with other data. The most prominent of these ‘pseudo numerical’ dating methods is li‐ chenometry, which centers on the idea that lichen growth is widely constant at a certain area of in‐ vestigation (Beschel, 1950, 1961). These growth rates can be determined using known‐age surfaces, such as buildings or grave stones, to establish growth curves (Benedict, 1967, 1968). Owing to their abundance on moraine boulders beyond the altitudinal and latitudinal tree lines, lichens have been used for dating numerous late Holocene moraines in, for example, Scandinavia (e.g., Denton and Karlén, 1973; Erikstad and Sollid, 1986; Matthews and Shakesby, 2008), Iceland (e.g., Mckinzey et al., 2004), Alaska (Wiles et al., 2010, and references therein), South America (e.g., Rodbell, 1992) and New Zealand (e.g., Gellatly, 1982). Lichen growth rates are, however, non‐linear with thallus size (Bradwell and Armstrong, 2007) and depend highly on local stand parameters, such as stone aspect (e.g., Dabski, 2007). To my knowledge, no robust lichen growth curve for High Asia was published until now. Further ‘pseudo numerical’ dating approaches relevant in the context of moraine dating are measurements of weathering rind thickness (cf. Chinn, 1981) and attempts to calibrate Schmidt hammer rebound measurements (cf. Shakesby et al., 2006; Matthews and Owen, 2010).

3 Regional setting

The study area is located in the central part of the eastern Nyainqêntanglha Range at the southeast‐ ern fringe of the Tibetan Plateau (Fig. 3.1). In the east‐west direction, the eastern Nyainqêntanglha Range extends for approximately 550 km, stretching from the Maidika Basin (∼ 93°E) to the western tributaries of the Three River Gorges (∼ 97°E) in the Hengduan Mountains. The Nu Jiang (Salween) catchment delimits the Mountain Range to the north (∼ 32°N) and the foothills of the orogen to the south (∼ 29°N).

Fig. 3.1: Topographic overview map of the study area. The two cities, Nyingchi and Bomi, are also the locations of the meteorological stations.

3.1 Geology Prior to the India–Asia collision, terranes originating from the Gondwana breakup were added to Asian lithosphere in a process of sequential accretion (cf. Molnar et al., 1993; Royden et al., 2008; Liebke, 2010). Beginning in the latest Paleocene and early Eocene (∼ 50 Ma), the India–Asia collision induced the rise of the Pamir‐Tibet‐Himalaya orogenic system (cf. Rowley, 1996; Hodges, 2000; Leech et al., 2005). Within the collision process, the collage of accretional terranes was substantially de‐ formed; its basic structure, however, predetermines the modern tectonic arrangement of the TP, consisting of a set of approximately east–west trending crustal blocks and interjacent suture zones 22

(Gloaguen and Ratschbacher, 2011). The mechanisms underlying the complex deformations, includ‐ ing the relatively homogeneous uplift of the TP to an average altitude of ∼ 5 km, and the patterns of crustal motion, have been the subject of scientific debate for decades (cf. Argand, 1924; Allégre et al., 1984; Molnar et al., 1993; Murphy et al., 1997; Chemenda et al., 2000; Yin and Harrison, 2000; Royden et al., 2008, and references therein) and continue to be so (e.g., Capitano and Replumaz, 2013; Faccena et al., 2013; Kohn, 2014; Replumaz et al., 2014). Measurements of crustal motion have revealed rotating patterns centering on syntaxes, of which the two major ones are located at the eastern and western boundaries of the rigid Indian plate indenter (Gloaguen and Ratschbacher, 2011). The eastern Himalayan syntaxis is located within the study area at the Namche Barwa–Gyala Peri massif (cf. Quanru et al., 2006; Booth et al., 2009). Recent studies indicate that rock uplift is par‐ ticularly high at this syntaxis, promoting high rates of erosion (Finnegan et al., 2008; Lang and Hun‐ tington, 2014). The study area thus marks the transition zone between Himalayan thrust tectonics and eastern Tibetan extrusion tectonics (Clark et al., 2004; Finnegan et al., 2008). Consequently, the region is frequently affected by earthquakes (e.g., Mukhopadhyay and Dasgupta, 2015). A prominent feature of the active strike‐slip tectonics is the Karakorum‐Jiali fault zone which exhibits dextral dis‐ placements of 10–20 mm/a (Armijo et al., 1989; Lee et al., 2003).

Within the TP’s tectonic collage, the study area is part of the Lhasa terrane. The main lithological components are metasediments, gneisses and amphibolitic to granulitic metamorphic rocks, and magmatic rocks of the Gangdese batholith and the so‐called northern plutonic belt (Finnegan et al., 2008; Wen et al., 2008; Booth et al., 2009; Zhang et al., 2014). Additionally, clastic sediments of ter‐ restrial and shallow marine origin as well as a few marine carbonates occur in the eastern Nyainqên‐ tanglha Range (Yin and Harrison, 2000; Seward and Burg, 2008; Zhang et al., 2012).

3.2 Climate The eastern Nyainqêntanglha Range marks a transition zone between the subtropical warm‐humid climates in the forelands and the autochthone continental high‐mountain climates on the Tibetan Plateau (Domrös and Peng, 1988; Maussion et al., 2014). As a consequence of this situation, climate in the eastern Nyainqêntanglha Range exhibits substantial heterogeneity in several respects, includ‐ ing (i) a distinct interannual seasonality and (ii) substantial effects of topography, particularly the orographic situation, as well as geographic location.

(i) Located on the track of monsoonal water vapor from the Bay of Bengal to the Tibetan Plateau, the regional climate in the eastern Nyainqêntanglha Range is predominately controlled by the ISM (Molnar et al., 2010). To a lesser extent, the EASM may also penetrate into the study area, even though most authors assume that its zone of influence does not reach far beyond 100° E (Chang, 2004; Fu et al., 2008). Typically, the phase of strong EASM precipitation covers spring and early summer, whereas the ISM precipitation peaks in July and lasts until September (Weischet and End‐ licher, 2000; Chang, 2004). As a consequence, summer is characterized by abundant precipitation in the eastern Nyainqêntanglha Range (Peng et al., 2000; Lu et al., 2004). Conversely, dry and relatively cold conditions prevail from October to March (Leber et al., 1995; Peng et al., 2000). Spring and au‐ tumn are both relatively short phases of transition between these dominant modes (Domrös and Peng, 1988; Molnar et al., 2010).

(ii) Continuing the Himalayan ridge to the east, the eastern Nyainqêntanglha Range poses a substan‐ tial orographic barrier to moisture being advected toward the Tibetan Plateau (cf. Boos and Kuang, 2010, 2013). Moisture‐bearing air masses are gradually lifted up on their way from the foothills in the south towards the high plateau, leading to orographic precipitation (Anders et al., 2006; Bookhagen and Burbank, 2006). In consequence, the southern foothills of the eastern Nyainqêntanglha Range are characterized by above‐regional‐average temperatures and precipitation (Yang, B., et al., 2008; Zhou et al., 2011; Figs. 3.2, 3.3). Toward the TP, annual precipitation sums decrease gradually along a southeast–northwest gradient (Zhu et al., 2013). On a catchment scale, the valley system exerts a substantial influence on the monsoonal air masses; particularly, the longitudinal Three River Gorges and the south‐trending segment of Yarlung Tsangpo‐Brahmaputra valley (between the Great Bend and the southern foothills) are known to act as major transport routes for monsoonal moisture to‐ wards the TP (Peng et al., 2000; Lu et al., 2004).

Fig. 3.2: Average annual air temperature at 2 m above ground according to the High Asia Refined Analysis (Maussion et al., 2014), averaged for the period 2001–2013. Boundaries of the detail are equal to Fig. 3.1.

Two meteorological stations providing accessible data are located in the study area; one near the city of Nyingchi (29.64° N, 94.36° E, ∼ 3000 m asl), and one near Bomi (29.96° N, 95.77° E, ∼ 2750 m asl, Fig. 3.1). Average annual precipitation for the period between 1971–2000 amounts to 683 and 895 mm, respectively; the average temperature is 8.7 C at both stations (cf. Fig. 4.2). Reflecting the sub‐

24 stantial influence of the monsoonal air masses, approximately 85% and 75% of the annual precipita‐ tion falls during the summer months at Nyingchi and Bomi stations, respectively.

Fig. 3.3: Spatial distribution of average annual precipitation according to the High Asia Refined Analy‐ sis (Maussion et al., 2014), averaged for the period 2001–2013. Boundaries of the detail are equal to Fig. 3.1.

Paleoclimatic proxy records are scarce in the study area. Out of the wide variety of archives available in High Asia (see chapter 2.2.1), the vast majority is located either on the TP or in its forelands. Con‐ versely, few archives are found in the high mountain ranges framing the plateau, particularly in those with rugged topography. For example, to date only one of the many High Asia limnological studies investigated lakes in the vicinity of the study area due to the lack of suitable archives (Wischnewski et al., 2011a). The most suitable archive in the region is tree rings because of the abundance of for‐ ests. Modelling results indicate that climate in the eastern Nyainqêntanglha Range generally yielded a distinct drying trend during the last 6 ka (Dallmeyer et al., 2013). The few available proxy records show, however, that wetter‐than‐today conditions prevailed throughout the LIA (Grießinger et al., 2011; Wernicke et al., 2015).

3.3 Geomorphology Similar to the Himalayas, the high mountain range of the eastern Nyainqêntanglha is bounded by sediment‐filled peripheral basins in the southern forelands and structurally controlled valleys on the

TP (Summerfield, 1991). The average elevation of the neighboring portion of the TP is widely exceed‐ ed by the eastern Nyainqêntanglha’s summit altitudes, including –besides numerous other peaks > 6000 m asl– the prominent massif of Mount Namcha Barwa (∼ 7800 m asl) and Mount Gyala Peri (∼ 7300 m asl) at the eastern syntaxis of the Himalaya orogen (Fig. 3.1). Conversely, abundant mon‐ soonal precipitation facilitates substantial erosion in the region, promoting deeply incised valleys (Fielding, 2000). Average elevation is thus only slightly higher compared to the TP, probably because the height of the peaks is isostatically balanced by the depth of valleys (Wager, 1933, 1937). Several studies investigated these interactions with a focus on the Namcha Barwa–Gyala Peri region; their findings support a distinct relationship between rapid bedrock exhumation and strong erosion driven by fluvial and glacial processes (Finnegan et al., 2008; Korup and Montgomery, 2008; Wang, P., et al., 2014; Tu et al., 2015). Geodynamic modelling experiments by Finnegan et al. (2008) indicate that the rapid uplift of the Namcha Barwa–Gyala Peri massif compensated for most of Yarlung Tsangpo’s inci‐ sion, preventing rapid headward erosion toward the TP. Relief increases from the southern foothills to the Namcha Barwa–Gyala Peri massif where the drop from the highest summit to the Yarlung Tsangpo–Brahmaputra valley floor is ∼ 6800 m within a horizontal distance of < 30 km (Fig. 3.1). Re‐ lief gradually decreases from this rugged topography toward the TP, tracing decreasing erosional potential (Finnegan et al., 2008).

The study area’s first order valley and drainage networks are widely prescribed by the aforemen‐ tioned major tectonic features. As such, the Yarlung Tsangpo River follows the trace of the Indus– Tsangpo suture for ∼ 1300 km to the Namcha Barwa–Gyala Peri massif, then cuts through the dome and subsequently rejoins the suture at the knickpoint of the so‐called ‘Great Bend’ (Seward and Burg, 2008). The Yarlung Tsangpo’s major tributaries in the region, the Yigong and Parlung Tsangpo Rivers, follow the course of the Jiali fault which crosses the study area in an almost straight line trending ENE–WSW for more than 400 km (Korup and Montgomery, 2008; Mukhopadhyay and Dasgupta, 2015). As a consequence of high relief, active tectonics, and high precipitation, the eastern Nyainqêntanglha Range is highly exposed to natural hazards, including earthquakes, landlsides and GLOFs (Shang et al., 2003; Korup and Montgomery, 2008; Mukhopadhyay and Dasgupta, 2015).

3.4 Glaciers The combination of high mountain topography and high precipitation also provides the basis for the extensive glaciation evident in the study area, estimated to account for 22.2% of China’s total glaci‐ ated area (Su and Shi, 2002). Glaciers in southeastern Tibet are predominantly of the monsoonal temperate‐type which is highly sensitive to climate change (ib.). According to a classification by Maussion et al. (2014), main accumulation phases are heterogeneous in the eastern Nyainqêntan‐ glha Range, shifting from predominant accumulation during spring in the eastern part to predomi‐ nant summer accumulation toward the TP in the west. The vast majority of monsoonal glaciers showed dramatic shrinking rates and negative mass balances during the past decades and are ex‐ pected to further retreat in the future (e.g., Smiraglia, 1997; Su and Shi, 2002; Li et al., 2008; Qiu, 2008; Yang, B., et al., 2008). The study area constitutes the regional center of Pleistocene glaciation (Zhou et al., 2011). Two major late Pleistocene moraine stages are evident, termed ‘Guxiang’ and ‘Baiju’ in reference to valleys in the Bodui Tsangpo catchment (Zhou et al., 2007). Sophisticated gla‐ cial chronologies for the Holocene are, however, widely lacking. Prominent LIA moraines were

26 mapped and dated in several case studies at multiple sites in southeastern Tibet (e.g., Bräuning and Lehmkuhl, 1996; Lehmkuhl et al., 2002, Klinge and Lehmkuhl, 2004). At some places, older Neoglacial moraines are also reported (Jiao et al., 2005).

3.5 Vegetation As a consequence of the characteristic topoclimatic gradients described before, the eastern Nyainqêntanglha Range constitutes an orobiom at the transition zone between tropical conditions in the southern forelands (zonobiom II), continental conditions on the eastern margin of the TP (zono‐ biom VII), and continental conditions with an artic‐like temperature regime on the central TP (zono‐ biom VII‐tIX; cf. Walter and Breckle, 1991, 1994, 2004; Miehe, 1994). Regarding floral kingdoms, the whole Himalayan orogen is considered a transition zone between Paleotropis and Holarctis (Walter and Breckle, 1991; Miehe, 1994). Heterogeneous lithological and climatic boundary conditions facili‐ tated a variety of different soils in the study area (Klinge and Lehmkuhl, 2003). The relatively favora‐ ble climate in the eastern Nyainqêntanglha Range –particularly in valley positions– supports conifer‐ dominated, mixed forests as potential natural vegetation (Cui and Graf, 2009). Pervasive tree species include spruce (e.g., Picea balfouriana), fir (e.g., Abies delavayi var. motouensis), larch (e.g., Larix griffithii), oak (e.g., Quercus serrata) and various rhododendrons (Singh and Singh, 1987; Bräuning and Mantwill, 2004). Further, epiphytes are abundant in the region (Singh and Singh, 1987). A lack of comprehensive forest planning and a low level of management led to shrinking forest areas in the study area during recent decades (Peng et al., 2000). Favored by the climatic and topographic boundary conditions, numerous settings where glaciers reach down into the forests of the montane altitudinal zone are evident in the eastern Nyainqêntanglha Range (e.g., Bräuning and Lehmkuhl, 1996; Bräuning 2006; Zhu et al., 2013). After glacier retreat, proglacial areas are rapidly settled by pioneer species, particularly seabuckthorn (Hippophae; Hochreuther et al., 2015).

4 Reconstructing glacier retreat since the Little Ice Age in SE Tibet by glacier mapping and equilibrium line altitude calculation

David Loibla, Frank Lehmkuhla, Jussi Grießingerb a RWTH Aachen University, Department of Geography, Templergraben 55, 52062 Aachen, Germany b FAU Erlangen‐Nürnberg, Institute of Geography, Kochstraße 4/4, 91054 Erlangen, Germany

Temperate glaciers in the eastern Nyainqêntanglha Range, southeastern Tibet, are highly sensitive to climate change and therefore of particular high interest for research on late Holocene changes of the monsoonal climate in High Asia. However, because of the remoteness of the area, the scarcity of empirical data, and the challenges to remote sensing work posed by cloud and snow cover, knowledge about the glacier dynamics and changes in this region is still very limited. In this study, we applied a remote sensing approach in which 1964 glaciers were mapped from a Landsat ETM+ scene and subsequently parameterized by DEM‐supported measurements. Geomorphological evidence, i.e., trimlines and latero‐frontal moraines, were used to obtain quantitative data on the glaciers’ morphological characteristics and the changes since the Little Ice Age (LIA) maximum glacier ad‐ vance. Statistical analysis of glacier length change revealed an average retreat of ∼ 27% and a trend toward stronger retreat for smaller glaciers. An evaluation of different methods to calculate equilib‐ rium line altitudes (ELAs) indicates that an optimized toe‐to‐ridge altitude method (TRAM) is more suitable than other methods in settings with complex topography and a lack of mass balance meas‐ urements. A large number of glacier measurements is crucial for high quality of TRAM results, and special attention has to be paid to different glacier characteristics. In order to determine the best‐ fitting TRAM ratio value and to test the quality of the calculated ELAs, a remote sensing approach was applied: for each investigated glacier, the altitudes of transient snowlines visible in the late summer Landsat scene were measured from the DEM and compared to TRAM results. The interpo‐ lated ELA results show a SE‐NW gradient ranging from 4400 to 5600 m asl and an average ELA rise of ∼ 136 m since the LIA. Because of the high spatial resolution of measurements, the ELA distribution reveals topographic effects down to the catchment scale, specifically orographic rainfall and leeward shielding. The interpretation of these patterns reveals that the eastern Nyainqêntanglha Range is influenced by both, the Indian (ISM) and East Asian summer monsoon (EASM). However, the EASM does not reach the western part of the study area. The results indicate that the monsoonal temper‐ ate glaciers’ high sensitivity to climate change is driven by two double forcings owing to the coinci‐ dence of accumulation and ablation phases.

4.1 Introduction The Tibetan Plateau is of central importance for the global circulation systems and the regional cli‐ mate of High Asia (e.g., Murakami, 1987). The Asian summer monsoon is characterized by the double branch system of ISM and EASM. From May to September, a complex interplay of both pressure gra‐

28 dients is responsible for the transport of humid air masses from the Bay of Bengal (ISM) and the Pa‐ cific Ocean (EASM) to the Tibetan Plateau. In particular, the southeastern part of the Tibetan Plateau and its adjoining highlands are characterized by a hydrological gradient of decreasing precipitation toward the inner Tibetan Plateau (Benn and Owen, 1998). Thus, monsoonal summer precipitation is of major importance for the accumulation of the region’s glacier ice, which acts as a long‐term stor‐ age and source for the major rivers of south and southeast Asia supplying water for more than 1.4 billion people (Immerzeel et al., 2010).

Glaciers in the southeastern part of the Tibetan Plateau, especially in the eastern Nyainqêntanglha Range, have a maritime temperature regime and are highly sensitive to climate change (Zhou et al., 1991). Higher ice‐layer temperatures lead to reduced nourishment, therefore temperate glaciers show remarkably high ice flow velocities. For example, Zheng et al. (1999) measured flow rates of 438 m/y and 533 m/y for Azha (29°32.05′ N., 96°06.06′ E.) and Minyong (28°26.33’ N., 98°41.01’ E.) glaciers, respectively. In the study area, dynamics of glacier advances and retreats are directly con‐ trolled by monsoon intensity and temperature. However, details of the spatiotemporal interaction of climate and glaciers in southeast Tibet are still not fully understood. A major problem is the lack of empirical data: mass balance measurements and meteorological stations are very sparse and data cover only short timespans (e.g., Yang, W., et al., 2010a). Furthermore, the meteorological stations are situated in valley bottoms, whereas glaciers are located within the alpine or nival zone. Conclud‐ ing, the knowledge on local climate and glacier response is as yet very limited. Particularly, the influ‐ ence of monsoonal systems and precipitation is not yet well understood.

Remote sensing data offers the opportunity to obtain detailed information concerning the modern state of glaciation (Raup et al., 2007b; Paul, 2010) and its recent changes (e.g., Bolch et al., 2008). However, the changes before the oldest available satellite imagery from the mid‐twentieth century cannot be mapped from ice extents. In order to obtain information about former glacial history, proxy archives have to be used. Geomorphological records such as glacial landforms of former maxi‐ mal glacier stands and lateral or terminal moraines can serve as archives of changes within the land‐ scape. In contrast to the European Alps, where surveys of glaciology and glacial geomorphology date back until the early twentieth century (e.g., Penck and Brückner, 1909; see Fiebig et al., 2011 for an overview), historical data sets on glaciers in southeast Tibet are scarce. In this context, the so‐called LIA is of particular interest. This worldwide cooling period began in the sixteenth century and lasted until the end of the nineteenth century (Grove, 2004). Dendrochronological studies in southeastern Tibet have shown that vegetation succession delineates moraines of the LIA maximum glacier ad‐ vance (Bräuning, 2006). The outer walls of the LIA moraines are typically covered by successional forests, whereas younger moraines host only few pioneer species. In contrast, the debris‐covered glacier tongues are free of higher vegetation species. Because this pronounced vegetation threshold represents a clear marker within the landscape, remote sensing can serve as a promising tool to de‐ lineate the spatial extent of glacier retreat since the LIA.

The ELA is the theoretical line at which accumulation and ablation are in balance over the period of one year (Porter, 1975; Benn and Lehmkuhl, 2000). Several studies have demonstrated the possibility to derive reliable estimations of present and former ELAs by methods that are based on measure‐ ments of geometric parameters of glacial geomorphological features (e.g., Meierding, 1982; Sa‐

29 voskul, 1997). However, the methods based on glacier morphology have been criticized to be rather crude in comparison to ELAs derived from actual mass balance measurements (Benn et al., 2005). It is therefore important to carefully assess the applied ELA method and validate the results (Benn and Lehmkuhl, 2000). The ELA calculations for different time slices represent the relative changes in mass balance over the given time span. In the context of ELA calculations, the LIA maximum is of particular high relevance; it represents the last major turning point in glacial development, switching from an advancing to a retreating regime. Theoretically, at this point all glaciers were in a steady‐state, i.e., ice mass and geometry were in equilibrium with climate (Benn and Lehmkuhl, 2000). Taking into account different glacier response times caused by differences in size, topography, etc., the ELA of the LIA maximum represents an important reference level for the evaluation of post‐LIA glacier changes (Maisch, 1981). The ELA estimates of previous studies in the eastern Nyainqêntanglha di‐ verge widely: Shi and Li (1981) suggested a range from 4400 to 5200 m asl, whereas Yao et al. (2010) presented ELAs of 3370–5490 m asl for the Yigong Tsangpo and 3560–5980 m asl for the Parlung Tsangpo catchment.

In order to establish links between glacier change and climatic or topographic parameters, the situa‐ tion at the individual glacier scale has to be analyzed in detail. Detailed glacier mapping has a long scientific tradition and provides insight into each glacier’s state (Ashwell, 1982). Haeberli (2004) pointed out that changes in glacier length are among the most reliable and most easily observed terrestrial indicators of climate change. A topographical factor of particular strong influence in sub‐ tropical mountain environments is the aspect, which affects the amount of insolation received at the surface. In High Asia, slopes orientated toward the north thus offer much more favorable conditions for glaciation than their southern counterparts (Evans and Cox, 2005; Evans, 2006).

In this study a remote sensing approach to infer information on late Holocene glacier changes is pre‐ sented and factors influencing the observed pattern are discussed. In a first step, detailed glacier‐ mapping case studies were carried out in order to gain insight into the regions’ glacier characteristics. This included the role of different glacier types, the influences of different morphological catchment attributes, and the problem of delineating glacier outlines under debris cover. The resulting knowledge provided the basis for the next steps. An inventory of modern glaciers in southeastern Tibet was created by delineation from Landsat ETM+ satellite imagery. The glacier change since the LIA was measured from geomorphological evidence, i.e., moraines and trimlines. Subsequently, the data set was parameterized with morphological attributes, supplemented by digital elevation data. In consideration of the rugged topography and the lack of empirical data, different methods of ELA calculation were evaluated. Afterward, ELAs were calculated on a regional scale for the modern and for the LIA stages. Measurements of tongue retreat distances and ELA results were then statistically analyzed in order to disentangle the complex climate–relief–glacier interactions.

4.2 Regional setting and glacier distribution The eastern Nyainqêntanglha Range covers an area of about 8 x 104 km² on the southeastern edge of the Tibetan Plateau. It stretches from the highlands in the west to the three river gorges in the east (Fig. 4.1). The east–west directional mountain range holds the eastern syntaxis of the orogenic sys‐ tem that is reflected by intense tectonic dynamics (Burg et al., 1998; Seward and Burg, 2008). This

30 resulted in a strong uplift, evidenced by average summit elevations of more than 6000 m asl up to almost 7800 m asl at mount Namcha Barwa. The first‐order valley system is prescribed by large faults (Liu et al., 2011). Key erosional features originating from strong linear erosion along these faults are the great bend of the Yarlung Tsangpo River (Brahmaputra in India) and the Yigong and Parlung Tsangpo Valley systems which run straight for more than 300 km. The interplay of a still ongoing tec‐ tonic uplift and erosion is the main reason for the rugged topography with steep valleys and slopes. The altitudinal differences between valley bottoms and surrounding peaks often reach 3000–3500 m within a short horizontal distance of < 10 km (Finnegan et al., 2008).

Fig. 4.1: Study area in southeastern Tibet. Elevations are based on SRTM data.

The intense erosion is triggered by high precipitation amounts during the summer monsoon season (May to September; Fig. 4.2). In contrast to more arid areas north of the Himalaya main ridge, the investigated area is characterized by a strong climatic influence of the ISM entering through the Yar‐ lung Tsangpo valley. Climate data from the meteorological stations Nyingchi and Bomi show that more than 80% of annual precipitation falls from June to September. In contrast, winter months are characterized by cold and dry conditions in the absence of monsoonal precipitation (Molnar et al., 2010; Fig. 4.2). A climatic classification of local meteorological station data by Leber et al. (1995) indi‐ cates that the eastern Nyainqêntanglha Range marks a transition zone between warm‐wet subtropi‐ cal and cold‐dry plateau conditions. Remarkable differences between the two meteorological sta‐ tions in Nyingchi and Bomi are apparent in the annual precipitation and its monthly distribution: Nyingchi is, in average, considerably less humid (683 mm vs. 895 mm; Fig. 4.2), and the phase of in‐ tense summer monsoon rainfalls is shorter (> 60 mm per month from May to September vs. March

31 to October). In Nyingchi the precipitation regime is monomodal, showing a maximum in July whereas the Bomi station is characterized by a bimodal precipitation regime with a first peak in June and a second in September. It has to be emphasized that the recorded precipitation data is not representa‐ tive for the high elevation glacier sites as both stations are located at the valley bottoms.

Fig. 4.2: Climate diagrams for the two stations in the study area, Bomi and Nyingchi (see Fig. 4.1 for exact locations). Dark blue areas show the average precipitation for the period 1971–2000, yellow lines the average temperature for the period 1971–2000. The conditions relevant for the evaluation of the Landsat scene used in this study are also depicted; blue bars and red lines represent the precipita‐ tion and temperature for 1999, respectively, and the grey bars highlight the month of September.

Climate models by Böhner (2006) and Maussion et al. (2014) indicate that average annual precipita‐ tion in most high‐elevation areas in the southern part of the study area exceeds 2000 mm. Farther north, the east–west striking ranges act as barriers forcing heavy orographic rainfalls. This results in a distinct precipitation gradient from the south toward the drier highlands of the Tibetan Plateau.

Because of the high elevation, the monsoonal summer precipitation accumulates as snow in the up‐ per reaches of the mountain range despite the subtropical latitude of about 29° to 31°N. This allows for a large number of maritime temperate glaciers, with a coincidence of accumulation and ablation maxima in summer (Kaser and Osmaston, 2002). The late summer (Aug–Oct) climate of the period 1960–2004 shows distinctively alternating temperature and precipitation, i.e., temperature and pre‐ cipitation are anti‐correlated (Fig. 4.3). In contrast to the prominent interannual variability, no long‐ term trends are obvious.

Fig. 4.3: Late summer (Aug, Sep, Oct) precipitation (left ordinate) and temperature (right ordinate) averages of Bomi and Nyingchi meteorological stations for the period 1960–2004. The black line marks the year 1999.

4.3 Methods and materials

4.3.1 Data selection The Glacier Inventory of China (GIC) represents a large data set including size and distribution of glac‐ iers supplemented by basic topographical and glaciological parameters (Jiao and Zhang, 1988; Mi and Xie, 2001). In order to evaluate whether this data could be used as a basis for the present study, a quality assessment for the eastern Nyainqêntanglha Range was carried out. The GIC was compiled following the guidelines of the world glacier inventory (Müller et al., 1977; Paul et al., 2010) and pro‐ vides glacier outlines as shapefiles. These have been extracted manually from maps and aerial pho‐ tographs from the mid‐twentieth century. A comparison with a Landsat ETM+ scene from 1999 (see below) revealed large discrepancies to the GIC shapes in terms of glaciers missing in the GIC, glaciers that exist in the GIC but have no pendant in the satellite image, and both, over‐ and underestima‐ tions of glacier sizes by inaccurate outlines (Fig. 4.4). Because precise glacier data is a prerequisite for the processing steps described below, the first step of the study was the delineation of glacier out‐ lines.

Fig. 4.4: Example of glacier outline quality assessment for the Glacier Inventory of China (GIC) data by comparison to a Landsat ETM+ (channels 5‐4‐3): (A) glacier missing in the GIC data set, (B) inaccurate outlines resulting in overestimation of glacier size, (C) inaccurate outlines resulting in underestimation of glacier size, (D) GIC glacier outline without a corresponding glacier in the Landsat ETM+ data.

Landsat ETM+ satellite imagery served as a basis for glacier mapping (Bindschadler et al., 2001). The scene had to meet three requirements: (i) minimal cloud cover to enable comprehensive mapping; (ii) no snow cover to make sure glacier outlines can be delineated properly; (iii) a late summer acqui‐ sition date to ensure that only multi‐annual ice is mapped (Paul et al., 2010). Because monsoonal cloud cover is usually dense until the end of September and snowfall often already starts in Novem‐ ber, potential scenes are limited to a small time frame around October. Furthermore, the failure of the Landsat 7 Scan Line Corrector made all scenes collected after 31 May 2003 unusable for compre‐ hensive scientific surveys. The scene from 23 September 1999 offered a good compromise with no snow cover and < 10% clouds. Most mapping was performed using a composite of channels 5‐4‐3 pansharpened to 15‐m resolution.

Absolute elevations were derived from ASTER GDEM2 and SRTM 4.1 data. A comparison of the two DEMs shows that the study area’s topographic and climatic conditions of the study area had negative effects on both data sets (Figs. 4.5B and D). The ASTER GDEM2 (∼ 30‐m resolution) shows intense noise and some areas of strong artifacts that probably originate from clouds in the original image. The SRTM data (∼ 90‐m resolution) provides smoother surfaces but summits and ridges are remark‐ ably low in many cases. Remarkably, the areas with the highest mountains in the region, Namcha Barwa and Gyala Peri, show gross errors in the SRTM DEM. These smoothing errors in the SRTM data are originated in data gaps caused by layover and shadowing effects during SAR scanning in rugged terrain. In the void‐filled SRTM 4.1 DEM, these data gaps were subsequently filled with interpolated, low‐resolution DEM data from other sources (Rott, 2009; Jacobsen, 2010). An assessment of the alti‐

34 tudes of several higher summits confirmed that ASTER GDEM2 is more precise in terms of summit and ridge elevations while SRTM provides the better overall picture of the relief situation.

Fig. 4.5: Data sources, interpretation, and errors: (A) interpretation of key features from the Landsat ETM+ scene, channels 5‐4‐3; (B) SRTM DEM of the same area; note the smoothing effects on several ridges; (C) aspect calculated from the SRTM DEM; note the clearly visible ridgelines; (D) the ASTER GDEM2 showing extreme artifacts in this area. The scene shows the tongue of glacier (C) in Fig. 4.7.

4.3.2 Glacier mapping The mapping was carried out in ESRI ArcGIS 10.1 in several phases: (i) case study on debris cover and morphological characteristics, (ii) delineation of ice‐covered surfaces as continuous objects, (iii) cut‐ ting of the resulting objects along ridges and spurs to retrieve glaciers, and (iv) parameterization of the glaciers with relevant topographical, glaciological, and geomorphological attributes.

(i) A detailed understanding of the influence of supraglacial debris cover is important in order to ob‐ tain a reliable delineation of glacier extents. Debris cover has spectral characteristics very similar to those of surrounding materials (e.g., lateral moraines, bedrock) and therefore obscures the boundary between glacier ice and dead ice (Nicholson and Benn, 2006). In order to address this challenge, a detailed mapping case study was performed on several prominent glaciers. All of these are showing outstanding features such as remarkably little or much debris cover or remarkably few or many signs of stagnant ice beyond the debris cover. Amongst others, flow patterns, water discharge structures, melting ponds, and collapse structures were examined in the 1999 Landsat ETM+ scene (Fig. 4.5A). Another key aim of the case study was to evaluate the feasibility of mapping the maximum LIA glaci‐ er extent, i.e., latero‐frontal moraines and trimlines. In agreement with the findings of Bräuning

(2006) the outer walls of LIA moraines are typically covered by forests, whereas the proglacial areas are almost free of vegetation. Hence, the remote sensing delineation of LIA moraines was possible. Furthermore, the influences of morphological catchment characteristics and different glacier types were examined in the case study in order to better understand the topography–glacier interactions.

(ii) In the second step, ice‐covered surfaces were delineated for the whole Landsat scene by manual interpretative mapping from channel combination 5‐4‐3. The application of automated mapping rou‐ tines as discussed by Bhambri and Bolch (2009) was not suitable owing to the variety of difficult situ‐ ations such as cloud cover, shadows, and supraglacial debris. A set of different interpretative meth‐ ods was used to overcome these obstacles. In cases where debris cover was present, this was noted in the attribute table, and the debris cover was studied in detail to delineate the glacier extent as precisely as possible (see above). Cloud cover obscures parts of each available satellite image be‐ cause the time window for glacier mapping coincides with the end of the monsoon season. Following the instructions of the ‘World Glacier Monitoring Service’, cloud‐covered sections of glacier outlines were bridged by straight lines (Müller et al., 1977). The mapping scale varied between 1:10,000 and 1:50,000 depending on the settings’ complexity. Interpretations of critical settings were additionally validated with Google Earth’s 3D‐view. In accordance with the ‘Global Land Ice Measurements from Space’ conception, the minimum size of glaciers to be mapped was 0.1 km² (Paul et al., 2010).

(iii) Ridgelines were extracted automatically using ArcGIS’ stream delineation tools on a hole‐filled inverse DEM. Because of its more precise representation of topography, the SRTM DEM served as input data here. The result was then corrected by matching it to ridge areas against a topographic position index raster (Jenness, 2006) and checking it manually using ASTER GDEM2 and Landsat ETM+ data (see also Figs. 4.5A and 5C). Afterward, the ice cover shape file generated in step 1 was cut by the ridgelines to retrieve individual glaciers.

(iv) Step iv involved a variety of processes to collect attributes for each glacier. Topographical param‐ eters included the altitudes of the highest summit, the average headwall elevation, and the elevation of lowest recent and lowest LIA terminal moraines. These parameters were partly derived automati‐ cally (summit and recent terminal moraine) and collected manually from ASTER GDEM2 and SRTM data in the cases where this was not possible. Information on characteristics of the glacial morpholo‐ gy was entered into the data fields ‘glacier type’ (see chapter 4.3.3), ‘debris cover’ (see chapter 4.3.3), ‘glacier length’, and ‘retreat since LIA’. Glacier length was calculated automatically in ArcGIS, and glacier retreat was determined by semiautomatic measurements of the 3D‐distance between 1999 glacier tongue and LIA terminal moraine.

4.3.3 Glacier classification The mapped glaciers were categorized into four basic and two complex glacier types. These types were adopted from widely accepted concepts for glaciers constrained by topography (Rathjens, 1982; Benn and Lehmkuhl, 2000; Benn and Evans, 2010). Taking the topographic position as key dif‐ ferentiator, the four basic glacier types are (Fig. 4.6):

(1) ice caps and ice fields on plateaus, (2) slope glaciers in open slope positions,

(3) cirque glaciers in armchair‐shaped bedrock hollows, and (4) valley glaciers where iceflow reaches the valley bottom.

Additionally, two special types were introduced to describe valley glaciers in more detail. The type ‘complex valley glacier’ (4a) was assigned to valley glaciers connected with neighboring glaciers of other types, e.g., cirques. Situations where several valley glaciers join and form a network were clas‐ sified as ‘valley glacier network’ (4b).

Fig. 4.6: Typology of glaciers used in this study: (1) ice cap, (2) slope glacier, (3) cirque glacier, (4) val‐ ley glacier with debris‐covered tongue, (r) the special case of reconstitution.

Unobstructed and continuous ice flow is a prerequisite for the automated calculation of ELAs. Two special cases were identified in which the assumption of a coupling of climate changes and glacier size oscillations is invalid because of local relief parameters. The first is the case of ‘reconstituted glaciers’: these develop where very steep terrain causes intermittent ice avalanches that result in bare rock surfaces separating accumulation and ablation areas (Fig. 4.6r). The lower parts of such glaciers are typically cones with high debris content. These often reach down to altitudes far below the climatic conditions that support active glaciers. The second case applies to large valley glaciers reaching down to the lower reaches of valleys. Here, the inclination of the valley floor tends to be so low that glacier advances primarily lead to horizontal elongation. The ice surface exposed to ablative climate conditions is thus growing disproportionately large. At the same time, glacier advances into lower altitudes are hampered by the proximity to the base level of erosion (Kuhle, 1986; Lehmkuhl, 1989). Glaciers to which one of the two cases is applicable were marked and excluded from ELA cal‐ culations. Furthermore, calculation of ELA approximations for the LIA stage is not valid in settings where glacier retreat led to a separation of valley glaciers into tributaries. As a consequence, all glac‐ iers that were identified as part of a ‘valley glacier complex’ (4a) or a ‘valley glacier network’ (4b) for the LIA stage but not in the 1999 stage had to be excluded from ELA calculation. Another factor po‐ tentially influencing the coupling of climate changes and glacier size oscillations is supraglacial debris cover. Thick debris cover protects the underlying ice from insolation, whereas thin debris tends to accelerate ablation by lowering albedo (Benn et al., 2005). Yang, W., et al. (2010b) found highly het‐ erogeneous ablation rates strongly correlated to debris thickness for the 24K glacier in close vicinity to the study area. Consequently, the assumption of approximately linear mass balance gradients cannot be applied for most debris‐covered glaciers; these glaciers were also marked and excluded from ELA calculations.

The length of each individual glacier was calculated as maximum flow path length within ArcGIS. The retreat since the LIA was obtained by 3D measurements of the distance between LIA terminal mo‐ raine and the 1999 glacier snout with Google Earth’s measuring tool. To ensure equal glacier stages,

37 the same Landsat ETM+ scene was used as the measuring basis. Similar to the ELA calculations, a set of restrictions was applied to ensure the quality of data used for further analysis. Again, all glaciers that were either of the reconstituted type or were part of a larger glacier in the LIA and divided into tributaries since were omitted. Furthermore, all glaciers that did not allow precise length measure‐ ments, e.g., because of missing distinct terminal moraines or cloud cover, were also excluded. After the first tests, an additional minimum criterion of 1000 m glacier length was introduced because many smaller glaciers show very untypical geometries that result in unrepresentative length change values. In contrast to the ELA preselection, debris cover or reaching of the valley floor were not used as criteria for exclusion from the length change calculation.

4.3.4 Calculation of ELA estimates A number of indices have been developed as surrogates for extensive ELA measurements (Meierding, 1982; Torsnes et al., 1993; Aa, 1996; Benn and Lehmkuhl, 2000). Because of the remoteness of the study area and the large amount of glaciers, the evaluation and application of such indices was a critical part of the present study. In a first step, the strict glaciological approach of the balance ratio (BR; Furbish and Andrews, 1984) had to be omitted owing to the lack of mass balance data. The ac‐ cumulation‐area‐ratio (AAR; Porter, 1975) and the more complex area‐altitude balance ratio (AABR; Osmaston, 2005) were not applied because DEM artifacts prevented generating high‐quality contour lines for the whole study area, which are a fundamental prerequisite for these methods (see chapter 4.3.1). In the rugged high mountain environment, the application of AAR and AABR is also hampered by the fact that the assumption of linear accumulation/ablation gradients is questionable in many specific settings and no applicable ratios are published so far.

The ELA calculation methods that rely on hypsographic and geomorphological evidence (morphomet‐ rical approaches) are an alternative in cases where reliable data is lacking. Several of these methods have shown promising results in previous studies (e.g., Meierding, 1982; Savoskul, 1997; Lehmkuhl, 2012), specifically the toe‐to‐headwall altitude ratio (THAR), the toe‐to‐summit altitude method (TSAM; Louis, 1955), and the toe‐to‐ridge altitude method (TRAM; Höfer, 1879). Each of these meth‐ ods is based upon the assumption that the ELA lies at a fixed proportion of the altitudinal difference between highest and lowest point of the glacier. All three methods use the terminus altitude as low‐ est point, but they differ in the definition of the highest glacier limit: the THAR uses the headwall, the TSAM uses the highest peak, and the TRAM uses the average altitude of the ridgeline. For the TRAM, Höfer (1879) originally defined ‘ridge’ as ‘average altitude of the ridgeline above the ELA’. Because of the vagueness of this definition and the circular argument arising from calculating the ELA by meas‐ uring altitudes above the ELA, this classic concept was optimized. In this study the ‘average ridge altitude’ was defined as the arithmetic mean of the most prominent summits and passes in the glaci‐ er’s ridgeline. In practice, an equal number of summits and passes were measured, beginning with the highest summit and then proceeding in descending order of the highest relative elevation differ‐ ences compared to neighboring parts of the ridgeline.

Morphometrical approaches have generally been criticized for their high grade of simplification and the lack of sensible glaciological knowledge incorporated (Hawkings, 1985; Torsnes et al., 1993). The applicability of the THAR, TSAM, and TRAM methods was therefore evaluated in a comparative per‐ formance study. This was carried out on a subset of the glacier data, representing two mountain

38 ranges in the Bodui valley in the center of the study area (Fig. 4.1). It first revealed that determining glacier headwalls from Landsat imagery was not possible in most cases. An application of the THAR method was therefore not possible. In contrast, data required for the TSAM and TRAM methods could rapidly be gathered from the DEM for the year 1999 and for the LIA. For the latter, the terminal moraine altitude measured before was applied. Several studies have shown that the morphometrical methods provide the best results for small and geometrically simple glaciers with uniform area‐to‐ altitude distributions (Porter, 1975, 2001; Osmaston, 1989). Our tests support this finding and high‐ light the importance of excluding unfeasible glaciers from ELA calculation. In summary, the exclusion criteria were: debris cover, reconstitution, valley floor limitation, and detachment of tributaries (see chapter 4.3.3).

The appropriate THAR/TRAM/TSAM ratio value may differ between study areas (Benn et al., 2005). Kaser and Osmaston (2002) presented a graphic approach to determine the most appropriate THAR ratio from a set of toe and headwall measurements that can also be applied to TSAM and TRAM studies. Further ELA proxies were measured where possible. The transient snowline (TSL) is the tran‐ sition between wet snow and ice facies on the glacier surface (Lliboutry, 1964). The altitude of the TSL at the end of the ablation phase in late summer represents the ELA for one particular year (Østrem, 1975). Measureable indicators for the LIA ELA are trimlines (Ballentyne, 1997) and the max‐ imum altitudes of lateral moraines (Porter, 2001). These proxies were compared to the different calculated ELA estimate values in order to evaluate the most suitable method to subsequently be applied to the whole data set.

The inverse distance weighted (IDW) algorithm was applied to generate a continuous ELA surface from the resulting point data (Watson and Philip, 1985). Insolation effects had to be considered care‐ fully because of the low latitude and the predominant E–W striking of the ranges. Therefore, the interpolation was performed twice for different subsets of data: the first omitting the aspects south, southwest, and southeast; the second omitting north, northwest, and northeast. The resulting sur‐ faces were subsequently combined by calculating an arithmetic mean of both rasters. This procedure allowed for an approximation that retains local variance but smoothes extreme values. In order to disentangle the influence of aspect in further detail, ELA values were grouped by the predominant glacier orientations for each major mountain range in the study area. The delineation of these ranges primarily followed the valley system. However, in some cases large ranges were subdivided further because of striking changes in topography or relief, e.g., alteration of predominant ridge orientation or varying valley incision owing to petrographic and climatic differences. Based on average ELA val‐ ues for each aspect, range‐wise wind rose diagrams were created to visualize the spatial distribution.

4.3.5 Fieldwork In 2011, 2012, and 2013, three field campaigns were performed. The first two campaigns focused on the northern and southern glacier of Mt. Pulongu, an ∼ 6300‐m‐high summit in the southern range of the Yigong Tsangpo valley (Fig. 4.1). Here, the focus of the investigations was on the extent and structures of LIA moraines, the status of debris‐covered ice, and the altitudinal differences resulting from different aspects. The 2013 campaign aimed to gather regional data to ensure that the results from Mt. Pulongu can be assigned to other glaciers of the eastern Nyainqêntanglha Range. In the context of this study, the field trips provided reference data and a ground validation for the remote

39 sensing analyses. Specifically, distributions of active ice, debris cover, and the underlying landform– process interactions were studied. Furthermore, the feasibility of the delineation of LIA moraine ridges from rock color and vegetation contrasts was checked. This knowledge was subsequently ap‐ plied in the remote sensing work to ensure a correct interpretation of the geomorphological settings.

4.4 Results

4.4.1 Glacier case studies Prior to the comprehensive mapping, case studies were conducted in order to gain important insights into the study area’s glacial and morphological characteristics. This was necessary to evaluate the best practice for delineating the outlines of debris‐covered glaciers. Additionally, information on various aspects of the interplay of glaciers and climate, bedrock, topography, and vegetation was obtained. After an extensive survey, three large complex valley glaciers where chosen for a detailed case study on glacier mapping (Fig. 4.7; Fig. 4.1 shows the position within the study area). Despite their location in neighboring valleys, these glaciers show remarkably different features. In the follow‐ ing text, the three glaciers will be called (A), (B), and (C) according to the labeling in Fig. 4.7 to keep it simple.

In 1999, glacier (A) covered an area of 151.7 km² and was ∼ 39.5 km long. It retreated by 4.16 km (∼ 9.5%) since the LIA. The glacier surface shows signs of active ice flow over the whole course, i.e., ‘smooth’ debris surfaces and linear flow patterns (see also Fig. 4.5A). Only the lowest parts of the tongue are covered by debris; LIA trimlines and lateral moraines are in close proximity to the modern glacier. Glacier (B) is much smaller (area ∼ 46.4 km², length ∼ 12 km). Here, the lower part is exten‐ sively covered by debris, and the flow structures indicate active ice flow down to the modern glacier tongue. Horizontal retreat is relatively slight (1.65 km, ∼ 12%), but trimlines indicate massive down‐ wasting of up to 300 m. The debris‐covered glacier (C) shows distinct collapse structures comprising ice cliffs and ponds (see also Fig. 4.5A). Several tributary glaciers joining the mainstream on different positions of the glacier tongue are affected by the ice flow of the main glacier in different ways, indi‐ cating the absence of active ice flow in the last 7.5 km. This accounts for about 37.3% of the 20.1‐km overall length. The tributaries have obviously been detached from the ice flow of the main glacier; glacier retreat could therefore not be measured here.

The summits of all three catchments are at equal heights of 6000 to 6400 m asl; the highest summits are located in the northern ridge of glaciers (B) and (C). Catchment morphology, specifically the val‐ ley width, differs strongly: the valley of glacier (A) is the widest in the upper reaches (up to ∼ 10 km diameter ridge to ridge) but narrows to about 6 km downstream. In contrast, the valley glacier (B) widens (< 5 km up to > 10 km), and the valley of glacier (C) consists of several troughs. The late sum‐ mer TSL marking approximately the position of the ELA shows that glacier (A) has a very large and comparably flat accumulation area, whereas the accumulation areas of glaciers (B) and (C) are small‐ er and steeper. In all three catchments, the flattest parts of the accumulation areas are slightly above the observed TSL.

Fig. 4.7: Map of three selected glaciers and selected glacial and geomorphological features: (A) large, relatively flat accumulation area, ∼ 1% debris cover, little downwasting, ∼ 9.5% retreat; (B) smaller and steeper accumulation area, ∼ 7% of debris cover, up to 300 m downwasting, ∼ 12% retreat; (C) smallest and steepest accumulation area, ∼ 7.5% debris cover, up to 100 m downwasting; retreat could not be measured because of the detachment of several major tributaries. Altitudes are derived from SRTM digital elevation data. TSLs are mapped from the dark blue to light blue color gradient visible in the Landsat ETM+ scene form October 1999 (see also Fig. 4.5A).

Despite their close spatial proximity, the three glaciers show a high level of heterogeneity in terms of catchment morphology and glacial features. The results of the case studies provided crucial infor‐ mation on the local setting that was applied in the subsequent mapping and statistical analysis pro‐ cedures. In particular, the delineation of debris‐covered glaciers was based on the characteristics of active and dead ice inferred from the case study observations.

4.4.2 Glacier types and spatial distribution In the study area, a total of 1964 glaciers were mapped from the Landsat image, representing the region’s glaciated area as of late summer 1999. Out of these, 13 were classified as ice caps, 1398 as slope glaciers, 401 as cirque glaciers, 76 as valley glaciers, 61 as complex valley glaciers, and 15 as

41 glacier networks (Table 4.1). A comparison to the LIA class distribution reveals that the number of slope glaciers increased strongly (+224) whereas the number of ice caps (+2), cirques (+3), and valley glaciers (‐6) changed only slightly. Complex valley glaciers (‐152) and glacier networks decreased (‐ 71). Regarding the glacier type changes since the LIA, an overall trend from ‘higher’ toward ‘lower’ types with slope glaciers as end members can be observed. The decrease of complex valley glaciers and glacier networks represents the dissection of formerly continuous ice masses into several smaller glaciers. However, it should be kept in mind that these numbers cannot be interpreted as absolute values for the LIA stage because the data collection method only considers glaciers existing today. Thus, the numbers rather represent relative shifts in the frequency distribution of different glacier types.

Table 4.1: Comparison of numeric glacier type distribution for the LIA (columns) and 1999 (rows) stages.

LIA glacier types Sum Ice cap Slope Cirque Valley Complex Network (1999)

Ice cap 8 2 2 0 1 0 13 Slope 2 1164 118 29 73 12 1398 types

Cirque 1 8 278 24 52 38 401 Valley 0 0 0 29 35 12 76 glacier

Complex 0 0 0 0 52 9 61

1999 Network 0 0 0 0 0 15 15 Sum (LIA) 11 1174 398 82 213 86

Figure 4.8 shows the spatial distribution of glacier types for the 1999 and LIA stages based on the mapped 1999 glacier extent. It shows that the largest ice masses are located in the northwestern part of the study area, whereas in the central and western part only summits and high ridges are glaciated. In the south, the Namcha Barwa and Gyala Peri mountain massifs are the only glaciated features. A comparison of both maps reveals the dissection of ‘higher’ type glaciers, the transitions to ‘lower’ type glaciers, and the overall increase of slope glaciers.

Fig. 4.8: Spatial distribution of glacier types for (A) 1999 and (B) LIA. The black rectangle marks the area of the case study (Fig. 4.7).

4.4.3 Length change analysis The distance from LIA terminal moraines to modern glacier snout positions for individual glaciers gives an absolute measurement for glacier changes. All measured glaciers have retreated since the LIA. After the discrimination procedure (see chapter 4.3.3), 422 glaciers can be used for length change analysis. Figure 4.9A shows the absolute glacier length plotted against the relative length reduction [%]. The minimum value is at ∼ 6%, the maximum at ∼ 58.3%, and the average at ∼ 27%. However, many of the excluded glaciers with < 1000 m length showed higher values (the break at the bottom of the diagram is because of this exclusion). The plot shows that all glaciers longer than 5000 m, including the majority of debris‐covered glaciers, have retreated by max. 25% while smaller glaci‐ ers show a much higher variability, peaking close to the 1000‐m glacier length limit.

The plot of glacier area versus relative length reduction (Fig. 4.9B) shows that glaciers larger than 10 km² did not retreat by more than ∼ 25% and that there is a slight trend toward less retreat with in‐ creasing area. In contrast to the result in Fig. 4.9A, the highest relative retreat does not occur at the smallest glaciers but at those of about 0.7 to 2 km². Figure 4.9C shows that the latitudinal distribution of the sample set is similar to the overall glacier distribution in the study area with a maximum in the lower northern part (see also Fig. 4.8). A correlation between relative length reduction and latitude does not exist. For a subset of the glaciers (n = 331) both, ELA and length measurements were possi‐ ble. In order to compare rates of change, the difference between LIA and 1999 ELAs (ΔELA) was plot‐ ted against the relative length reduction (Fig. 4.9D). The result shows a nonlinear distribution with a concentration of values in the lower‐left quarter, indicating a loose connection between ELA and length changes. A rose diagram shows the relation of aspects to cumulative relative length changes (Fig. 4.9E). It is striking that north, northwest, west, and southwest aspects are all in a very close range of ∼ 27 to 28% length reduction. South (22.9%), southeast (25.2%), and northeast (24.4%) as‐ pects have less mean length reduction, whereas east facing glaciers show the strongest average re‐ treat (45.3%). Consideration of the glacier types reveals that slope glaciers (∼ 30%) and cirque glaci‐ ers (∼ 28%) retreated stronger than the different types of valley glaciers, which are all in a range of 14 to 24% length reduction (Fig. 4.9F). Except for ice caps, a general trend to less length reduction for ‘higher’ glacier types can be observed. However, in the steep high mountain environment only two glaciers in special topographic settings were classified as ice caps. Consequently, ice caps are a spe‐ cial case in the study area and shall not be compared directly to other glacier types.

Fig. 4.9: Average glacier length reduction plotted against different parameters: (A) glacier length, (B) glacier area (log. display), (C) latitude, (D) ΔELA, (E) aspect, and (F) glacier type.

4.4.4 ELA calculations After identifying and omitting all glaciers that met at least one of the criteria for exclusion (see chap‐ ters 4.3.3 and 4.3.4), a total of 688 glaciers was left that could be used for the calculation of ELAs with the TSAM and TRAM methods. The application of the graphical approach to identify adequate TSAM/TRAM values failed because, in contrast to the original study by Kaser and Osmaston (2002), a distinct linear relationship between ridge and glacier terminus altitude does not exist (Fig. 4.10). We therefore assumed a ratio of 0.5 and evaluated the quality of the results by comparison to TSL alti‐ tudes derived from the satellite image and the DEMs. Because of the Landsat scene’s acquisition date in late September, the TSL represents a marker of the approximate position of the ELA for the rele‐ vant year (Østrem, 1975). For 185 glaciers TSL altitude and ELA data were available. The TSAM and TRAM results both show significant correlations, with the TRAM reaching a slightly higher value (r = 0.847, Fig. 4.11). Calculation of average deviations between TSL altitude and ELAs results in a value of 118 m for the TSAM and 30.4 m for the TRAM. This is in agreement with results of Gross et al. (1976) that showed that the TSAM method typically overestimates the ELA by ∼ 100 m in the European Alps. Therefore, the following ELA results are based on data calculated with the TRAM. The best match between ELA and TSL is reached at a ratio of ∼ 0.39 for the TSAM and ∼ 0.47 for the TRAM.

Fig. 4.10: Plot of average ridge and terminus altitudes. Different signatures indicate different moun‐ tain ranges of the study area. Note the large scattering of the data and the few differences between different mountain ranges.

Fig. 4.11: Evaluation of ELAs calculated by (a) TSAM and (b) TRAM by comparison to TSL altitudes mapped from the Landsat ETM+ scene from 23 September 1999.

The interpolation of the calculated ELA point data using the IDW algorithm results in a map of spatial distribution of the ELAs over the study area (Fig. 4.12). It shows a distinct south–north gradient from ∼ 4600 to above 5700 m asl. The lowest ELAs are in the southeastern part in areas adjacent to the southward running Yarlung Tsangpo valley, whereas the highest values can be found in the transition zone to the Tibetan Plateau in the northwest. A trench of relatively low ELAs is roughly tracing the ENE‐WSW running Yigong and Parlung Tsangpo valleys. In the areas north of this valley, the distribu‐ tion becomes more heterogeneous, with several local minima and maxima. Many of these show a close spatial relationship to the local topography. For example, the trough at about 30° N., 96°40’ E. coincides with the entrance to the large Bodui valley and the northerly adjacent mountains. In con‐ trast, the ELA ridge at about 30°30’ N., 95°10’ E. traces a mountain ridge that is positioned at the heads of two long tributary valleys. These are running at low angles to the Yigong and Parlung Tsang‐ po valleys and are separated from them by major ridges.

Fig. 4.12: ELAs calculated for the 1999 stage. Major ridges and the river network are included to clari‐ fy topographic influences.

In order to analyze the ELA dynamics in further detail, ΔELA values were interpolated by the IDW algorithm, resulting in a map of the spatial distribution of ELA change from the LIA to 1999 (Fig. 4.13). Noticeably, all ΔELA values show an ELA‐rise since the LIA; the mean of all ΔELAs measurements is +136.4 m, the median 97.5 m. Even though an overall trend from high ΔELA values in the south and southwest to lower values especially in the northeast can be observed, the data set is dominated by local variability. In many cases small areas of strong ELA lowering are in the immediate vicinity of low ΔELA values.

Fig. 4.13: Interpolated differences between local ELA calculations (ΔELA) for the LIA and 1999 stages.

The map of range‐wise ELA wind rose plot of (Fig. 4.14) shows almost circular plots indicating quite regular distributions of ELAs by aspects over the whole study area. The strongest deviations from this scheme can be found in the central, south‐central, and southeastern ranges, of which the first two were calculated on a noticeably smaller number of glaciers (n = 21 and n = 9, respectively). Average ELAs of south‐facing glaciers are about 50 to 150 m higher than ELAs of north‐facing glaciers in most cases. An apparent exception is the second most northwestern range where north‐orientated glaci‐ ers have ∼ 160‐m higher average ELAs. In contrast, the strongest depression of north aspects in comparison to south (296 m) is displayed in the southeastern diagram.

Fig. 4.14: Rose diagrams for the calculated ELAs in different ranges of the eastern Nyainqêntanglha. Red lines indicate the 1999 and violet lines the LIA stage ELAs, respectively.

4.5 Discussion

4.5.1 Quality assessment Quality assessment played a critical role in all stages of the study. The detailed study of glacier set‐ tings primarily improved the understanding of the properties and influences of debris cover and catchment morphology. This was a prerequisite for a proper delineation of debris‐covered glacier tongues. Each glacier was investigated twice because the mapping process was split into a first phase of glacier delineation and a second phase of parameterization. Morphological and glaciological evi‐ dence, e.g., trimlines and TSL, provided further control of the results. The exclusion of unsuitable glaciers was an essential prerequisite for the ELA calculations. In this regard, the case study helped to identify criteria that prevent accurate ELA calculations, such as debris cover, ice avalanching of the glacier snout, detachment of tributaries, and limitation of the altitudinal oscillation of glaciers by reaching the valley floor. Furthermore, the relatively large data set had positive effects: smaller measurement errors were ruled out by averaging of the large data set and larger errors were identi‐ fied as outliers, checked, and corrected if necessary. The comparison to TSL altitudes provides a benchmark for the quality of the ELA calculations, demonstrating that both methods, TSAM and TRAM, provided reasonably good ELA calculations. Altogether, the TRAM performed better: with a ratio of 0.5, the TRAM’s average deviation from the TSL is ∼ 30 m, which is within the range of error of the DEM used for data acquisition. The TSAM method overestimates ELAs by more than 100 m,

50 which is consistent with the results of Gross et al. (1976). In order to assess the quality of the TSL as a marker for the 1999 ELA, an evaluation of the climatic and glaciological circumstances was important. Assuming a slight warming trend for the ending of the twentieth century, the TSL would adapt to changes each year whereas the morphological features on which the ELA calculation bases, i.e., the glacier snouts, react slower. At the end of the ablation period, the calculated ELAs should therefore be lower than the TSLs. However, taking together the high ice velocities in the study area (Zheng et al., 1999) and the moderate climate conditions in late summer 1999 (Fig. 4.3), we can conclude that the late summer TSL should be in close proximity to the ELA. The climate data (Fig. 4.2) supports the assumption that the end of the ablation season for 1999 was in October rather than in late Septem‐ ber. Hence, the calculated ELA values should be slightly above the TSL, and thus the application of the TRAM with a ratio of 0.5 seems to be valid.

4.5.2 Influences of topography and climate Several conclusions regarding the glacier–topography–climate interactions can be drawn from the distinct differences between the three glaciers that were analyzed in the case study (Fig. 4.7). Glaci‐ ers (B) and (C) are showing a strong retreat despite their high summits. Both have small and relative‐ ly steep accumulation areas. In contrast, the large accumulation area of glacier (A) is promoting rela‐ tively stable conditions. This suggests that the absolute elevation of the highest summit is of less importance for glacier dynamics than the size of the area above the long‐term highest ELA. Consider‐ ing the coincidence of accumulation and ablation in the study area, this implies that temperate mon‐ soonal glaciers with steep and relatively small accumulation areas are severely affected by climate warming. First, the temperature increase leads to rising ELAs and subsequent reduction of the accu‐ mulation area. The morphologies of the three glaciers in the case study indicate that in many cases the absolute loss of accumulation area will be disproportionately large owing to the aggregation of relatively flat areas slightly above the ELA (see chapter 4.4.1). Second, a larger portion of the incom‐ ing precipitation is falling as rain instead of snow. Consequently, less material is available for glacier accumulation. Furthermore, melting of glacier ice is increased through lowered albedos and the ef‐ fects of liquid water in the glacier system. Similar effects have been recently observed by Benn et al. (2012) for glaciers in the Mount Everest region. Whether tributaries can connect to the main glaciers or not primarily depends on valley width in the ablation zone. The steepness of the accumulation area controls the rate of avalanching and thus the amount of debris introduced into the glaciers. Thick debris cover protects glacier ice from melting and therefore may conceal glacier dynamics (Yang, W., et al., 2010b; Scherler et al., 2011b). For example, glacier (B) retreated by only 10%, less than glacier (A); but the massive downwasting of up to 300 m indicates strongly negative mass bal‐ ances. We can assume that when the downwasting reaches the valley floor the glacier tongue will rapidly collapse. This seems already to be the case at glacier (C), where the debris‐covered ice body shows numerous signs of disintegration.

The regional distribution of glaciers in the study area indicates that large ice masses are typically located in high mountain ranges that are in the vicinity of major valleys running approximately in S–N direction. This could be explained by the major source of snow accumulation in the region, the ISM. It propagates northward through the main valleys until intercalated E–W running ridges lead to heavy orographic rainfalls. The distinctively alternating late summer temperature and precipitation patterns (Fig. 4.3) reveal the substantial impact of monsoonal rainfall on the glaciers: an intensive monsoon is

51 leading to intensive and persistent precipitation and at the same time reduces insolation because of enhanced cloud cover. Thus, both relevant factors for glacier mass balances, temperature and precip‐ itation, are coupled during summer monsoon. Considering the coincidence of accumulation and abla‐ tion phases in the study area, this implies that temperate monsoonal glaciers are also affected in more than one way by changes of monsoon intensity.

In the ΔELA map many local effects are preserved, resulting in a heterogeneous pattern after the IDW interpolation (Fig. 4.13). This is in contrast to the ELA maps, in which the strong insolation dif‐ ferences between different aspects were smoothed by generating a mean surface from separate data sets for north and south. Consequently, more local variability is preserved in the in the ΔELA map than in the ELA map. This variability shows the limits of the TRAM, i.e., that it is not possible to retrieve exact ELAs for each glacier. In addition, the ΔELA map shows the most detailed picture of local effects obtained in this study: in many cases, the direct neighboring of high and low values is predominantly caused by shielding effects of summits and ridges owing to different aspects. As men‐ tioned before, detailed interpretations of each local situation have to be considered individually be‐ cause shielding effects can apply to insolation and to precipitation. As a consequence of this substan‐ tial influence of the local topography, an earlier concept to derive an AAR from the interpolated TRAM data was dismissed. The attempts to apply the calculated TRAM ELAs to individual glaciers revealed strong heterogeneity concerning the fit of ELA levels to evidence in the satellite imagery, i.e., TSL, including good matches, over‐ and underestimation in close proximity. This highlights that (i) the calculated ELAs must be interpreted as medium scale average and that (ii) the deduction of mountain range‐wide AAR and BR ELAs from measurements on single glaciers can lead to critical errors, at least in subtropical high mountain environments.

4.5.3 Glacier change All results show that the investigated glaciers in the eastern Nyainqêntanglha Range have retreated since the LIA. However, the detailed glacier type, length change, and ELA analyses reveal a very het‐ erogeneous picture of different forcings and subsequent reactions.

A general trend toward less relative length reduction for longer (Fig. 4.9A) and larger (Fig. 4.9B) glaci‐ ers can be observed. Several studies report similar findings for other glaciated ranges in High Asia (e.g., Bolch, 2007; Bolch et al., 2010). Longer response times of larger glaciers are frequently identi‐ fied as the main cause for these size‐dependent differences (e.g., Gross et al., 1976; Kulkarni et al., 2007). The protective influence of the debris covers present at many of the larger glaciers in the study area probably intensifies this effect. However, the data shown in Fig. 4.9B indicates that length reduction does not increase uniformly with decreasing glacier size but that length reduction peaks at about 0.7 to 2 km² in the study area. In many cases this seems to be a result of changes in morpho‐ logical glacier types: ongoing glacier retreat leads to the degeneration of glacier tongues, resulting in oval‐ to sickle‐shaped slope glaciers in near‐ridge positions. Here, a further ELA‐rise often affects the glacier as a whole, leading to rather uniform shrinkage that is not properly represented by the length change measure. Generally, the analyses of glacier types show a succession starting with large valley glacier networks that get separated into tributaries, typically valley glaciers and cirques. Continuing retreat leads to slope glaciers on the upper reaches of the mountains. Further negative mass balance then results in the extinction of glaciers. Many settings, especially in the northeastern part of the

52 study area, show morphologic evidence of LIA glaciers that vanished and were therefore not investi‐ gated further in this study.

The results of aspect analysis reveal that glaciers facing north and west are generally reacting quite uniform, whereas south‐ and east‐facing glaciers deviate from these values (Fig. 4.9E). The observa‐ tion that south‐facing glaciers show the smallest average retreat rates indicates that another factor must play a key role for the mass balance. The length reduction rates of glaciers facing east are above average. This observation is especially remarkable because it runs against the overall trend toward lower length reduction for northeastern to southern aspects.

Several distinct climatic and topographic north–south gradients have been observed in the study area: the mean annual precipitation is decreasing from south to north, average altitude is rising, and the relief is getting flatter toward the Tibetan Plateau. However, none of these parameters seems to have a sufficiently strong impact on glacial dynamics to leave a clear footprint in the correlation of length change and latitude (Fig. 4.9C). This highlights the importance of the local effects of topogra‐ phy, glacial geomorphology, and climate to the glacier dynamics. The same seems to apply for the correlation of ΔELA and glacier length change (Fig. 4.9D), indicating that different morphological glac‐ ier and catchment characteristics can lead to heterogeneous glacier reactions to similar changes in mass balance.

The calculated ELAs show the influences of overall topographic and climatic gradients and local set‐ tings. A general ELA rise of about 1000 m from the southeast to the northwest of the study area can be observed for both investigated time slices, 1999 and LIA. For 1999, the lowest ELAs are at ∼ 4400 m asl in the southeast and the highest ELAs are at slightly above 5600 m asl in the north‐ western part of the study area. The results of Shi and Li (1981) and the presented results show simi‐ lar lowest ELA values, whereas the maximum ELAs are about 400 m higher in this study (∼ 5200 vs. ∼ 5600 m asl). Taking into account that their early assessment was based on aerial photos and maps from the mid‐twentieth century, a recent ELA rise for the northwestern part of the study area and rather stable ELAs in the south for the last 50 to 60 years might be deduced. The values provided by Yao et al. (2010) for ELAs in the Yigong (3370–5490 m asl) and Parlung (3560–5980 m asl) Tsangpo valleys seem to substantially underestimate the lowest ELAs. Consideration of the very small scale of the resulting maps and the problems concerning the interpretation of the early aerial photo maps used by Shi and Li (1981) and the GIC used by Yao et al. (2010; see also Fig. 4.4) implies that the data and methods used in our study improved the quality of measurement and the level of detail. The average post‐LIA ELA rise of about 100 m is slightly higher than mid‐latitude values (e.g., Maisch, 2000: 90 m for the European Alps) but lower than tropical values (e.g., Rabatel et al., 2005: tropical Andes 160 m). Kayastha and Harrison (2008) reported an ELA rise of 80 m for the Nepalese Himalaya for a period from the LIA to 1959 but point out that further glacier retreat has occurred since.

4.5.4 Monsoon and glaciers In accordance with the findings of the glacier distribution analysis, the spatial patterns of ELAs sup‐ port the assumption that monsoonal air masses from the Bay of Bengal are mainly entering the study area from the south via the Brahmaputra‐Yarlung Tsangpo valley. In the study area they are propa‐ gating farther north, predominantly along the major valley systems. Relatively low ELAs occur in are‐ as where these valleys encounter major obstacles, i.e., high, approximately E–W running ridges. Rela‐ 53 tively high ELAs can be found in settings where orographic features have shielding effects or the course of valley systems hampers the northward‐streaming trends. The most prominent example is the major ESE–WNW running mountain range in the northwestern part of the study area where oro‐ graphic rainfalls and shielding effects lead to lower ELAs on southern than on northern slopes (Figs. 4.12 and 4.14). Both effects can also be observed at the catchment scale, indicating a good perfor‐ mance of the method in depicting the overall trend and local effects, respectively.

In the southeastern part of the study area several features suggest that precipitation is also coming from eastern directions. The remarkably clear imprint of the topography in the spatial distribution patterns of the ELA indicates that the bulk of precipitation‐relevant monsoonal moisture is being transported in altitudes that are constrained by the relief in the study area, i.e., well below 6000 m asl.

For southern aspects, neither the distribution of length change nor ELAs for different aspects show the expected patterns (Figs. 4.9E and 4.14). Considering the increase of solar radiation since the LIA (Solanki et al., 2004; Muscheler et al., 2007), the general warming trend for the northern hemisphere (Mann and Jones, 2003; IPCC, 2013), and the twentieth century warming trend on the Tibetan Plat‐ eau (Yang, B., et al., 2010), one may expect that south‐facing glaciers show the strongest negative length change and the strongest ELA rise in reaction. The observations however are contrary, show‐ ing the least length reduction for southern aspects (Fig. 4.9E) and a comparatively small ELA differ‐ ence between northern and southern aspects (Fig. 4.14). Applying the principle of exclusion for topographic and large‐scale climatic influences, we can deduce that these effects originate from the monsoonal system. In consequence, this indicates that the monsoon system must have been highly dynamic since the LIA and still has substantial influence on glacier mass balances in the study area. Notably, Kayastha and Harrison (2008) observed that glaciers in the Himalaya Range of eastern Nepal do indeed show highest ΔELA values for south‐facing glaciers.

Knowledge about the spatiotemporal variability of the summer monsoon system in the eastern Nyainqêntanglha Range is still sparse. Until now, only one proxy data set on monsoon intensity avail‐ able for the study area. Liu et al. (2013) found that δ18O values in tree rings from sites near Bomi (29°26′ N., 96°29′ E., 4150 m asl) are decreasing since 1700 CE. The authors interpret this observation as a signal of decreasing ISM intensity owing to a decreasing thermal contrast between the Bay of Bengal and the Tibetan Plateau. However, their data shows a δ18O minimum in the late nineteenth century and a slight increase until today. Grießinger et al. (2011) studied δ18O in tree rings near Ret‐ ing (30°18.5′ N., 91°31′ E., 4100–4200 m asl) in the southern part of the Tibetan Plateau. Their find‐ ings indicated an increase in monsoonal precipitation from the fifteenth to the early nineteenth cen‐ tury, followed by a drop to a minimum in the mid‐twentieth century and only a slight increase from there on. A similar trend can be observed in speleothem records from Dongge cave in southern China (25°17’ N., 108°5’ E., 680 m asl; Dykoski et al., 2005; Wang et al., 2005) and several other speleothem records from central and eastern China (Cai et al., 2010), which are all supposed to reflect the dy‐ namics of the EASM. The only high resolution monsoon record available located directly south of the study area, a speleothem record from Wah Shikar cave in northeastern India (25°15′ N., 91°52′ E., 1290 m asl) shows a different trend: here, where the ISM branch from the Bay of Bengal propagates toward the study area, strong monsoonal activities are being reported from the beginning of the data

54 set at ∼ 1400 CE to 1650 CE followed by a decreasing trend until about 1750 CE and a new equilibri‐ um of weaker monsoon from then on (Sinha et al., 2011). In contrast, ice core data from the Dasopu glacier in the Himalaya main range (28°23’ N., 85°43’ E.) shows an increasing δ18O trend, which is, however, interpreted as a temperature signal by the authors (Thompson et al., 2000).

The fact that none of the records can explain the observed glacier changes straightforward supports the hypothesis that the eastern Nyainqêntanglha Range is a transition zone of different monsoonal influences. This is substantiated by a comparison of the two meteorological stations in the study area (Fig. 4.2): precipitation data from Nyingchi in the west is showing a monomodal course, whereas Bomi in the east is characterized by a bimodal precipitation regime. The latter indicates the influence of EASM and ISM in the Bomi area. The remarkably strong length reduction of east‐facing glaciers may indicate a decreasing influence of the EASM in the study area toward the west. This is in accord‐ ance with the orographic rainfall and shielding effects observed in the southeastern part of the ELA map (Fig. 4.12; see also chapter 4.5.2) and the depression of northern‐aspect ELAs showing in the southeastern wind rose plot (Fig. 4.14). In summary, these indications of monsoonal patterns suggest that a slight EASM influence exists in the study area. However, in contrast to the ISM, which seems to affect the whole investigated area, the indications of EASM are limited to the eastern part. A sub‐ stantial source of glacier ice accumulation in western to northern directions cannot be observed in the results of this study. Consequently, precipitation transported by westerly winds seems to be of less importance in the study area than in the Himalayas and on the Tibetan Plateau (Benn and Owen, 1998; Mölg et al., 2014). The results show that the local precipitation dynamics are mostly orograph‐ ically induced owing to rugged high mountain topography and the position at the transition zone between the Indian subcontinent and the Tibetan Plateau. Arguably, glacier change in the study area does not directly correspond to the surrounding palaeoclimate proxy records because of this high grade of topoclimatic variability.

4.6 Conclusions The approach presented in this study allowed for efficient data collecting of the modern and LIA states of glaciation in the remote eastern Nyainqêntanglha Range. First, detailed glacier mapping and parameterization provided a sound data basis, revealing numerous clues to the principles of the cli‐ mate–relief–glacier interactions. Subsequently, statistical methods were applied to the large data set (n = 1964) in order to determine the importance of individual variables and their interrelations. Final‐ ly, ELA calculations provide an overview of spatial distribution gradients and a quantitative measure of changes through ELA depressions.

The utilization of the remotely sensed late summer TSLs proved to be particularly useful for the eval‐ uation of different ELA calculation methods and their results. In this regard, special attention should be paid to the climatological and glaciological interpretation of the TSL. The use of the TSL as a proxy for the ELA is only valid if measured at the end of the ablation phase; if this is not the case, the TSL will be below the ELA. A prerequisite for TSL‐based evaluations of morphometrically calculated ELA approximations is that the climatic conditions of the year under consideration for TSL measurement do not deviate too heavily from the multiannual average. For the study area, the evaluation of the

TRAM results using climate data and TSL altitudes showed that ELA calculations yield good results by application of a 0.5 ratio.

The details of spatial ELA distributions could only be identified by the large number of glaciers that were analyzed by individual‐glacier empirical data. In the present study, this was a great advantage of the crude TSAM and TRAM methods that allows for the rapid collection of comprehensive data sets. This implies that they may perform better than AAR and BR methods in complex high mountain relief settings with only few empirical glacier data available.

The results clearly show that glaciers in the study area have retreated strongly since the LIA. Statisti‐ cal analysis of glacier length change revealed an average retreat of ∼ 27% and a trend toward stronger retreat for smaller glaciers. The interpolated ELA results show a southeast–northwest gradi‐ ent ranging from 4400 to 5600 m asl and an average ELA rise of ∼ 136 m since the LIA. Contrary to expectations for subtropical settings, glaciers on south‐facing slopes have not retreated strongest and ELAs on south‐facing slopes did not rise the farthest. Instead, highly heterogeneous spatial pat‐ terns emerge, showing a strong imprint of topography and of monsoonal dynamics. The influence of the relief situation proved to be of particular high relevance in different scales: regional, catchment, and individual glacier scales. In many settings the imprint of local orographic rainfall and shielding effects on late Holocene glacier dynamics exceeds the overall influence of the regional topographic and climatic gradients. The spatial distribution patterns of the calculated ELAs show the controlling influence of the major valleys and mountain ridges on the monsoonal air masses. This also implies that the majority of monsoonal moisture is being transported in altitudes below the average ridge altitudes in the study area, i.e., well below 6000 m asl. Distinct gradients in ELAs and subsequently in precipitation and temperature emphasize the region’s role of a mixing zone of different influences, including the ISM and EASM monsoon currents and the transition from foreland to high altitude plat‐ eau topography. However, indications for EASM influence are limited to the southeastern part of the study area, implicating that it does not reach the Tibetan Plateau.

The synoptic interpretation of the observed glacier–relief–climate interactions revealed indications for two ‘double impact’ couplings. (i) Because of the steep high‐mountain relief, many glaciers in the study area have small and steep accumulation areas with an aggregation of relatively flat areas slightly above the ELA. For these cases the reduction of accumulation area due to climate warming will be disproportionally large. Furthermore, a larger portion of the incoming precipitation is falling as rain, leading to lowered albedos and additional thawing. (ii) Owing to the coincidence of accumu‐ lation and ablation phases in summer, changes in monsoon intensity affect glacier mass balance through precipitation and temperature (cloud cover). In combination with the knowledge on the high ice flow velocities (Zheng et al., 1999), the presented findings support the frequently expressed the‐ sis that glaciers in the eastern Nyainqêntanglha Range are highly sensitive to climate change.

The presented results also show that the dynamics of the glacier–relief–climate interactions are ex‐ tremely complex. A better understanding of the glacier and climate dynamics in the eastern Nyainqêntanglha Range would contribute important information about the late Holocene and mod‐ ern climate system for the whole area, from the southern and eastern forelands to the Tibetan Plat‐ eau. The overview provided in this study highlights several research fields in which further empirical

56 data need to be gathered in order to disentangle the details of the complex glacier–relief–climate interactions:

‐ detailed glaciological and glacial geomorphological studies (process understanding, mass balance data); ‐ glacial chronologies (timing and extent of glacier retreats and advances); ‐ high altitude meteorological stations (modern precipitation and temperature data); and ‐ monsoon proxy data, i.e., ice cores, speleothems, or tree‐ring parameters (hydroclimatic, precipi‐ tation, and temperature data).

Furthermore, measurements of snout positions in recent satellite imagery, e.g., in Google Earth, indi‐ cate that the trend of glacier retreat continued since 1999. Therefore, a follow‐up survey with up‐to‐ date data, e.g., a scene from the new Landsat 8 mission, is highly recommended.

5 Glaciers and equilibrium line altitudes of the eastern Nyainqên‐ tanglha Range, SE Tibet

David Loibla, Frank Lehmkuhla a RWTH Aachen University, Department of Geography, Templergraben 55, 52062 Aachen, Germany

A set of maps that illustrate the key results of a study on glaciers and late Holocene glacier change in the eastern Nyainqêntanglha Range, southeastern Tibet, is presented. The cartographical challenge was to present this information in a way that focuses on the most important findings but retains as many details as possible. The layout is based on two main maps showing two different sets of infor‐ mation for the same area; one map of glacier types and distribution and one map of calculated ELAs. Additionally, two inset maps provide a regional topographic overview and information on the ELA change since the Little Ice Age maximum glacier advance. By this approach, the viewer is enabled to combine the different sets of information individually for different parts of the study area. Insights into the complex couplings of climate, relief, and glaciers can thus be obtained. The maps reveal highly heterogeneous spatial patterns of glacier and ELA distributions. These reflect the interplay of the high‐mountain topography and the primary moisture sources, the Indian Summer Monsoon and the East Asian Summer Monsoon. Specifically, luv‐lee effects and the channeling of monsoonal air masses through the valley system can be observed at catchment to regional scales. The patterns of ELA change since the Little Ice Age indicate that glacier sensitivity to climate change is decreasing toward regions with increasingly continental climate in the eastern Nyainqêntanglha Range.

5.1 Introduction The remote eastern Nyainqêntanglha Range in southeastern Tibet is located in a transition zone be‐ tween warm‐wet subtropical and cold‐dry plateau climate conditions (Maussion et al., 2014). In this high mountain environment, intense monsoonal precipitation during summer supports numerous temperate glaciers despite the latitude of ∼ 29° to ∼ 31°N. Due to the outstanding importance of the monsoonal air masses for the water cycle of the whole region, it is a key area to study climate change and glacier dynamics in High Asia. These glaciers act as a long term water storage and source for the major rivers of South‐ and Southeast Asia supplying water for more than one billion people (Im‐ merzeel et al., 2010).

Recent large‐scale remote sensing studies have shown that glaciers in the eastern Nyainqêntanglha Range are highly sensitive to climate change and retreated strongly in the last decade (Gardelle et al., 2013; Neckel et al., 2013). However, climatic records and glacier mass balance measurements are scarce in the study area. It is therefore a key challenge to improve the understanding of the coupling between glacier and climate changes in this region. Data on past glacier changes and their linkage to climate potentially provide fundamental insights into the spatio‐temporal dynamics of the complex

58 climate‐relief‐glacier interactions (Owen and Dortch, 2014). A key measure for the state of glaciers is the ELA, i.e., the theoretical line at which accumulation and ablation are in balance (Porter, 1975; Benn and Lehmkuhl 2000). Several studies have demonstrated the possibility of deriving reliable es‐ timations of present and former ELAs by methods that are based on measurements of geometric parameters of glacial geomorphological features (e.g., Meierding, 1982; Savoskul, 1997; Heyman, 2014). However, the possible effects of local topography and climate have to be considered carefully (e.g., Evans, 1990; Warren, 1991; Anderson et al., 2014). Multi‐temporal ELA calculations provide a promising tool to obtain information on glacier change and regional climate patterns, e.g., atmos‐ pheric circulation patterns and moisture sources (Benn and Lehmkuhl, 2000).

In the context of ELA calculations for the late Holocene, the maximum of the LIA glacier advance is of particular high relevance. This worldwide cooling period began in the 16th century and lasted until the end of the 19th century (Grove, 2003). Its maximum thus represents the last major turning point in glacial development, switching from an advancing to a retreating regime. Theoretically, all glaciers were in a steady‐state at their LIA maximum advance, i.e., ice mass and geometry were in equilibri‐ um with climate (Benn and Lehmkuhl, 2000). Taking into account different glacier response times due to size, topography, etc., the ELA of the LIA maximum represents an important reference level for the evaluation of post‐LIA glacier changes (Maisch, 1981; Anderson et al., 2014). In the study ar‐ ea, dendrochronological studies have shown that vegetation succession delineates moraines of the LIA maximum glacier advance (Bräuning and Lehmkuhl, 1996; Bräuning, 2006). It is thus possible to determine the LIA maximum glacier extent from satellite imagery.

The map presented here is the result of a study in which 1964 glaciers in the region were mapped by remote sensing from a Landsat ETM+ scene from September 1999 (Loibl et al., 2014). The resulting dataset was subsequently parameterized by DEM‐based measurements, e.g., of glacier terminus, summit, and ridge altitudes, and glaciological attributes, e.g., glacier types. Geomorphological evi‐ dence allowed for the determination and measurement of terminus positions of the LIA maximum glacier advance. The large data set provided the basis for calculating ELA estimations with an opti‐ mized TRAM method for both stages, 1999 and LIA.

The key cartographical challenge was to present this wealth of information in a way that focuses on the most important findings but retains as many details as possible. A layout with two main map frames showing different information sets for the same area was chosen. Two additional inset maps provide a topographic overview of the wider area and insights to the late Holocene ELA changes.

5.2 Study area The eastern Nyainqêntanglha is an east‐west oriented mountain range that stretches from the high plateau of Tibet in the West to the Great Gorges in the East (Fig. 5.1). The core study area covers ∼ 35 x 103 km² in the central part of the range. Its outlines are prescribed by the Landsat ETM+ scene that served as basis for the mapping (Fig. 5.2, see also 5.3.1). The region is a high mountain environ‐ ment with numerous summits > 6000 m asl and a maximum elevation of ∼ 7800 m asl at Namcha Barwa (Fig. 5.1). Average ridge altitudes are increasing from ∼ 4500 m asl in the southeast to ∼ 5700 m asl in the main range of the eastern Nyainqêntanglha. As a consequence of intense tectonics relat‐ ed to the eastern syntaxis of the Himalaya orogen, great fault systems are apparent which prescribe 59 the major valley system (Burg et al., 1998; Seward and Burg, 2008). Many of these valleys are deeply incised, resulting in a rugged topography with typical catchment reliefs of 2500 to 3500 m (Finnegan et al., 2008). Maximum relief in the study is at the eastern side of Namcha Barwa, where the drop from the highest summit to the Yarlung Tsangpo–Brahmaputra valley floor is ∼ 6800 m within a hori‐ zontal distance of ∼ 30 km (Fig. 5.1).

Fig. 5.1: Overview of the study area in southeastern Tibet.

In summer, the local climate is controlled by warm‐wet monsoonal air masses entering the study area predominantly from the south through the Yarlung Tsangpo–Brahmaputra valley (Yang, B., et al., 2008; Molnar et al., 2010). The monsoonal air masses convey large amounts of moisture and make the eastern Nyainqêntanglha Range the region with the highest average precipitation rates of the entire Tibetan Plateau (Zhou et al., 2007). More than 80% of the annual precipitation occurs from June to September (Zhu et al., 2013). During the winter month the westerlies dominate the circula‐ tion patterns, conveying mainly cold‐dry air masses from the Tibetan Plateau.

Fig. 5.2: The study area in the Landsat ETM+ scene from 23 September 1999 which was used for mapping (channels 5‐4‐3). Note the vegetation gradient form south to north, the partial cloud cover and the absence of fresh snow. Background hillshade is based on SRTM data (Jarvis, Reuter, Nelson, & Guevara, 2008).

For the glaciers in the area, the climatic situation poses a coincidence of accumulation and ablation phases during the summer months. The glaciers have a maritime temperature regime and show re‐ markably high ice flow velocities (Zhou et al., 1991). For two glaciers in close proximity to the study area Zheng et al. (1999) measured flow rates of 438 m/yr (Azha glacier, 96°06.06′E, 29°32.05′N) and 533 m/yr (Minyong glacier, 98°41.010E, 28°26.330N). In contrast, the climatic conditions make win‐ ter a phase of relative stability for the glaciers. Wagnon et al. (2013) measured no winter accumula‐ tion for monsoonal glaciers in Nepal.

Large moraines encompassing the paraglacial areas are abundant in the eastern Nyainqêntanglha Range. Several case studies have focused on mapping and dating these landforms at different glaci‐ ers, strongly supporting LIA age of formation (e.g., Bräuning and Lehmkuhl, 1996; Bräuning, 2006; Zhu et al., 2013; Xu and Yi, 2014).

5.3 Methods

5.3.1 Data and glacier mapping For glacier mapping, a late summer satellite image scene is required to ensure only multi‐annual ice is mapped (Paul et al., 2010). Furthermore, cloud cover and fresh snow are not desirable because they may obscure glacier outlines. In the study area, monsoonal cloud cover is usually dense until the end of September and the last precipitation events of the monsoon seasons are often light snowfalls in November. Consequently, potential scenes are limited to a small time frame around October. Fur‐ thermore, the failure of the Landsat 7 Scan Line Corrector made all scenes collected after 31 May 2003 unusable for comprehensive scientific surveys. The scene from 23 September 1999 offers a good compromise with very few snow cover on unglaciated surfaces and less than 10% clouds. Most mapping was performed using a composite of channels 5‐4‐3 pansharpened to 15 m resolution (Fig. 5.3).

The glacier mapping procedure was split into three separate phases. In the first phase, ice‐covered surfaces were delineated in an expert‐driven manual mapping process. Automation of the mapping procedure as discussed by Bhambri and Bolch (2009) was not possible due to the challenges posed to the delineation of glacier outlines by supraglacial debris or cloud cover in many settings. Cloud‐ covered sections of glacier outlines were bridged by straight lines in accordance to the instructions of the world glacier monitoring service (Müller et al., 1977). Debris cover obscures the outlines of glaci‐ er tongues because its spectral characteristics are very similar to those of surrounding materials, such as lateral moraines and bedrock (Nicholson and Benn, 2006). In order to map glacier extents in these settings correctly, typical debris‐cover‐related structures were examined in a case study (Fig. 5.3, see Loibl et al., 2014, for further details). This included flow patterns, water discharge structures, melting ponds, and collapse structures. Even though the insights from this case study helped with the interpretation, a relatively high level of uncertainty remained in debris‐covered glacier settings (see also Scherler et al., 2011b). In most clean ice settings, ice margin and moraine positions could be easily identified (Fig. 5.3). According to Nyquist sampling theory, however, the Landsat ETM+ resolu‐ tion of 15 m does not allow mapping of elements < 30 m (Arrell et al., 2008). In the study area’s complex high mountain environment, the detectable minimum object size is often larger, because the effects of shading, different lithologies, different vegetation, etc., are often exacerbating object identification.

Fig. 5.3: Mapping concepts for glaciological and geomorphological features as applied in this study.

The second phase of the mapping workflow obtained a shapefile of ridgelines. These were extracted automatically using ArcGIS’ stream delineation tools on a hole‐filled inverse SRTM DEM with 90 m resolution (Jarvis et al., 2008). The result was then corrected by matching it to ridge areas in a topo‐ graphic position index raster (Jenness, 2006) and checking it manually using ASTER GDEM2 (30 m resolution) and Landsat ETM+ data. The generated ridgelines allowed for cutting the ice‐cover shape‐ file into individual glaciers.

The final phase of the mapping process was the parameterization of the dataset. This included sever‐ al DEM‐based measurements such as summit, ridge, and glacier terminus altitudes. For glaciers that fulfil the criteria to be used for ELA calculation (see chapter 5.3.3), the position of the terminal mo‐ raine of the LIA maximum glacier advance was determined by remote sensing from geomorphologi‐ cal evidence (Fig. 5.3). At most glaciers in the study area, the LIA terminal moraines are well devel‐ oped and can be easily identified in the Landsat ETM+ image. Furthermore, in agreement with the findings of Bräuning (2006), at many glaciers the outer walls of LIA moraines are covered by forests, whereas the inner walls and the paraglacial areas within the LIA moraines are almost free of vegeta‐ tion. In cases where no clear terminal moraines were detectable, either because they were not de‐ veloped or because the Landsat ETM+ data did not allow identifying them precisely, no measure‐ ments of LIA terminal moraine altitude were conducted. Otherwise, the altitudes were obtained from the DEM data. For a total of 862 glaciers LIA terminal moraine altitudes were measured, includ‐ ing all 688 glaciers that were finally used for ELA calculation and 174 that were sorted out in later phases of the ELA quality evaluation (see chapter 5.3.3).

5.3.2 Glacier classification Glacier morphology and topographic position have a substantial influence on glacier dynamics, i.e., accumulation, ice flow, and ablation, and are thus highly relevant for ELA calculation (Benn and Lehmkuhl, 2000). Therefore, the mapped glaciers were classified based on a differentiation that focuses on these aspects. The classification concept is mostly in line with the standard concepts for glaciers constrained by topography (Rathjens, 1982; Benn and Evans, 2010; Cogley et al., 2011). However, some details differ because of the focus on ELA‐relevant glacier morphology. The four basic glacier types used in this study are (Fig. 5.4):

(1) plateau glaciers on generally flat summit tracts, (2) slope glaciers in open slope positions, (3) cirque glaciers in armchair shaped bedrock hollows, and (4) valley glaciers where iceflow reaches the valley bottom.

In order to distinguish between different types of large valley glaciers, two additional categories were introduced. ‘Complex valley glaciers’ consist of one main valley glacier and one or more smaller tributaries, e.g., cirque or slope glaciers. Cases in which several valley glaciers join were defined as ‘valley glacier networks’.

Fig. 5.4: Typology of glaciers used in this study: (1) ice cap and ice fields, (2) slope glacier, (3) cirque glacier, (4) valley glacier. Brown material symbolizes the debris‐covered tongue of many valley glaci‐ ers in the study area.

5.3.3 ELA calculation In the high‐mountain environment of the eastern Nyainqêntanglha Range neither reliable glacier mass balance data nor high‐quality DEMs are available. Several sophisticated methods to calculate ELAs could therefore not be applied. For example, the AAR (Porter, 1975) and the AABR (Osmaston, 2005) both require reasonably accurate DEMs and the BR (Furbish and Andrews, 1984) method is based on reliable mass‐balance measurements. Several simplifying morphometrical approaches al‐ low for ELA calculations from elevation measurements: the THAR (Richmond, 1965), the TSAM (Louis, 1955), and the TRAM (Höfer, 1879). All of these methods are based on the assumption that the ELA is at a fixed portion between the highest point and the lowest point of a glacier. As indicated by the methods’ names, the differences between the concepts regard the definition of the highest point of glacier (Fig. 5.5). For the TRAM method, the ‘average ridge altitude’ was defined as the arithmetic mean of the most prominent summits and passes in the glacier's ridgeline (Loibl et al., 2014). The values were obtained through manual DEM‐based measurements.

Fig. 5.5: Schematic diagram of the three morphometric approaches of ELA calculation: toe‐to‐summit altitude method (TSAM), toe‐to‐ridge altitude method (TRAM), and toe‐to‐headwall altitude ratio (THAR). The arrows indicate the approach of calculating an average ridge altitude as the arithmetic mean of an equal number of peaks and passes from the back wall.

A comparative evaluation of the three methods first showed that the THAR could not be applied be‐ cause the headwall cannot be identified in the Landsat ETM+ imagery. In contrast, the TSAM and TRAM allowed for collecting data rapidly. A more detailed comparison including ELA‐proxies such as the late summer glacier snowline (transient snowline, see also Fig. 5.3) revealed that the TRAM pro‐ vides slightly better results and that a TRAM ratio of 0.5 fits well (Loibl et al., 2014). Consequently, the TRAM method was applied using a 0.5 ratio. However, the quality of the calculations was highly dependent on the choice of appropriate glaciers. Therefore, a set of criteria was defined which regu‐ late the exclusion of glaciers from the ELA calculation. Specifically, glaciers had to be omitted which have glaciological or geomorphological features that lead to non‐linear mass balance gradients, ob‐ structed ice flow, or obstacles which constrain the glacier’s potential to react to climate changes. The following exclusion criteria were applied (see Loibl et al., 2014, for details):

(1) ‘debris cover’ on glacier tongues, (2) ‘reconstitution’ of glaciers through intermittent ice falls, (3) ‘valley floor limitation’ where glaciers reach flat valley bottoms, and (4) ‘detachment of tributaries’ where parts of a glacier detached during post‐LIA retreat.

At 688 of the 1964 glaciers, remote sensing and DEM‐based measurements were possible and none of the exclusion criteria was met; these were subsequently used for ELA calculation. This included both phases, 1999 and the LIA. The ELA calculation of the LIA maximum glacier advance used the same ridge‐altitude value but the toe‐altitude measured using the geomorphological evidence (see chapter 5.3.1). 65

The resulting point‐data was interpolated using the IDW algorithm (Watson and Philip, 1985). How‐ ever, in the subtropical latitudes of the study area pronounced ELA differences between northern and southern aspects exist (Evans and Cox, 2005; Evans, 2006). The ELA calculation was thus per‐ formed twice; in the first run northwestern, northern, and northeastern aspects were left out, in the second southwestern, southern, and southeastern. Subsequently, an arithmetic mean of the two resulting raster datasets was calculated in order to obtain a result that smoothes out the most ex‐ treme effects of different insolation intensities but retains local variability. By this procedure the E and W aspect measurements, which were incorporated in both raster datasets, were relatively in‐ creased in weight. The ΔELA data was obtained by subtracting the 1999 from the individual LIA val‐ ues. For the interpolation of this dataset, the ‘universal kriging’ algorithm using a full second‐order polynomial drift function was applied (Zimmermann et al., 2006). In comparison to IDW, less local variability is retained but general trends in the data are visible more clearly (Chaplot et al., 2006).

5.3.4 Post‐processing Several post‐processing steps were necessary in order to obtain clean vector datasets. First, reclassi‐ fied versions of the ELA (50 m intervals) and ΔELA (20 m intervals) were generated from the raw ras‐ ter results of the previous procedures. Local noise was removed and the boundaries between the resulting zones were smoothed by a shrinking and expanding algorithm with descending sorting. Subsequently, the raster files were converted into polygons and smoothed further by applying a bezier interpolation. In both ELA maps, interpolation results in marginal areas without data points in a distance of more than ∼ 10 km were excluded manually. Furthermore, all line shapefiles, i.e., rivers and ridges, were slightly smoothed in ArcGIS. A hillshade raster file was generated from the DEM data.

5.4 Map Design The key cartographical challenge was to present the wealth of information obtained in the study in a well legible and intuitively interpretable way. The map design therefore focuses on the presentation of the most important findings. However, scales from catchment to regional are relevant. The ap‐ proach therefore also aims to retain as many details as possible. The layout is based on two main maps showing two different sets of information for approximately the same area; glacier types and distribution in the left frame and calculated ELAs in the right. The idea behind this approach is to enable the viewer to combine the different sets of information individually for different areas.

The left map shows the distribution of glaciers and glacier types as mapped from the 1999 Landsat ETM+ scene. In the background, a subtle greyish hillshade provides an impression of the overall to‐ pography. The major rivers support this and additionally highlight the course of the major valleys. In order to allow for a distinct and subtle, yet intuitively perceivable representation of the river net‐ work, blue color, thin lines, and arrows showing the flow direction were chosen. The symbol design of the glaciers in the foreground was particularly challenging because two requirements had to be met: (1) very small and very large glaciers must both be clearly visible and, (2) different glacier types must be clearly identifiable, no matter whether glaciers in immediate vicinity or far away from each other are considered. Consequently, outlines were omitted and a quite intensive color ramp was chosen for the fills. Here, a color‐temperature gradient was applied in which cool colors represent

‘higher’ glacier types. These types comprise most large glaciers. ‘Lower’ glacier types are displayed with warm colors. Consequently, most small glaciers are colored in purple and beige and therefore clearly visible. Furthermore, the glaciers used for ELA calculation were marked by black circle sym‐ bols. These glaciers were marked by small black points in order to show the distribution but not to clutter the overall layout too much. Unglaciated areas in the northeastern and southwestern parts of the map were used for an overview map and the legend, respectively. This in‐map positioning helped to keep the layout compact despite the wealth of information.

Spatial patterns of ELA‐ and ΔELA‐distributions are the information displayed in the right map. In order to support easy cross‐referencing, the river‐network map was adopted from the left map. Ad‐ ditionally, major ridges were added to clarify the effects of topography on the ELAs, specifically lee‐ ward shelter and orographic precipitation. ELAs were classified in 50 m intervals and symbolized by a color gradient from blue to red. These colors are supposed to underline the message of how favora‐ ble the conditions are for glaciers; specifically, cool bluish colors imply low ELAs and warm reddish colors high ELAs. Contour‐style labels were added in order to improve the legibility of absolute ELA values. Marginal areas without data points in a distance of more than ∼ 10 km were excluded manu‐ ally, because here the interpolated data does not represent empirical measurements. These areas in the northeastern and southwestern parts of the map were filled with a white background and used for inline display of the auxiliary ΔELA map. The ΔELA map shows the post‐LIA ELA change in 20 m intervals. All measured ΔELA values were negative. Therefore, a color ramp from white to red was chosen in order to visualize the spatial distribution of negative trends.

5.5 Interpretation The maps depict several aspects of the complex high mountain environment in the eastern Nyainqêntanglha Range. The spatial distribution of glaciers shows a high level of heterogeneity. The largest glaciated areas and the greatest number of ‘high’ glacier types (i.e., the different types of valley glaciers) are concentrated in the northwestern part of the study area. Further to the North and particularly to the Northeast, the number of glaciers mapped for 1999 is decreasing and glacier types are ‘lower’ in comparison. In general, glaciated areas are attached to approximately east‐west run‐ ning ridges in most settings in the study area.

Similar patterns can be observed in the ELA map where steep ELA‐gradients coincide with the posi‐ tions of the major ridges in many settings. These local ELA phenomena can be interpreted in relation to the moisture sources and atmospheric circulation patterns, i.e., the monsoonal system and its pathways through the study area (Benn and Lehmkuhl, 2000). A clear imprint of the valley system, which is represented by the river network in the map, can be observed in the spatial distribution of ELAs. This implies that the monsoonal air masses are channeled by the valley system. Furthermore, patterns of orographic precipitation and leeward shelter can be observed in many settings. These are particularly obvious in the large northwestern range but many other settings with ELA troughs in front of ridges and peaks in lee positions exist (Fig. 5.6). In most parts of the study area, these pat‐ terns indicate moisture advection from southern directions. This fits well to observations on the modern circulation in Asia and can be attributed to the Indian Summer Monsoon (e.g., Wang et al., 2010). However, in the eastern part of the Parlung Tsangpo Valley and in the Bodui Valley, luv‐lee

67 effects can also be observed on north‐south running ridges. This indicates another moisture source in the East, probably the East Asian Summer Monsoon. The meteorological data available support this assumption: Nyingchi in the southwestern part of the study area shows a monomodal precipitation regime (683 mm/year) whereas Bomi in the eastern part has two precipitation peaks and considera‐ bly higher total precipitation (895 mm/year, Loibl et al., 2014). The clear dependence of ELA troughs and peaks on topography further strengthens the hypothesis that the pathways of monsoonal air masses are prescribed by the valley system. Consequently, the majority of moisture is probably being transported in altitudes below the average ridge elevation in the study area, i.e., considerably below 6000 m asl.

Fig. 5.6: Interpretation of the spatial distribution patterns of ELAs. Peaks and troughs are closely linked to topography, i.e., valley system and major ridges. Consequently, ELA‐troughs and ‐peaks in close proximity to ridges are interpreted as a result of orographic precipitation and leeward shelter, respectively. Note that the overall distribution of ELAs roughly traces the valley system on different scales, regional to catchment.

The ΔELA map shows the spatial distribution of regional ELA rise since the LIA. The highest values are located in the south‐central part of the study area; the maximum of 160–180 m is close to the con‐ fluence of Parlung and Yarlung Tsangpo. Thus, the strongest change in ELAs is also observed in the areas where the main stream of monsoonal moisture encounters the major mountain ranges. Rela‐ tively low ΔELA values, i.e., less than 100 m, are in the northeastern part of the study area. Several

68 leeward shelter positions mentioned above also show relatively few ELA rise. Thus, the frequently observed negative correlation of increasingly continental climate and decreasing glacier sensitivity to climate change seems to be also valid in the eastern Nyainqêntanglha Range (e.g., Hoelzle et al., 2003; Bolch, 2007). The trench of < 100 ΔELA values that crosses the Yarlung Tsangpo valley in the southernmost part of the map is probably an artifact of interpolation since there are no data points in this area.

5.6 Conclusions The maps provide insights into the spatial distribution of glaciers and the associated spatial ELA‐ patterns for the eastern Nyainqêntanglha Range. Furthermore, the measurements of geomorpholog‐ ical evidence of the LIA maximum glacier advance allowed the analysis of spatio‐temporal phenome‐ na of post‐LIA glacier retreat and ELA change. A close link between the spatial patterns of ELAs and topo‐climatic parameters is evident. The dataset thus provides important clues to improve the un‐ derstanding of the coupling between glacier and climate changes in this region. Key findings com‐ prise:

‐ All glacier margins in the study area retreated since the LIA and all ΔELA values are negative ‐ ELA patterns are closely linked to topography, indicating (a) a strong influence of luv‐lee effects and (b) an important channeling influence of the valley system on monsoonal air masses ‐ ELA calculations yield good results on different scales, regional to catchment, indicating a good performance of the calculation and interpolation methods.

The layout consisting of two main map frames and two additional inline maps enables the viewer to investigate the complex setting. This is achieved by a step‐wise approach that aids intuitive immer‐ sion into the subject matter; i.e., by arranging elements following their complexity approximately in the direction of reading. The layouts of the different elements are all designed to supports an intui‐ tive perception of the data. Specifically, all colors and symbols were chosen to support the viewer in understanding the data.

5.7 Software All data processing, mapping, and map design was performed in ESRI ArcGIS 10.1. Raster files, i.e., the hillshading raster, were processed in Adobe Photoshop CS4. Adobe Illustrator CS4 was used for vector file optimizations and the final PDF‐file export.

6 Toward a late Holocene glacial chronology for the eastern Nyainqêntanglha Range, southeastern Tibet

David Loibla, Philipp Hochreutherb, Philipp Schultea, Daniela Hüllec, Haifeng Zhu d, Achim Bräuning b, Frank Lehmkuhla a RWTH Aachen University, Department of Geography, Templergraben 55, 52062 Aachen, Germany b FAU Erlangen‐Nürnberg, Department of Geography, Am Wetterkreuz 15, 91058 Erlangen, Germany c University of Cologne, Institute of Geography, Albertus‐Magnus‐Platz, 50923 Cologne, Germany d Chinese Academy of Sciences, Institute of Tibetan Plateau Research, Lincui Road 16, 100101 Beijing, China

Monsoonal‐affected temperate glaciers in the eastern Nyainqêntanglha Range, southeastern Tibet, are highly sensitive to climate change. Knowledge about their late Holocene dynamics is still, howev‐ er, widely lacking. The few studies on glacial chronology which are available for this region tend to mainly focus on dating results without sufficiently considering the geomorphological setting, often leading to misinterpretations in this complex high mountain environment. In this study, two selected glacier forelands are investigated using a multi‐proxy approach encompassing detailed geomorpho‐ logical mapping, dendrochronology, sedimentology, and optically stimulated luminescence as well as radiocarbon dating. The starting point was the creation of a consistent morphosequence which was validated by remote sensing of further glacier forelands from the wider region. Similarities and dif‐ ferences between the investigated settings were analyzed in detail to identify the relevant morpho‐ logical and topoclimatic forcing mechanisms. We found evidence of climatic events affecting the whole region during the Little Ice Age, resulting in similar configurations and numbers of moraines. The geomorphological settings of the glacier forelands are, however, remarkably different, making investigations of the landform and sediment configuration an indispensable condition for their inter‐ pretation. Subsequently, constraints from different methods of relative and numerical dating were evaluated critically and included into a conceptual chronosequence if applicable. Our results suggest that the late Holocene maximum glacier advance comprised several successive advances from mid‐ 17th to mid‐18th century. None of our observations supports an earlier Neoglacial advance reaching further than the LIA maximum. After the LIA maximum, continued retreat that was only interrupted by short phases of stability followed, as evidenced by 2–3 recessional moraines in the investigated settings.

6.1 Introduction The southeastern part of the Tibetan plateau is strongly exposed to the moisture‐bearing monsoonal air masses and, as a consequence, is one of the most heavily glaciated parts of High Asia (Shi et al., 2006). The mountain range hosting most peaks with heights over 6000 m asl, the Nyainqêntanglha Range, exhibits ∼ 8000 km² of glaciated area. The glaciers belong to the ‘monsoonal temperate’ type which is often reported to be highly sensitive to climate change (e.g., Zhou et al., 1991). Recent stud‐

70 ies have shown that glacier mass loss in this area distinctly exceeds the average in High Asia (Gardelle et al., 2013; Neckel et al., 2014). A better understanding of these changes is of great importance, i.a. because of the glaciers’ relevance for Southeast Asia’s water supply (Immerzeel et al., 2010) and natural hazards such as landslides (e.g., Korup et al., 2010), glacier lake outburst floods (e.g., Benn et al., 2012), debris flows, flash floods, etc. (e.g., Milly et al., 2002; Fort, 2014).

However, there is still a considerable lack of information on late Holocene glacier fluctuations and the driving factors. It is recognized that glaciers in Tibet react heterogeneously to climate change (e.g., Scherler et al., 2011a; Yao et al., 2012; Owen and Dortch, 2014; Murari et al., 2014). In the eastern Nyainqêntanglha Range, a strong influence of topography and local topoclimatic effects on glacier changes was found (Loibl et al., 2014; Loibl and Lehmkuhl, 2014). Previous studies on the late Holocene glacier chronology in this region presented variable results that are contradictory in some parts (e.g., Yang, B., et al., 2008; Yi et al., 2008; Xu and Yi, 2014). Many of the glacier advances postu‐ lated in these studies are dated to 3–1 ka, typically referred to as ‘Neoglacial’ advances in the litera‐ ture (cf. Wang and Fan, 1987). A key problem is that many questionable radiocarbon and lichen da‐ ting results from pioneer studies (e.g., Li et al. 1986; Iwata and Jiao 1993) are still being used even though they do not comply with modern standards (e.g., Aitken, 1998; Hajdas, 2008; Preusser et al., 2008; Coe, 2010). In most cases, no information on methods, laboratories, sampling sites, sampled material, etc., is provided. This is particularly problematic because valid interpretation of moraine dating results is strongly dependent on such information (cf. Lehmkuhl, 1997).

Dendrochronology is an alternative method that has been widely used to obtain information on the timing of late Holocene glacier fluctuations (e.g., Wiles et al., 1999; Luckman, 2000; Larocque and Smith, 2005). Most of these studies focused on the so‐called LIA, referring to cold phases within the last millennium (Grove, 2004). In the eastern Nyainqêntanglha Range, such dendroglaciological ap‐ proaches are particularly useful because many glacier tongues are reaching down below the timber‐ line; consequently, a wealth of dendrochronological case studies from different glaciers in the region has been published during the last decade (e.g., Bräuning, 2006; Xu et al., 2012; Zhu et al., 2013). In comparison to the aforementioned 14C dating results, most tree ring minimum ages for beginning glacier retreat are remarkably young, i.e., < 300 years. The applied sampling strategies are heteroge‐ neous and interpretations do not always consider the implications of different tree stand positions, e.g., inner moraine slope, moraine crest, outer moraine slope. This is particularly problematic for studies missing detailed information on the sampled sites (e.g., Li et al. 1986), causing similar chal‐ lenges regarding the interpretation as with the 14C dating results. Consequently, the results available so far hardly allow a regional integration. In most cases, the deduced timing of LIA glacier fluctua‐ tions differs from reconstructions for other regions, such as the European Alps (Holzhauser et al., 2005; 1300–1380 CE, 1600–1670 CE, 1800–1860 CE) or the Canadian Rockies (Luckman, 2000; 1190– 1250 CE, 1280–1340 CE, 1690–1700 CE, 1800–1810 CE). Specifically, the LIA maximum glacier ad‐ vance is often reported to be relatively early in eastern Nyainqêntanglha Range (e.g., ∼ 1780 CE in Bräuning et al., 2006) and evidence of LIA glacier advances prior to 1750 CE is widely lacking.

As a consequence of this challenging situation regarding numerical dating results, the starting point for this study was the development of a reliable morphosequence, based on detailed geomorpholog‐ ical mapping (cf. Kirkbride and Winkler, 2012; Barr and Lovell, 2014). This was conducted by a combi‐

71 nation of fieldwork and remote sensing at two large debris‐covered glaciers on the northern and southern slopes of Mt. Pulongu (∼ 6300 m asl). Additionally, the large foreland of Zepu glacier, ∼ 60 km ENE from Mt. Pulongu, was mapped by remote sensing techniques to validate the mor‐ phosequence in another setting. To allow conclusions regarding the chronology of events and the processes involved, fieldwork followed a multi‐proxy approach including geomorphological, sedimen‐ tological and dendrochronological methods. Geomorphological and sedimentological analyses were applied to provide insights into the processes at work and a basis for setting up a relative chronose‐ quence using weathering indices (cf. Bäumler, 2004). Different archives were sampled to test the feasibilities of different dating techniques, including dendrochronological dating, radiocarbon dating and OSL dating. Finally, a synoptic analysis of the results from this study and others was carried out to disentangle the complex interactions between glaciers, topography, paleoclimate and vegetation, and thus proceed toward a comprehensive chronosequence of late Holocene glacier advances for the eastern Nyainqêntanglha Range.

6.2 Regional setting The eastern Nyainqêntanglha Range is a high mountain range in southeastern Tibet (Fig. 6.1), contin‐ uing the Himalaya orogen toward the east and encompassing its eastern syntaxis at mount Namcha Barwa and Gyala Peri (Burg et al. 1998). The region is subject to active tectonics (Finnegan et al., 2008; Kornfeld et al., 2014). Amongst the most prominent tectonic features is the Jiali‐Parlung Fault (Seward and Burg, 2008) which prescribes the course of the Yigong and Parlung Tsangpo Valleys, running almost straight in ENE‐WSW direction for more than 400 km. Mount Pulongu (∼ 6300 m asl) is one of the highest summits in the range south of the Yigong Tsangpo Valley.

Fig. 6.1: Regional setting of the study sites at Mt. Pulongu and Mt. Zepu Kangri in southeastern Tibet.

The Yarlung Tsangpo River (Brahmaputra in India) separates the Namcha Barwa and Gyala Peri mas‐ sifs and subsequently turns southward in what is called the ‘Great Bend’. Beyond the Great Bend, the Yarlung Tsangpo River incised deeply and formed a large, south running valley. During summer, this valley is the most important corridor for the transport of monsoonal air masses toward the Tibetan Plateau; particularly, the Indian Summer Monsoon conveys moisture from the Bay of Bengal in the south into the study area (Molnar et al., 2010). As a consequence of high mountain topography and high precipitation, the eastern Nyainqêntanglha Range is extensively glaciated (Liu et al., 2008). The glaciers are of the monsoonal temperate type and thus characterized by high ice‐flow velocities (e.g., Zheng et al., 1999) and a coincidence of accumulation and ablation periods in summer (Kaser and Osamaston, 2002). A regional remote sensing study revealed highly heterogeneous patterns of post‐ LIA glacier retreat which are distinctively coupled with local topographies (Loibl et al., 2014; Loibl and Lehmkuhl, 2014). Specifically, the predominately E–W‐running mountain ridges act as barriers for the approximately S–N‐directional, valley‐channeled monsoonal moisture transport. As a result, oro‐ graphic precipitation and lee‐ward shielding have strong effects on local topoclimates (Maussion et al., 2014) and in turn on the glacier dynamics. In combination with the strong influence of insolation on the mass balances of subtropical glaciers (Evans and Cox, 2005; Evans, 2006), highly complex glac‐ ier–relief–climate interactions are evident in which the slope aspect is particularly relevant. Many glaciers in the region, particularly larger valley glaciers, have debris‐covered tongues. Large moraines

73 encompassing the glacier forelands are abundant in the eastern Nyainqêntanglha Range. Several case studies have focused on mapping and dating these landforms at different glaciers, strongly sup‐ porting LIA age of formation (e.g., Bräuning and Lehmkuhl, 1996; Bräuning, 2006; Zhu et al., 2013; Xu and Yi, 2014).

Mount Pulongu summit (∼ 6300 m asl) is > 1000 m higher than the adjacent ridges. From this topo‐ graphically exposed peak, glaciers are descending toward all different aspects (Fig. 6.2A, B). All prom‐ inent glaciers at Mt. Pulongu are valley glaciers with debris‐covered tongues. Inferred equilibrium line altitudes (ELA) obtained from single‐year (1999, cf. Fig. 6.2) late summer transient snowline by remote sensing are at ∼ 4800 m asl at the northern glacier (‘Xinpu’) and at ∼ 5000 m asl at the southern glacier (‘Gongpu’). Zepu Glacier is located ∼ 60 km ENE from Mt. Pulongu in an east–west running valley (Fig. 6.2C, D). The highest peak, Zepu Kangri (6364 m asl), is located close to the valley head in the east. Most tributaries are northward facing glaciers, with ELAs ranging from ∼ 4600 to 4800 m asl. Tributaries with southern aspects are smaller and their ELAs are considerably higher (∼ 5050–5100 m asl). Baitong Glacier is located ∼ 4 km east of Zepu’s present glacier snout. In the following text, the names ‘Xinpu’, ‘Gongpu’, ‘Zepu’ and ‘Baitong’ always refer to the glaciers; other parts of the system will be addressed explicitly (e.g., ‘Xinpu Valley’).

Fig. 6.2: Sites of case studies; (A) Mount Pulongu (6300 m asl) in Landsat ETM+ scene (23 September 1999, channels 5‐4‐3) and, (B) mapped glacier extents and moraine crests for Xinpu and Gongpu; (C) Mount Zepu Kangri (6364 m asl) in Landsat ETM+ scene and, (D) mapped glacier extents and moraine crests for Zepu and Baitong. Please note the different scales of maps A and B in comparison to C and D.

6.3 Methods Fieldwork was conducted on the glaciers descending from Mount Pulongu toward the north (Xinpu) and the south (Gongpu). The logistics posed a great challenge and allowed only for a maximum of 74 four days of work at the glaciers, owing to the remoteness of the area and the poor infrastructure. Consequently, close attention had to be paid on sampling strategies. The set of applicable methods was restricted to time‐efficient approaches that do not require heavy equipment. Specifically, this included geomorphological mapping of the proglacial settings, and sampling of various archives.

6.3.1 Geomorphological mapping A combination of field mapping and remote sensing basing on a Landsat ETM+ scene (30 m pansharpened to 15 m resolution) and Google Earth imagery (< 1 m resolution) was applied to create detailed geomorphological maps of the investigated glacier forelands (see also Loibl and Lehmkuhl, 2013, for details on the approach). This allowed for obtaining positions and identification criteria of different moraine ridges. Specifically, the outer walls of the maximum glacier advance moraines were usually covered by trees, whereas the inner part of the glacier forelands were mostly free of phaner‐ ophytes, which is in accordance with the findings of Bräuning (2006). Consequently, the moraine ridges appear as distinct boundaries in the satellite images. In contrast to this clear delineation, smaller moraine ridges within the glacier forelands were much harder to distinguish and required a combination of all available material. GPS waypoints marking moraine ridges and panoramic photos proved to be particularly important for matching these less prominent features to the satellite im‐ agery. On basis of the case studies, GIS and remote sensing techniques were applied to evaluate whether similar settings exist in the wider region or not. For one prominent example, Zepu Glacier in the Bodui Valley, a geomorphological overview map of less detail was created. This mapping was based solely on Landsat ETM+ imagery and the SRTM DEM, supported by several photographs taken during a dendrochronological campaign in 2013.

6.3.2 Sedimentological analysis Sediment sampling focused on moraine crests, terraces of proglacial lakes and meltwater channels. At moraine crests, the material finer than gravel was sampled; in most cases, this material formed a matrix between larger boulders. As a consequence of the analysis of the setting at Xinpu, a higher priority was assigned to sampling glaciofluvial and glaciolimnic environments in the second campaign to Gongpu. Thus, the dataset comprises three glaciolimnic sediment sections from Gongpu but only two such samples from Xinpu. Two additional ‘down‐valley’ sections were dug during the descent from Gongpu to better understand the dynamics of weathering and soil formation in the study area.

For particle size analysis, all samples were dried at 35 °C, homogenized and passed through a 2 mm sieve. Organic matter was removed by repeatedly treating the samples with 0.70 ml 30% H2O2 at 70 °C for several hours (Allen and Thornley, 2004). To keep particles dispersed, the samples were treated with 1.25 ml Na4P207 for 12 hours on an overhead shaker (Pye and Blott, 2004; DIN ISO 11277, 2002). Particle size was measured with a Laser Diffraction Particle Size Analyzer (Beckman Coulter LS 13 320). Grain size distributions (0.04–2000 μm) were determined applying the Mie theory (Fluid RI: 1.33; Sample RI: 1.55; Imaginary RI: 0,1; Özer et al., 2010; ISO 13320‐1, 1999; cf. Lehmkuhl et al., 2014).

Element concentrations were measured on the fine‐grained fractions. These were extracted by siev‐ ing out material <63 µm and drying it at 105 °C for 12 hours. Eight g of the sieved material was mixed with 2 g Fluxana Cereox, homogenized, and pressed to a pellet with a pressure of 20 tons for 120 s. All samples were measured twice with a Spectro Xepos X‐ray fluorescence (XRF) device. Subsequent‐ 75 ly, mean values were calculated from the two measurements following the instructions of Spectro (2007). A small portion of each sample was treated with 10% HCl to evaluate the state of Ca (Loeppert and Suarez, 1996). Because none of the samples reacted to this treatment, Ca concentra‐ tion results of the XRF measurements are interpreted as parts of silicates and not as CaCO3. Thus, no further correction of the XRF results was necessary (Buggle et al., 2008; Smykatz‐Kloss, 2010). To evaluate the XRF results, a set of widely used geochemical indices was applied (Table 6.1, see suppl. Fig. B.10 for ratios of these and further elements).

Table 6.1: Geochemical indices used in this study and their interpretation.

Index Interpretation References

TiO2–Al2O3 Provenance (higher values ≈ more maﬁc parent materi‐ Sheldon and Tabor (2009), al). Remains constant during soil formation (both ele‐ Smykatz‐Kloss and Roy (2010) ments are relatively immobile)

(Na2O+K2O)–SiO2 Relative contents of alkaline feldspar and Quartz, parent Gauthier and Burke (2011), rock composition of weakly weathered sediments Nesbitt and Young (1984)

Ti–Zr Degree of physical weathering Zech et al. (2008), Lehmkuhl et al. (2014)

Rb–Sr Degree of chemical weathering and pedogenesis, posi‐ Gallet et al. (1996); Chen et al. (1999) tive water balance

Al2O3–CaO Degree of decalcification and chemical weathering Mason and Moore (1982), Buggle et al. (2011)

Na2O–K2O Degree of pedogenesis and availability of organic matter Nesbitt et al. (1980), Smykatz‐Kloss and Roy (2010)

6.3.3 OSL analysis Fifteen cm long steel tubes with a diameter of 56 mm were used for OSL sampling. Sampling sites were checked carefully for roots and cracks in the sediment body to ensure that the recovered mate‐ rial was not bleached post‐depositionally. For the same reason, sampling positions were chosen to be at least 30 cm beyond ground surface where possible. Only material from the central ∼ 5 cm of the tubes was used for the OSL measurement. Additionally, mixed samples of adjacent material were recovered for each OSL sample for determination of background radiation. The OSL samples were prepared following standard procedures (e.g., Preusser et al., 2008, Hülle et al., 2010).

Quartz is commonly used in luminescence dating studies on Holocene deposits. Luminescence dating studies in Central Asia have, however, shown that samples may exhibit feldspar contamination of the luminescence signal (Hülle et al., 2010; Lehmkuhl et al., 2011). Stimulating each coarse‐grained quartz sample with infrared light (cf. Duller, 2003) testified for a feldspar contamination of the sam‐ ples, inhibiting accurate paleodose determinations. Consequently, potassium‐rich feldspars were used for dating in preference to quartz. These were measured with infrared stimulated luminescence (IR LEDs, detection wavelength 380–440nm, interference filter D410, L.O.T. Oriel), carried out on automated Risø TL/OSL readers (TL‐DA‐20), equipped with 90Sr/90Y β‐sources for irradiation and EMI 9235 photomultiplier tubes for luminescence detection (Bøtter‐Jensen et al., 2003). 76

Preheat tests combined with dose recovery tests (the samples were illuminated for 24 h in a Hönle Sol2 solar simulator followed by beta irradiation comparable to the expected paleodose) revealed that the standard Single Aliquot Regenerative (SAR) dose protocol (Murray and Wintle, 2000; Wallinga et al., 2000) is not suitable for the examined samples. Increasing dose underestimations of 16–70% were observed for all preheat temperatures within the range of 210/210 °C to 290/290 °C (ph regeneration dose/ph test dose). This result was observed for all three investigated samples from different sections and thus appears to be systematic. Therefore, the postIR‐IR180°C‐protocol by Reimann et al. (2011) was tested and modified, using a preheat at 200 °C, an IR bleach at 50 °C and IR‐stimulation at an elevated temperature of 180 °C (followed by an IR‐hotbleach at 280 °C only after the test‐dose cycle). For this protocol, the applied dose could be recovered within ±3% and the gen‐ eral quality criteria (recycling ratio limit of ±10%, recuperation <5%, test dose error <10%) further indicate applicability of the protocol. Furthermore, this post‐IR IRSL signal using elevated stimulation temperatures has significant potential to derive a dating result that is far more independent from fading‐correction than the conventional IRSL50°C‐signal.

Dose rates are based on laboratory high resolution ‐spectrometry results using 800 g of bulk materi‐ al. The dose rate was calculated taking variations in the water content of 5–15 weight‐% into ac‐ count. An internal potassium content of 12.5  0.5% (Huntley and Baril, 1997) and an α‐efficiency factor of 0.07  0.02 (Preusser et al., 2005) was assumed. The contribution of cosmic rays to the dose rate was calculated following Prescott and Hutton (1994). All dose rate results are given in Table 6.2.

6.3.4 14C analysis A buried tree was discovered in the steep upper part of the largest lateral moraine at Xinpu (see Fig. 6.4 for location). A slice of the tree was retrieved for 14C dating which was conducted in the Erlangen AMS Laboratory. Wood samples were purified by acid–alkali–acid treatment, then combusted in an elemental analyzer, reduced to graphite, and measured according to the standard laboratory proto‐ cols documented in Morgenroth et al. (2000). Ages were calibrated basing on the data set of Hua et al. (2013) using the in‐house program Cal09/IntCal09 (Reimer et al., 2011). For wiggle‐matching, Ox‐ Cal v 3.10 software (Bronk Ramsey, 2009) was employed, using IntCal09 data.

6.3.5 Tree sample analysis Tree sampling strategy at Xinpu included sampling of both lateral moraines as well as one of the frontal moraines for dating. Another site that was most probably unaffected by direct ice contact during LIA was chosen for comparison and finding missing rings in the other chronologies. Since Picea balfouriana was the dominant species with the oldest trees, all site chronologies were composed of this species; for species comparison, a few samples of Larix griffithii were taken. The southern glacier at Mt. Pulongu, Gongpu, was sampled differently since all sampled specimen on the lateral moraines of species P. balfouriana were rotten from the core. Therefore, sampling concentrated on young tree stands on the frontal moraine. Samples of L. griffithii and P. balfouriana were taken in chronology strength (25 trees/50 cores per species).

Tree ring widths of all cores were measured following standard dendrochronological techniques (Stokes and Smiley, 1968). After air‐drying, the cores were mounted on sample holders. Wood sur‐ faces were sanded using sandpaper from grain 80 to 1200 to maximize contrast and visibility. Ring widths were measured with an LINTAB VI measuring system with a resolution of 0.01 mm. All cores 77 were cross‐dated visually and using statistical test (GLK and t‐test) with the software package TSAP‐ Win (Rinn, 2003). In cases where the corer had missed the pith, the missing rings (‘pith offset’) was estimated using curvature and width of the innermost 5 rings. Additionally, 10 years were added as a correction for the time needed to grow to breast height (Bräuning 2006). Resultantly, five chronolo‐ gies were established at site Xinpu: one for each lateral moraine, one for the frontal moraine, one below the LIA moraines (longest) and a weak Larix chronology. At Gongpu site, two short frontal mo‐ raine chronologies were established (P. balfouriana and L. griffithii).

6.4 Results

6.4.1 Proglacial geomorphology In order to set up a morphosequence and to correctly interpret the dating results, the geomorpho‐ logical configurations of two glaciers were investigated in detail. The glacier at the northern slopes of Mt. Pulongu (Xinpu) is a debris‐covered valley glacier. Its late Holocene glacier foreland reaches down to ∼ 3200 m asl and covers an area of ∼ 1.13 km² (Figs. 3A, 3B, 4A). The most prominent fea‐ tures are the two lobes of the outer latero‐frontal moraine. The western of these two lobes is con‐ siderably smaller and hydrologically decoupled from the active glacier (cf. Fig. 6.4A). Instead of melt‐ water from the active glacier, a large debris‐covered dead ice body is draining into the western lobe’s proglacial lake. Periodic runoff was witnessed at the channel draining this lake through the outer moraine during fieldwork in 2011. At the vertex of both moraines lobes, the tip marks a peak with a steep slope toward the eastern side (cf. Fig. 6.4A). Between the dead ice body and this peak, fluvio‐ glacial sediments indicate that meltwater discharge ran through this channel in former times. Today, this paleochannel is ∼ 20 m above the active drainage way (Figs. 6.3B, 6.4A). The outer slopes of the western moraine lobe are covered by contiguous forests (N4 and N5 in Figs. 6.3A, 6.3B, 6.4A). At the moraine peak, moraines are > 50 m high and sufficiently steep to prevent the tree succession by rockfall. Forests are patchy throughout the inner moraine walls of the eastern lobe (Fig. 6.3B). Three smaller moraine ridges in ice‐proximal positions were mapped (N1, N2 and N3 in Figs. 6.3A, 6.4A, referred to as ‘inner moraines’ in the following). The glacier’s snout is a large backwasting ice cliff with grounded frontal margin, discharging into a small lake which is surrounded by moraine ridges and debris‐covered dead ice (between N1 and N2 in Fig. 6.3A). The runoff drains beyond the debris of the second moraine and discharges again at another backwasting ice cliff. The glacier tongue is extensively debris‐covered and has a hummocky surface with numerous ice cliffs and smaller melting ponds. Further less prominent lateral moraines are evident besides the glacier tongue.

Fig. 6.3: Panorama photos of the studied glacier forelands; see Fig. 6.4 for exact locations and viewsheds. Moraines were numbered according to their positions relative to the modern snout of each glacier; therefore, corresponding numbers at Gongpu and Xinpu do not necessarily mean equal ages. (A) View from the true left moraine at Xinpu, showing the central part of the glacier foreland, the innermost recessional moraines, and the debris‐covered glacier; (B) downvalley view at Xinpu, showing the two large moraine lobes and several smaller recessional stages; (C) view from the true left moraine at Gongpu, showing the central part of the glacier foreland, the inner recessional mo‐ raines, and the debris‐covered glacier; (D) terminal moraine setting at Gongpu, including push mo‐ raines and a lobe‐shaped depression with lake sediment fill. Photos by D. Loibl 2011, 2012.

A comparison of the glacier forelands at Xinpu to the situation at Gongpu (Figs. 6.3C, 6.3D, 6.4B) re‐ veals several remarkable similarities and differences. The similarities comprise: (i) overall setting, i.e., retreating debris‐covered valley glacier, (ii) three separate proglacial lakes, (iii) a moraine peak at the vertex of several outer moraines, (iv) two major outer and at least three smaller inner moraine stag‐ es. Noticeable differences concern: (i) elevation, ∼ 500 m higher at Gongpu, (ii) size of glacier fore‐ land (∼ 0.37 km²), ∼ 1/3 of Xinpu, (iii) no lateral moraine breach and different relative positions of moraines. A tight set of push‐moraines around the ∼ 25 m high moraine peak shows that strong ad‐ vances occurred at Gongpu as well. Furthermore, a lobe‐shaped depression of ∼ 130 m diameter is evident in the frontal part of the terminal moraine. The center of this structure is filled with horizon‐ tally bedded fine sand sediments, very similar to the lake facies in the active lake environments at Xinpu and Gongpu. The large proglacial lake at Gongpu (left part of Fig. 6.3C) was of particular inter‐ est, because it was the only such setting still connected to meltwater discharge of the active glacier. Field observations revealed water flow in many shallow channels near the lake shore. Satellite im‐ agery shows that most of the lake is only periodically filled with water whereas large areas are flood‐ plains. Many of the paleolake settings are in similar morphological positions; consequently, a combi‐ nation of different process regimes will probably have shaped all of those settings.

Fig. 6.4: Geomorphological maps of the glacier forelands of the glaciers (A) Xinpu and (B) Gongpu (see Fig. 6.2 for locations). All three maps are at the same scale. Note the different elevations and sizes as well as the similar arrangement of features and equal number of moraine stages. The mapping of (A) and (B) was based on field data (GPS waypoints and tracks, sketches, photos) and satellite imagery (Landsat ETM+ from 23 September 1999, Google Earth). A corresponding map of Zepu glacier is pro‐ vided as supplement (suppl. Fig. B.14).

Two additional glaciers were studied by remote sensing techniques for regional integration. The glac‐ ier foreland at Zepu glacier is considerably larger (∼ 3.7 km²) than those of the glaciers at Mt. Pulongu (cf. suppl. Fig. B.14). Even though the mapping results obtained solely from remote sensing data are less detailed, two major outer moraine lobes can be clearly identified. Between those lobes, there seems to be one additional, less prominent lobe which has not been observed at the other glaciers. Conversely, only one inner moraine could be delineated. This is, however, probably owing to the resolution of the Landsat ETM+ imagery (∼ 15 m) which is insufficient to resolve smaller objects.

The same applies for Baitong glacier, but its glacier foreland comprises only ∼ 0.5 km². Detailed ge‐ omorphological mapping from Landsat data was thus not applicable; however, it is clearly visible that its tongue is narrower and more elongated than those of the other three glaciers (Fig. 6.2, Fig. 6.4). Indications of multiple moraine lobes are missing.

6.4.2 Sedimentology From field observations, Xinpu and Gongpu have a similar lithologic debris composition consisting of felsic‐type plutonic rocks, i.e., diorite, as the main component. Subordinate types are various igneous and metamorphic rocks. At Xinpu, the crests of the LIA laterofrontal moraines are typically covered by dense vegetation and relatively young forests (Fig. 6.5A). However, the sediment sections re‐ vealed that humic topsoils are very thin (∼ 1–4 cm). Fresh, unweathered glacial sediments, mostly in form of clast‐supported massive diamictons, were found beneath the topsoils of these shallow lepto‐ sols (Fig. 6.5B). A different structure was observed at the edge of the dead‐ice body (see Fig. 6.4A for section locations). This loosely stratified section was free of larger boulders and consisted only of fine gravels to fine sands (Fig. 6.5C). The ground surface was stabilized by a thin cover of pioneering plants without any topsoil. Grain size measurements yielded similar grain size distributions for the fine fractions of all moraine sections except for those from the dead‐ice body (Fig. 6.6A). Typically, the moraine samples show relatively heterogeneous multimodal distributions with maxima in the middle and coarse sands. In contrast, most sediments from the dead ice body are well sorted with distinct monomodal maxima in different fractions of the sand spectrum (Figs. 6.5C, 6.6A). This sec‐ tion’s well sorted layers of fine sand closely resemble sediments from the glaciofluvial terrace in the vicinity of Xinpu’s large proglacial lake (Figs. 6.5E, 6.6B). However, the sediments of the terrace sec‐ tion are often characterized by a distinct cross‐bedding (Fig. 6.5E). As a consequence of the detach‐ ment from the active glacier’s meltwater discharge, the large modern proglacial lake at Xinpu is a stillwater system. Here, sediments were horizontally bedded and grain‐size analysis revealed a mon‐ omodal distribution with a maximum in the coarse silt and a shoulder toward the fine sand (Figs. 6.5D, 6.6B).

Fig. 6.5: Different sample sites at Xinpu (left, A–E) and Gongpu (right, F–K); see Fig. 6.4 for exact loca‐ tions. (A) section on a moraine ridge of N4, covered by relatively young forests; (B) detail of the sec‐ tion (N‐P03); note the thin organic topsoil on fresh glacial sediments; (C) > 1 m long section showing a heterogeneous sequence of well sorted fine sands and badly sorted layers consisting mainly of coarse sand and fine gravel (N‐P02, yellow cap marks OSL3); note that there is no topsoil and only a thin cover of pioneer plants; (D) glaciolimnic site (OSL1); (E) glaciofluvial terrace (OSL2); (F) compact fine sands with slight horizontal bedding (S‐P02, OSL7); (G) lake terraces consisting of a sequence of fine– medium sands (detail shows the upper terrace S‐P15 including OSL9, in the foreground of the over‐ view is the pit of S‐P06); (H) material at moraine S2, with no visible sign of weathering and only few loose pioneer plants (S‐P07); similar morphological position as (B); (I) section on top of the large lat‐ eral moraine, very thin organic topsoil with pioneering plants; (J) section on a small fan ∼ 1.5 km down‐valley of the LIA frontal moraines (S‐P16); note the distinct soil horizons; (K) section ∼ 7.5 km down‐valley of the LIA frontal moraines (S‐P17); soils looked much more advanced than at (J). Photos by D. Loibl 2011, 2012.

The glaciofluvial and glaciolimnic sections at Gongpu all show evidence of varying depositional pro‐ cess regimes, ranging from moderately fast‐flowing to stillwater (see also chapter 6.4.1). All such sections consisted of relatively homogeneous horizontal layers with grain size modes between coarse silts and coarse sands (Figs. 6.5F, 6.5G, 6.6D). In many cases, alternating light‐grayish and reddish‐ brown layers indicated changing hygric regimes promoting redox‐processes, e.g., by waterlogging. Moraine sediment sections at Gongpu had a similar composition to those at Xinpu, i.e., clast‐ supported massive diamictons of fresh‐looking rocky material (Figs. 6.5H, 6.5I). However, the topsoils of comparable settings were always much less developed and plant succession was in a more initial state (e.g., Fig. 6.5B vs. 6.5I). In contrast, the sections dug 1.5 and 7.5 km down the valley revealed

83 hydromorphic soils developed in silty–sandy sediments with distinctly weathered soil horizons (Figs. 6‐5J, 6‐5K).

Fig. 6.6: Grain size distributions for different sediment types at Xinpu (left) and Gongpu (right).

The analysis of the XRF results focused on indices that allow obtaining inferences regarding the li‐ thology at the different sites and weathering indices. Besides the insights into weathering dynamics, this aimed to test whether it is possible to set up a relative morphochronology for LIA moraines in this potential high‐weathering environment as it has been demonstrated for last glacial maximum (LGM) moraines in other regions (e.g., Bäumler et al., 2004). In general, the different archive types tend to form relatively homogenous and well separated groups in the investigated geochemical rati‐ os (Fig. 6.7, suppl. Fig. B.10). In the plots of the two indices applied to test on lithological differences in the composition of the materials, TiO2–Al2O3 (Fig. 6.7A) and (Na2O+K2O)–SiO2 (Fig. 6.7B), samples from both glacier forelands are overlapping. This indicates that the parent material at Xinpu and 84

Gongpu is widely similar. In contrast, the lower down‐valley sediments are forming a clearly separat‐ ed group in both plots, suggesting that other sources of material added to the substrate there.

Fig. 6.7: Geochemical indices plotted for different archives: (A) TiO2–Al2O3; (B) (Na2O+K2O)–SiO2; (C) Ti–Zr; (D) Rb–Sr; (E) Al2O3–CaO (F) Na2O–K2O.

Weathering indices included Ti–Zr, Rb–Sr, Na2O–K2O and Al2O3–CaO. In the Ti–Zr plot, Gongpu and Xinpu are mainly separated by different Zr proportions (Fig. 6.7C). The different archive types, i.e., glaciofluvial/‐limnic and moraines, can be differentiated by their Ti content. The outlier with the highest Ti and Zr values is the topmost sample from S‐P06. The upper down‐valley samples are plot‐ ting close to the samples from Xinpu, whereas the lower down‐valley samples form a separate group of relatively high Zr and low Ti proportions. This shows a trend to stronger physical weathering in the glaciofluvial and –limnic settings in comparison to the moraines. Expectably, the down‐valley sedi‐ ments are weathered strongest. A clear differentiation between Xinpu and Gongpu samples is evi‐ dent in the Rb–Sr plot only (Fig. 6.7D). Interestingly, the lower down‐valley samples from the south‐ ern side of Mt. Pulongu are nearly consistent with the samples from Xinpu on the northern side in this plot. According to the Rb–Sr index, chemical weathering and pedogenesis is stronger at Xinpu glacier. The Al2O3–CaO plot shows a relatively clear grouping, implying an increasing decalcification from Gongpu moraines to Gongpu glaciofluvial, Xinpu moraines and the down‐valley sections (Fig.

6.7E). The grouping by setting and a similar sequence can also be observed in the Na2O–K2O plot (Fig. 6.7F). However, transitions are smoother and Xinpu moraines and Gongpu glaciofluvial and –limnic sediments are very close to each other. This indicates that the availability of organic matter and the grade of pedogenesis are increasing parallel to the grade of decalcification. In summary, similar li‐ thologies and clear patterns of weathering are observed. In most weathering indices, a sequence 85 from Gongpu moraines to Gongpu glaciofluvial and –limnic to Xinpu moraines to down‐valley sedi‐ ments can be observed. Conversely, the intra‐archive variability is very low; distinguishing different moraine stages by their grade of weathering is not possible.

6.4.3 Tree ring dating The minimum age of sediment surfaces can be determined by the living age of trees that have estab‐ lished shortly after the sedimentation of the respective morphological forms, i.e., moraines (Bräuning and Lehmkuhl, 1996; Bräuning 2006). At Xinpu glacier, the oldest living tree found on the true right lateral moraine dates back to 1677 CE. Adding an ecesis of 5 years and a height correction of 10 years from the age adjustment function yields a minimum moraine age of 1662 CE. At the true left mo‐ raine, the cores of the oldest tree missed the pith; thus, the pith offset was calculated using curva‐ ture and ring width of the innermost five rings, providing a pith date of 1761 CE. Including height correction and ecesis, the date the glacier left the moraine ridge is reconstructed to 1746 CE. Trees growing on the terminal moraines were distinctly younger with a maximum of 69 years (1942 CE). The inner rings of these samples were considerably smaller with signs of suppressed growth. Thus the oldest tree did not grow on the terminal moraine before the early 20th century, indicating a high degree of morphological activity with reduced pedogenesis. At Gongpu, the minimum age dating method could not be applied, since all trees sampled on the lateral moraine were rotten from the core. Their true age could therefore not be determined. Trees growing on the terminal moraine showed similar ages as at Xinpu, indicating a similar setting. To obtain climatic information that al‐ lows conclusions on glacial processes, trees below the terminal moraine were sampled. This led to no statistically relevant results, so minimum age dating at this site was omitted. The oldest tree found on top of the lateral LIA moraine at Baitong covers a time span of 468 years, dating back to 1530 CE.

6.4.4 OSL and 14C dating results Nine OSL samples were taken from different archives at Xinpu (OSL 1–OSL 6) and Gongpu (OSL 7–OSL 9; see Fig. 6.4 for locations, Table 6.2 and suppl. Table A.1 for data, and suppl. Figs. B.1–B.9 for pho‐ tos). Sample OSL 1 was retrieved from horizontally bedded glaciolimnic fine sands close to the shore‐ line of the proglacial lake. The well sorted, greyish material was probably transported through the channel to the west of the lake. The main source of the episodic stream is the meltwater outflow of the large dead ice block (Fig. 6.4A). Adjacent to this channel, a ∼ 0.8 m high terrace of crossbedded glaciofluvial fine sands of similar sediment properties was sampled (OSL 2). For sample OSL 2, the risk of post‐depositional bleaching is relatively high, because the sediment body was partially cracked by roots and the sample was taken close to the surface (15 cm) to retrieve an age indicating the termi‐ nation of deposition. Sample OSL 3 was taken from the top of the glacier‐ward facing side of the dead ice block. The setting is interpreted as ridge of a push moraine where supraglacial glaciofluvial outwash was deposited, possibly after being transported beyond supraglacial debris. The sampled material was a ∼ 15 cm thick layer of well sorted, crossbedded fine sands beyond ∼ 80 cm of badly sorted coarse sands. Sample OSL 4 was acquired from unstratified, relatively bad sorted fine sands at the off‐glacier facing slope of a scree. The material was presumably accumulated by a combination of melt‐out and outwash processes during the retreat of the debris‐covered glacier tongue. Samples OSL 5 and OSL 6 both originate from crossbedded fine to medium sands in glaciofluvial settings with‐ in the proglacial scree. At Xinpu, unstratified, well sorted fine sands accumulated in a depression within the outer moraine (OSL 7). The samples OSL 8 and OSL 9 were taken from two stacked terrac‐

86 es adjacent to the meltwater‐fed proglacial lake. Sample OSL 8 originates from a layer of crossbed‐ ded medium to fine sand at the lower terrace. At the upper terrace, OSL 9 was taken from horizontal‐ ly bedded fine sand.

All investigated feldspar samples showed high signal intensities using the post‐IR‐IR‐180°C‐protocol and all measured aliquots fulfilled the required quality criteria. The luminescence dating results are given in Table 6.2 (see suppl. Table A.1 for further details). Mean equivalent doses were calculated with the central age model of Galbraith et al. (1999) including an error level of 1σ. The relative standard deviation of the individual paleodose measurements are quite broad for some of the sam‐ ples (cf. Table 6.2: 1, 2, 4, 6, 8), which has to be expected when dating glaciofluvial and glaciolimnic sediments. Nevertheless, the dose distributions are more or less symmetric, with skewness values between 0 and 1.3 (typically 0.6), which suggests that the Minimum Age Model (after Galbraith et al. 1999) is not applicable. All age results of 47 to 233 ka are far from the expected age range of the Late

Holocene. Even if only the lowest De‐values of the dose distributions are tentatively used for age calculation, the lowest ages are >28 ka.

In contrast to the OSL dating results, the 14C dating result fits well to the expected age range. The age information obtained for the tree trunk retrieved from a lateral moraine of the true right lobe at Xinpu glacier is, however, highly ambiguous. Since radiocarbon ages have a particular high relevance for the existing literature on glacial chronologies in the study area, we decided to publish the whole dataset, not only the measured age of 60 ± 45 BP or a single calibrated age (Table 6.3, suppl. Fig. B.11).

Table 6.2: OSL dating results. All samples were measured using the post‐IR‐IR180 method, assuming 10±5% water content (see Table A.1 for further details).

Table 6.3: Calibrated 14C dating results of the tree trunk from the inner side of the outer moraine at Xinpu glacier (see Fig. 6.4A for location, Fig. B.11 for further details).

1 σ confidence 2 σ confidence Age [AD] Probability [%] Age [AD] Probability [%] 1695 ‐ 1725 19,2 1682 ‐ 1735 25,5 1813 ‐ 1836 14,3 1804 ‐ 1935 69,7 1844 ‐ 1850 3,0 1955 ‐ 1956 0,2 1869 ‐ 1871 1,0 1876 ‐ 1917 30,7 1955 ‐ 1955 0,1

6.5 Discussion

6.5.1 Dating methods One objective of this study was to test the reliability of different dating methods (14C, OSL) in particu‐ lar young proglacial environments. With 47 to 233 ka, the OSL age results are much higher than ex‐ pected (< 1ka by the initial tree succession as well as dendrochronological and sedimentological evi‐ dence). The acquired ages do thus not reflect the timing of final deposition and there is no indication what event they are referring to, even though the dominant occurrence of early‐to‐mid last Glacial ages may speculatively be interpreted as a hint toward the initial breakup of solid rock bodies and subsequent physical weathering. Luminescence dating might be complicated by the depositional processes. If the time of sunlight exposure during short distance transport was too short to reset the stored luminescence signal, the obtained ages for these sediments would not represent the timing of deposition. The fact that the distributions in this study are not extraordinarily broad and that they are symmetrical suggests that the grade of bleaching is equal throughout the samples. A possible explanation is the relatively short distance of sediment transport, possibly beyond thick debris cover or blocks. As stated by Fuchs and Owen (2008), glacial sediment is often produced by glacial abrasion and transported within or at the base of a glacier, or deposited as a sediment package – and there‐ fore may never be exposed to daylight before being deposited. Additionally, glaciofluvial sediments are often transported in turbid glacial meltwaters which attenuate daylight efficiently so that bleach‐ ing may be hampered, even when the sediments are being transported exclusively supraglacial (Lu‐ kas et al., 2007; Fuchs and Owen, 2008; Alexanderson and Murray, 2012; Gaar et al., 2014). The sam‐ ples investigated in this study comprise different bleaching potentials. Even though transport path‐ ways are generally short, particularly fine‐grained glaciofluvial terrace sediments indicate relatively slow subaerial transport and deposition which should be favorable for bleaching. Considering the short transport pathways and attenuation, luminescence dating of quartz would be favourable, as the quartz signal bleaches faster than the feldspar signal (e.g., Godfrey‐Smith et al., 1988). The quartz fraction was, however, not suitable for dating owing to feldspar contamination and low sensitivity

(see chapter 6.3.3). Furthermore, usage of the post‐IR‐IR180°C‐protocol might also result in age overes‐ timation, as this protocol uses a part of the feldspar signal which is on the one hand more stable and less prone to fading, but on the other hand might be harder to bleach than the conventional IRSL50°C‐ signal. Since the IRSL50°C‐protocol was not applicable because of the feldspar contamination, the post‐ 89

IR‐IR180°C‐protocol was the best solution. Using the post‐IR‐IR225°C‐ or even the post‐IR‐IR290°C protocol might even have worsened the results.

The 14C result obtained from the tree trunk shows several age ranges of equally high probability. This ambiguity of 14C dating results often occurs when dating material of the past centuries (until ∼ 1650 CE) because of a plateau in the radiocarbon calibration curve (Reimer et al., 2011). However, the moraine setting from which the tree trunk was recovered is typical for the region. It thus seems very likely that the radiocarbon dating results from the aforementioned pioneer studies (e.g., Li et al., 1986; Iwata and Jiao, 1993) were subject to similar challenges, underlining that a critical re‐ evaluation of these results is necessary and publication of full datasets is important for future stud‐ ies. In the case of this study, further insights into the plausibility of different ages are possible from analysis of the setting (chapter 6.5.3).

In comparison to other dating methods, one of the main advances of tree ring dating is replication. This provides yearly precision through eliminating possibly missing or false rings and increases chances of finding the oldest individual. Nonetheless, the ecesis time lag can vary widely, ranging from only a few years (e.g., Sigafoos and Hendricks, 1969) to decades (e.g., Villalba, 1990). It is strongly dependant on local morphological‐, geobotanical‐ and topo‐climatic effects, including dis‐ tance to seed source, length of vegetation period (as a function of altitude, shading a.o.), and sub‐ strate grain size (van der Burght et al., 2012; Bräuning, 2006, and references therein). Buried ice, if being part of a moraine, may constrain the establishment of seedlings. This possible problem can be circumvented by sampling a reasonable large fraction of the moraine arc since this phenomenon can be expected to be of limited spatial extent. Furthermore, at all considered monsoonal temperate glaciers, the debris cover was found to be a sufficient isolation for the germination of seedlings (see also suppl. Fig. B.13). Therefore, an incorporated dead ice block can be expected to be only an excep‐ tional and short temporary obstruction. In summary, large glaciers descending to low elevations are preferable to minimize inflating effects on ecesis and height‐growth delay. Furthermore, a sampling strategy that is based on a good understanding of the geomorphological settings is crucial for validly interpreting tree ring dating results as minimum ages for glacier advances.

During the past five years, cosmogenic nuclide techniques have become a promising approach for dating LIA moraines. A pioneering study by Schaefer et al. (2009) showed that the age range of 10Be dating can be expanded to ages as young as 130 years due to the increasing sensitivity of measure‐ ments for glaciers in the Southern Alps, New Zealand. Recently, successful applications of cosmogen‐ ic nuclide dating on LIA moraines have also been reported for other regions, such as the European Alps (Schimmelpfennig et al., 2014) and the Peruvian Andes (Stroup et al., 2014). However, the high‐ energy environments of glaciers in the eastern Nyainqêntanglha Range pose great challenges to the application of this method, particularly regarding prior or incomplete exposure (cf. Balco et al., 2011; Heyman et al., 2011). Furthermore, cosmogenic nuclide dating requires independent data for calibra‐ tion which is widely lacking in the study area (cf. Rother et al., 2014).

The results of the different dating methods clearly indicate that even carefully planned sampling schemes may provide ambiguous results which cannot be assigned without knowledge of the mor‐ phostratigraphic setting and the transport processes. Specifically, the challenges that exacerbate correct dating and interpretation of dating results in proglacial environments comprise: 90

‐ high energy processes, often subglacial or below thick debris, leading to rapid transport and dep‐ osition (insufficient bleaching) ‐ complex combinations of different processes, i.e., glacial, gravitational, glaciofluvial, variable in space and time ‐ destruction and reworking of archives ‐ plateaus in the radiocarbon calibration curves

As a consequence, careful consideration of the proglacial environments is indispensable:

‐ Knowledge on the morphosequence through detailed geomorphological mapping ‐ Knowledge on the local stratigraphy, i.e., glacial geomorphology, soils and sediments ‐ Knowledge on the interpretation of different positions for tree sampling, e.g., lateral moraine crest, frontal moraine crest, moraine slopes, etc.

In the following, we will demonstrate this approach to disentangle the challenging dating result.

6.5.2 Comparison of different glaciers The establishment of a consistent morphostratigraphy is a necessary precondition for further inter‐ pretations. A comparative analysis of the different glaciers revealed striking similarities between all studied glaciers, i.e., several relatively small inner moraines, a large outer latero‐frontal moraine and similar configurations of proglacial lakes. This indicates that all glaciers were influenced by the same regional climatic forcing. Conversely, remarkable differences between the different settings concern‐ ing (i) size of the glacier foreland, (ii) altitude of glacier terminus, (iii) vegetation, and (iv) morphologi‐ cal details exist, all of which can be attributed to differences in local geomorphological situations and subsequent topoclimatic effects (cf. Rasmussen, 2013; Anderson et al., 2014).

(i) Differences in glacier size depend on aspect of the glaciers, controlling the amount of insolation, and the potential of the adjacent valley system to channel moisture toward the glacier, influencing the amounts of monsoonal precipitation (Evans, 1990; Evans and Cox, 2005; Loibl et al., 2014). This is particular obvious at Mt. Pulongu, because Gongpu and Xinpu are both descending from the same summit and exhibit very similar morphological catchment characteristics. Gongpu (south aspect) receives more insolation and is less susceptible to monsoonal precipitation (see also Maussion et al., 2014), resulting in a relatively small proglacial system. In contrast, Xinpu (north aspect) receives less insolation and more precipitation; consequently, the glacier foreland is remarkably larger here. At Zepu, low insolation and strong advection of monsoonal moisture are combined at a ∼ 20 km long, north‐facing range with ridge elevations >5500 m asl. This outstanding feature provides the setting for one of the largest glacier tongues and glacier forelands in the whole eastern Nyainqêntanglha Range.

(ii) Glacier tongue altitudes primarily depend on the same topoclimatic factors. For the investigated glaciers, the observed tongue altitudes from highest to lowest are at Gongpu (∼ 3800 m asl), Zepu (∼ 3350 m asl), and Xinpu (∼ 3200 m asl). Geochemical weathering analysis is showing the great im‐ pact that altitude and aspect have on the process regimes; in many indices the upper down‐valley samples (∼ 3670 m asl) from Gongpu valley at the southern side of Mt. Pulongu are closer to the samples from Xinpu’s foreland than to those from Gongpu (e.g., Ti–Zr, Rb–Sr, Na2O–K2O; Fig. 6.7C, D,

F). Noticeably, geochemical analysis of the different settings at Mt. Pulongu shows, in agreement with field observations, that bedrock lithologies are widely similar at the Mt. Pulongu glaciers (Fig. 6.7A, B). The grouping of samples by archive type indicates that sorting processes are more relevant for the differentiation than weathering or differences in parent material.

(iii) In close dependency of tongue altitudes, remarkable differences in vegetation cover of morpho‐ chronologically similar elements of the proglacial setting are observed. In altitudes of ∼ 4000 m asl at Gongpu, only pioneering plant species are present in the glacier foreland. Forests on the outer walls of LIA terminal moraines are sparse and consist of relatively few species, primarily P. balfouriana and L. griffithii. In contrast, vegetation is almost completely covering the proglacial area at Zepu glacier, which is reaching down to ∼ 3350 m asl. Here, the LIA moraines are covered by dense forest with a much more heterogeneous species composition (e.g., Juniperus, Acer, Aesculus). Even though the Xinpu foreland is reaching further down, the vegetation setting is intermediate between Gongpu and Zepu. This can be explained by shading effects imposed by the steep valley slopes and the glacier tongue’s position north of the 6400 m asl high summit.

(iv) Morphological details, such as size and distribution of moraines, development of glaciofluvial channels, meltwater outflows, proglacial lakes, etc., differ between all studied glaciers. This is a con‐ sequence of the high level of variability of the glacio‐geomorphic systems in the eastern Nyainqên‐ tanglha Range. Baitong is one of many glaciers in the region (cf. Loibl et al., 2014) where terminal moraines of previous advances are not preserved. This is clearly not a consequence of different cli‐ matic forcing mechanisms but of the individual glacier morphology with a narrow and relatively steep tongue. Therefore, a unifying sequence for glaciers in different settings, i.e., counting moraines, as proposed by early studies in the European Alps (e.g., the first morphostratigraphy by Penck and Brückner, 1909) is not appropriate in this complex high mountain environment (cf. Lehmkuhl, 1998; Winkler and Matthews, 2010). In many cases, however, different details still show evidence of similar forcing (see chapter 6.5.3). This high local variability introduced by different local geomorphological settings is in accordance with the findings from a previous remote sensing study by Loibl et al. (2014) which emphasizes the high relevance of such effects for glacial settings in the wider region.

6.5.3 Glacial chronology Two review articles attempted to obtain a late Holocene glacial chronology for the eastern Nyainqên‐ tanglha Range through a synoptic assessment of available dating results (Yang, B., et al., 2008; Yi et al., 2008). The results of these studies imply that glacial advances in southeastern Tibet were highly heterogeneous in space and time, with different glaciers advancing at different times throughout the last 2–3 ka. Both studies include, however, questionable 14C and lichenometric dating results for which information on sampling sites, sampling methods, sampled materials, and laboratory methods is widely lacking. In contrast, many dendrochronological results published from the region solely cov‐ er approximately the last 300 years (cf. Yang, B., et al., 2008), often concluding that the LIA was lim‐ ited to this time frame in southeastern Tibet (e.g., Xu et al., 2012).

The large outer latero‐frontal moraines show morphological evidence of more than one advance in all three settings that were studied in detail: two large lobes at Xinpu, a set of push moraines at Gongpu, and two large plus two minor lobes at Zepu. These morphostratigraphic parallels point to‐ ward a regional climatic signal which triggered several glacier advances. As evidenced by the lateral 92 moraine breach at Xinpu, there must have been an obstacle blocking the preexisting true right dis‐ charge, assumedly a dead ice relict from an immediately preceding advance (Fig. 6.8). This is also supported by the formation of a moraine peak and adjacent steep inner slopes at the vertex of the two outer moraine lobes (Fig. 6.8C, cf. Fig. 6.3B), implying a switch in the dominant moraine for‐ mation process from dumping to pushing. Thus, the succession of glacier advances probably occurred during a relatively short time span. Dendrochronological analysis of the oldest tree ages provided minimum ages of beginning glacier retreat: by this evidence, the advances at Xinpu must have oc‐ curred before 1662 CE (true right lobe, Fig. 6.8A) and 1743 CE (true left lobe, Fig. 6.8C).

Conversely, geochemical analysis of moraine sediments showed very few indications of weathering. In the applicable indices (Fig. 6.7C–E), moraine samples are predominantly grouped by different glac‐ iers, not by sequence. Therefore, a differentiation of the moraine stages by their grade of weather‐ ing, as demonstrated e.g., by Bäumler et al. (2004) for late Glacial and LGM moraines, is not possible for the investigated settings. A pre‐LIA age of moraine formation is very unlikely considering this low grade of weathering despite the extremely humid and relatively warm climate at the glacier termini. This is further supported by the state of plant succession in the proglacial domain, which is very young in all investigated settings, even though pioneering trees were observed on top of debris‐ covered ice at Xinpu. As demonstrated by the push moraine sequence at Gongpu, evidence of the earlier of these two major LIA advances may be obscured in many settings because the second ad‐ vance overrode and masked its landforms.

At none of the studied glaciers, moraines were found between the LIA moraines and the moraines of local LGM –typically called “Baiyu moraines” in the literature (cf. Zhou et al., 2007)– neither in the field nor in the subsequent remote sensing‐aided mapping. During the survey of a remote sensing‐ based study of ∼ 2000 glaciers in the wider region (Loibl et al., 2014) no moraines between these stages were encountered as well. Thus, an early neoglacial advance reaching beyond the LIA maxi‐ mum glacier extent has to be rejected for the eastern Nyainqêntanglha Range. However, for many of the questionable 14C dating results > 1000 BP (Li et al., 1986; Iwata and Jiao, 1993; Jiao et al., 2005), a reworking of material from earlier advances by the LIA maximum advance is highly probable, sug‐ gesting that minor late Holocene glacier advances prior to the LIA maximum advances of the 17th and 18th century existed. The old tree from Baitong glacier shows that the lateral moraine it grew on was built before ∼ 1540 CE, substantiating the existence of an early LIA advance which probably occurred as a consequence of colder or wetter climatic conditions associated with the Spörer minimum (Fig. 6.9). No reliable dating data exists for the inner recessional moraine stages at the investigated glaci‐ ers. However, morphological constraints, sedimentological analysis results, and fieldwork impres‐ sions are indicating young ages of formation. Geochemical results support a similar sequence for Xinpu and Gongpu glaciers. In agreement with dating results from morphostratigraphically similar settings in other studies (e.g., Bräuning and Lehmkuhl, 1996; Xu et al., 2012) it is thus suggested that these moraines represent the advances of the late 19th and early 20th century.

Fig. 6.8: Conceptual morphochronology for Xinpu (see Fig. 6.4 for legend): (A) early phase of the LIA maximum glacier advance until ∼ 1660 CE, formation of true right latero‐frontal moraine, (B) rela‐ tively short phase of retreat ∼ 1660–1680 CE, trees beginning to grow on lateral moraines, (C) second LIA maximum advance ∼ 1680–1740 CE, creating the true left lobe as a consequence of a lateral mo‐ raine breach that probably resulted from a remaining debris‐covered dead ice body; creation of very steep slope and peak, trees are getting pushed down and incorporated into the moraine body, (D) short phase of stability or slight re‐advance during ongoing retreat, probably early 19th century, de‐ tachment of dead ice block in true left lobe, tree growth on true left moraines, (E) short phase of sta‐ bility or slight re‐advance during ongoing retreat, probably late 19th to early 20th century, (F) short phase of stability or slight re‐advance during ongoing retreat, probably mid to late 20th century.

Integrating all constraints on morphostratigraphy and chronology, three main phases of LIA glacier advances in the eastern Nyainqêntanglha Range emerge: an early phase before ∼ 1500 CE, an inter‐ mediate phase from ∼ 1650 to ∼ 1740 CE, consisting of several strong advances and intermediary slight retreats, and a late phase manifested in a set of inner recessional moraines, dated to ∼ 1800– 1820 CE, ∼ 1880–1920 CE, and ∼ 1960–1985 CE in many studies (e.g., Bräuning and Lehmkuhl, 1996; Xu et al., 2012; Zhu et al., 2013). In comparison to observations from other high mountain environ‐ ments, noticeable differences in the timing of LIA glacier fluctuations are obvious; specifically, the last prominent LIA glacier advance in the eastern Nyainqêntanglha Range consists of multiple ad‐ vances and was earlier than the last large LIA advance in the European Alps (1800–1860 CE; Holzhauser et al., 2005) or the Rocky Mountains (1800–1810 CE; Luckmann, 2000). However, the 94 emerging multi‐phase chronological pattern which also includes a pre‐Maunder advance fits much better to other observations than results from previous chronological review studies.

Considering only probabilities >10% of the ambiguous 14C dating result, three time slices remain eli‐ gible for the growth of the tree: 1695–1725 CE, 1813–1836 CE, and 1876–1917 CE. Since the oldest trees on top of the moraine body in which the tree was incorporated started growing in the mid‐18th century it can, however, be deduced that, in this particular case, the oldest values are probably clos‐ est to the true age of the tree (cf. Fig. 6.9A, Fig. 6.9B). The age uncalibrated age of 60 ± 45 cal BP is clearly rejected by this observation. In combination with the field observations showing that the tree was lying almost horizontally and was hardly damaged, it seems probable that the tree grew on the true left moraine of the first Maunder advance and was then pushed down and covered in situ with debris by the second advance. If this concept is correct, the tree must have grown between the first and the second Maunder advance. Considering the possible age results of the 14C analysis and den‐ drochronologically derived minimum moraine ages, this is only true for the oldest calibrated 14C age, i.e., 1695–1725 CE (cf. Fig. 6.7A,B). This would, in turn, allow for narrowing down the time frame for the first advance at Xinpu to ∼ 1650–1680 CE and for the second advance to ∼ 1695–1740 CE.

6.5.4 Glaciers and paleoclimate In order to evaluate whether identifying the paleoclimatic events that triggered the advances is pos‐ sible, the aforementioned morphostratigraphic and chronologic constraints were compared to rele‐ vant paleoclimate proxy and forcing data from other studies (Fig. 6.9). Specifically, negative tempera‐ ture and positive precipitation excursions prior to oldest tree ages were sought. For this purpose, two regional and one northern hemispheric proxy datasets were chosen (Fig. 6.9D–F). Additionally, three global forcing curves (Fig. 6.9G–I) and phases of prolonged low sun activity, also referred to as ‘prolonged sunspot minima’ (e.g., Eddy, 1976), are displayed (vertical bars in Fig. 6.9).

The spatially closest precipitation anomaly reconstruction available to date was conducted by Grießinger et al. (2011) using tree δ18O analyses at Reting near Lhasa. Their results show that rela‐ tively high precipitation prevailed throughout the LIA. Temporal correlation of precipitation peaks and glacier advances is weak, implying that precipitation changes probably played a subordinate role in forcing the investigated glacier fluctuations. The temperature reconstructions for the northern hemisphere (Shi et al., 2013) and for the eastern Nyainqêntanglha by Zhu et al. (2011) generally re‐ semble the minima which are evident in records of solar forcing, e.g., reconstructions of solar modu‐ lation from 14C data (Muscheler et al., 2007; Solanki et al., 2004) and solar activity from 10Be data (Steinhilber et al., 2009). The strength and distribution of negative deviations differs, however, be‐ tween the reconstructions. Noticeable examples of contrary trends between the regional and north‐ ern hemispheric curves are the 18th century and the phase of the Spörer minimum. Taking the uncer‐ tainties of both reconstructions into account, this demonstrates interfering directly from paleocli‐ mate proxies to glacier fluctuations is still highly problematic. Consistently in both datasets, the Da‐ mon and Dalton minima are showing only relatively slight and short negative excursions whereas the Maunder minimum is characterized by strong and long lasting negative values. In the regional recon‐ struction, a succession of distinct ups and downs is evident at the pinnacle of the Maunder minimum (∼ 1660–1710 CE); these cannot be correlated to the solar activity curves but fit relatively well to a succession of volcanic events during that time (Crowley and Unterman, 2013). This pattern is less

95 pronounced in the northern hemispheric reconstruction, even though a cool–warm–cool succession is evident in the 17th century as well. Such patterns fit to the rapid succession of advances and re‐ treats which is evident in the moraine chronology (5.3, cf. tree ring dating results from Xinpu in Fig. 6.9). For the smaller re‐advances after 1800 CE, oldest tree ages correlate closely to climatic pessima, possibly indicating an accelerated succession in a generally warmer climate.

Fig. 6.9: Compilation of dating results from this and selected other studies, and paleoclimate recon‐ structions: (A) oldest‐tree moraine dating results, this study, vertical black lines show the calculated germination date, (B) 14C dating result, this study, light grey bars indicate alternative interpretations of the result which are invalidated by morphological constraints, (C) oldest‐tree moraine dating re‐ sults younger than 1800 CE from selected other studies (see text), 1 Lhamcoka (Bräuning and Lehmkuhl, 1996), 2 Lhamcoka (Bräuning 2006), 3 Gawalong, only terminal moraine crest positions (Zhu et al., 2013), 4 Midui, only terminal moraine crest positions (Xu et al., 2012), (D) temperature reconstruction for southeastern Tibet (Zhu et al., 2011; solid line is the 11 yr running mean), (E) recon‐ struction of summer precipitation anomalies from Reting (Grießinger et al., 2011; solid line is the 11 yr running mean), (F) multi‐proxy reconstruction of northern hemisphere temperature anomalies (Shi et al., 2013; base period 1001–1960 CE, solid line is the 20 yr running mean), (G) reconstruction of volcanic forcing derived from ice core sulphate (Crowley and Unterman, 2013), (H) 14C based recon‐

96 struction of solar modulation (Muscheler et al., 2007; solid line is the 20 yr running mean), (I) recon‐ structed sun activity (Steinhilber et al., 2009).

6.6 Conclusions Our results demonstrate that the reliability of inferences concerning the timing of glacier advances can be increased with multi‐proxy approaches including morphostratigraphic, sedimentological and different dating methods. The morphostratigraphical analysis revealed clear patterns reoccurring at all investigated sites, i.e., a major outer moraine ridge with indications of several advances and for‐ mation of proglacial lakes. Except for the problematic OSL dating results, all constraints from dating, morphostratigraphy and sedimentology indicate that this maximum late Holocene advance occurred during the LIA. Specifically, results from all of our settings indicate that the major LIA moraine was formed by a relatively quick succession of glacier advances around the Maunder minimum. Mor‐ phostratigraphical, chronological and geochemical analyses yielded evidence that no neoglacial ad‐ vance of the same size or larger than the LIA advance occurred in the eastern Nyainqêntanglha Range. An almost 500 year old tree from a lateral moraine at Baitong glacier indicates that an earlier LIA advance existed. However, the fact that all trees on terminal moraines in this and other studies are considerably younger suggests that this advance was smaller than the LIA maximum advance. Even though no numerical dating results for the inner set of recessional moraines were obtained, their relative chronology is clear. Comparison with dating results from similar southeastern Tibetan settings in the literature suggests a formation during often described phases in the early‐19th, late‐ 19th to early‐20th and mid‐20th centuries. In summary, our findings indicate three phases of LIA glacier advances; a pattern that is remarkably much closer to other records from the northern hemisphere than previous studies suggested: Neither were strong neoglacial advances evident nor was the LIA maximum advance limited to the phase after 1750 CE. A conceptual glacier chronology for the east‐ ern Nyainqêntanglha Range has thus been achieved. This implies common large‐scale climatic forcing for all glaciers in the region. Our results suggest that temperature changes are predominantly con‐ trolling the LIA glacier fluctuations in the investigated settings. Proxy data indicates monsoonal pre‐ cipitation was high throughout the LIA, providing sufficient accumulation material. However, mon‐ soonal precipitation and temperatures are closely linked through cloud cover in the study area. Fur‐ thermore, the morphological details of the studied glaciers are heterogeneous, underlining the great influence of local topoclimatic settings. Numerical dating of Holocene moraines remains a great chal‐ lenge in this remote area, as demonstrated for OSL and 14C methods. Thus, further work is needed to obtain better dating results; beside minimum ages from tree rings, high‐sensitivity cosmogenic nu‐ clide dating is a promising approach to achieve absolute ages of LIA moraine deposits in the future. Additionally, more paleoclimate archives are needed to further clarify the regional climate forcing mechanisms during the last 1000 years. We suggest that future studies on glacier change and paleo‐ climate, at least in complex high mountain environments such as the eastern Nyainqêntanglha Range, should carefully consider whether and how the local geomorphological configuration may influence the investigated archives, owing to the great impact this factor may have on the interpreta‐ tion of the settings.

7 Synthesis

Resulting from the interdisciplinary nature of the topic, numerous aspects of glaciology, geomor‐ phology, sedimentology, and climatology, were touched in the previous chapters. The objective of this concluding chapter is to summarize and mutually link the main conclusions in order to synthesize an overall picture. With the objective to contribute to the understanding of late Holocene glacier change in the eastern Nyainqêntanglha Range, the research summarized in this dissertation had two main foci. The first focus was to assess the local ELA distribution and post‐LIA ELA changes, examine the underlying patterns, and draw inferences regarding the relevant forcing mechanisms. The second focus was to disentangle the complex proglacial settings and late Holocene moraine sequences in the study area.

7.1 Challenges and opportunities In the initial design of the research project, the two main foci were scheduled in reverse order, in a classic inductive approach. The idea was to go to the field first to conduct local case studies. Subse‐ quently, remote sensing analyses would be performed in order to extrapolate the local results to the wider region. However, the first field campaign was cancelled and further campaigns were substan‐ tially limited by local authorities due to security concerns, so fieldwork was delayed and restricted in terms of both staff and time. As a result, the research concept had to be altered and work com‐ menced with regional‐scale remote sensing analyses and subsequent concise fieldwork. This also included the acquisition of additional remote sensing data which was achieved by a successful pro‐ posal for TanDEM‐X data and intensive searches for freely available satellite imagery of sufficient quality. As a consequence, the overall research concept changed to a deductive approach. Starting on a regional scale with remote‐sensing‐based mapping and spatial analyses, the first phase aimed to obtain insights regarding broad distributions and patterns. The altered plan revealed that focused fieldwork on local settings would be increasingly important in the second half of the project, to allow for testing of the proposed hypothesis on the regional scale. Additionally, fieldwork sites that prom‐ ised to be of outstandingly high explanatory value were identified during the preceding remote sens‐ ing phase. This largely compensated for the disadvantages resulting from administrative obstacles, but also required prolonging of the project.

7.2 Glaciers and ELAs: patterns and driving forces of distribution and post‐LIA changes Originally, GIS‐based glacier analysis was supposed to rely on shapefiles provided by the CGI (cf. chapter 4.3.1). The substantial shortcomings of this dataset required, however, the creation of a new inventory by manual interpretative mapping based on freely available satellite imagery. Even though this required several months of work, it allowed for the collection of further data, e.g., debris cover and glacier morphology. Additionally, the time‐consuming map work also provided an in‐depth study of the mountain range. Many hypotheses and concepts that were extremely valuable in later phases of the project emerged during this phase. Finally, ∼ 2000 glaciers were mapped, covering glaciated areas in a full Landsat ETM+ scene comprising the whole central part of the eastern Nyainqêntanglha

Range. The quality and level of detail of the resulting local glacier inventory are considerably better than the CGI data from the area (Fig. 7.1). The delineation of glacier margins under debris cover re‐ mained, however, subject to substantial uncertainties, despite major efforts to deduce glacier mar‐ gins from morphological characteristics.

Fig. 7.1: Comparison of CGI glacier outlines and glacier outlines mapped in this study (see also Fig. 4.4).

Length measurements revealed that glaciers retreated by ∼ 27% on average since the LIA maximum advance. Smaller/shorter glaciers tend to exhibit higher retreat rates than larger/longer ones. The same applies for clean‐ice glaciers in comparison to debris‐covered ones. Both findings are in good agreement with reports from neighboring areas (e.g., Bolch, 2007; Bolch et al., 2010). These dynam‐ ics can be explained by the longer response times of larger glaciers (Gross et al., 1976; Kulkarni, 2007) and the protective influence of debris cover (Østrem, 1959; Nakawo and Young, 1981; Scherler et al., 2011b).

The study area’s data situation for ELA calculation was particularly challenging since no multiannual mass‐balance series were available. Additionally, all hypsometric data, i.e., all available DEMs, exhib‐ ited major artifacts. As a consequence, numerous concepts and methods were evaluated, leading to the insight that only a toe‐to‐summit‐ratio‐approach was suitable within the given boundary condi‐ tions. Since these methods are, however, known to be crude and susceptible to errors (Benn and Lehmkuhl, 2000), a substantial effort was made to improve robustness. This included (i) evaluating different concepts (TSAM, THAR, TRAM) and different ratios using the late summer TSL, (ii) identify‐ ing criteria to exclude unsuitable glaciers, i.e., valley floor limitation and reconstitution, (iii) and com‐ pensating local errors by averaging in a large dataset.

The TRAM method with a ratio of 0.5 performed the best, reaching a significant correlation of 0.847 to late summer TSLs. Owing to the distinct effects of insolation on local ELAs, regional interpolation 99 required the separate addressing of the northern and southern aspects. The resulting regionally av‐ eraged ELAs display effects of all major relevant topoclimatic features. This demonstrates the high potential of the refined TRAM method to provide ELA data at an intermediate level of detail, ranging between highly sophisticated AABR studies, which require lots of high‐quality input data, and large‐ scale interpolations, which lack detail (e.g., Shi and Li, 1981). The key advantage of the refined TRAM method is its potential to rapidly derive large datasets.

The topoclimatic features evident in the ELA calculation results include (i) channeling of monsoonal moisture by the valley system, (ii) orographic precipitation in luv positions, (iii) foehn effects in lee positions, and (iv) a general SE–NW gradient, i.e., mountain front to TP, from 4400 to 5600 m asl. These features reveal a closely interlinked geosystem: the tectonically prescribed mountain topogra‐ phy exerts strong influence on the pathways of monsoonal moisture which, in turn, adds further considerable local heterogeneity to conditions for glaciers in the upper reaches of the mountain ranges. This includes both approximately longitudinal running valleys acting as transport pathways and east–west trending ridges acting as orographic barriers. Successive depletion of moisture toward the TP yields the observed precipitation gradient which, in turn, also increasingly limits glacier accu‐ mulation and linear erosion by both glaciers and rivers. As a consequence, relief ceases toward the TP which also reduces the valley’s efficiency as a transport pathway for monsoonal moisture.

Once the ratio has been determined from the modern situation, the refined TRAM method can also be applied to past glacier advances by using moraine evidence. In the study at hand, this was done for the LIA maximum advance. The results show an average ΔELA of ∼ 136 m with substantial small scale variability. This variability is well‐visible in the IDW interpolation, displaying the differences at individual glaciers as an abundantly bumpy structure dominating the image (Fig. 4.13). Whether this variability is (a) also resulting from actual local heterogeneity caused by the reaction to changing climatic boundary conditions or (b) just noise resulting from methodological uncertainties cannot be determined at this point. However, the application of an interpolation algorithm that smoothes more local variability, i.e., kriging, resulted in a picture that reveals clear spatial trends (chapter 5.5). Spe‐ cifically, ΔELA values exhibit a substantial correlation to continentality of the local climate, i.e., glaci‐ ers at sites with high/low average annual precipitation react more/less sensitive. This is in accord‐ ance with findings from other regions (cf. chapter 2.3.1).

In addition to the aforementioned climatic forcing mechanisms, insolation exerts a strong influence on the whole region due to its position around 30° N. Aspect is therefore one of the most dominant controlling factors for the local climatic regimes. Aspect‐wise analysis of length reduction (Fig. 4.9E) as well as modern and LIA ELAs reveal complex patterns (Fig. 4.14). For the former, length reduction is –contrary to expectations– strongest on east‐facing glaciers. Beyond the gradients which were detected already in the other datasets, the rose plots of the latter yield indications of a moisture source to the east. This is also in agreement with the mono‐ and bimodal precipitation regimes in Nyingchi and Bomi, respectively (Fig. 4.2).

7.3 Proglacial settings and late Holocene moraine sequences During the field campaigns, two glacier forelands at Mount Pulongu were investigated using a multi‐ proxy‐approach including detailed geomorphological mapping, sedimentological analysis, and sam‐ 100 pling for different dating techniques. Geomorphological mapping revealed remarkable differences regarding the sizes of the glacier forelands, the altitudes of glacier termini, the abundance and com‐ position of vegetation, and many geomorphological details, including: the size and distribution of moraines; and the development of glaciofluvial channels, meltwater outflows, and proglacial lakes (Fig. 6.4, suppl. Fig. B.14). All observed differences were in accordance with expectations from the above‐stated climate–topography–glacier interactions; fieldwork thus corroborated these findings, showing that these forces also exert substantial influence on the characteristics of individual pro‐ glacial environments and moraine sequences. In contrast to previous studies that imply highly heter‐ ogeneous late Holocene glacier fluctuations (e.g., Yang, B., et al., 2008; Yi et al., 2010), our field in‐ vestigations revealed strong similarities between the morphological key features at different study sites. In particular, the late Holocene moraine sequence at both settings consisted of a large outer latero‐frontal moraine and three minor inner moraines. Both glaciers’ outer moraines show evidence of at least two advances that occurred in a relatively rapid succession; specifically, the latero‐frontal moraines at Xinpu glacier consist of two large lobes and the pendant at Gongpu glacier consists of a set of push moraines. Similar configurations are evident at the large Zepu glacier which was studied by a combination of remote sensing analysis and fieldwork (at Zepu, fieldwork was conducted exclu‐ sively by the Erlangen workgroup owing to permit restrictions; detailed results are published in Hochreuther et al., 2015).

This indicates that the same climatic event may trigger the formation of heterogeneous landform arrangements at different sites, making a good understanding of the glacial morphology a crucial prerequisite for interpreting the landforms correctly. Against this background, the insights obtained from remote sensing analysis were of great value for subsequent fieldwork. The findings regarding similar moraine configurations facilitated the establishment of a common morphosequence, substan‐ tiating the assumption of common climatic forcing in the eastern Nyainqêntanglha Range.

In order to obtain a late Holocene chronosequence of glacier fluctuations in the study area, the mor‐ phosequence had to be complemented with sedimentological laboratory analysis and dating results. The findings from grain size analysis corroborated the assumption that the sediments in the foremost lobe at Gongpu glacier originate from a proglacial lake and can thus be parallelized to the setting at Xinpu. In contrast to the large lake setting at Gongpu, grain size distributions from these lake rem‐ nant samples showed stillwater characteristics similar to those of the proglacial lake at Xinpu, indi‐ cating that the lake was decoupled from the main drainage system. Geochemical indices did not re‐ sult in a clear differentiation of the moraine stages at the individual settings; however, lacking differ‐ entiation by weathering processes, despite the relatively warm and humid conditions, provides a strong indication of a late Holocene age of formation. Conversely, the aforementioned topoclimatic factors left clear imprints that became obvious in the comparison of geochemical indices from the different glacier forelands. Against this background, the results obtained from OSL dating indicating ages > 40 ka for each sample clearly had to be rejected. Presumably, the reason for this failure was the poor sensitivity of feldspars which had to be used due to the contamination of quartz grains. Consequently, the chronosequence had to be based widely on tree ring dating results which, fortu‐ nately, reveal many details, particularly at Xinpu glacier.

101

A conceptual chronosequence for the region was established by complementing the constraints from Xinpu and Gonpu with data from the literature, further tree ring dating results from Zepu and Baitong glaciers in the Bodui valley (Hochreuther et al., 2015), and a 14C age from a buried tree at Xinpu. It consists of two main phases of LIA advances, specifically an early advance before 1530 CE (correlating approximately to the Spörer Minimum) and a pronounced advance at ∼ 1650–1740 CE (correlating approximately to the Maunder Minimum). Evidence for the early advance is rare, proba‐ bly because it was smaller than the succeeding one. The tree on the lateral moraine at Baitong, which presumably survived subsequent advances because of extraordinary favorable conditions at its stand, therefore marks an exceptional case. The regional maximum advance probably consisted of several sub‐advances in a relatively rapid succession as evidenced by multiple moraine lobes at Xinpu and Zepu and by a push moraine sequence at Gonpu (Fig. 6.4; suppl. Fig. B.14); the 14C age obtained from a tree buried in Xinpu’s major latero‐frontal moraine suggests a first advance at ∼ 1650–1680 CE and a second advance at ∼ 1695–1740 CE. Two to three less–pronounced recessional moraines are evident in the investigated settings. No dating results could be obtained for these moraines, but other studies from the region found ages of ∼ 1800–1920 CE for settings of similar morphostratigra‐ phy, indicating their formation can be broadly equated to the Dalton and Damon Minima (e.g., Bräuning and Lehmkuhl, 1996).

7.4 Synoptic conclusions, desiderata and outlook The studies presented in this dissertation were conducted under application of an interdisciplinary multi‐proxy approach. Many of the findings emerged from considerations that explicitly included different methods, concepts, scales or proxies. In the context of complex settings such as the eastern Nyainqêntanglha Range, this approach performed considerably well and has demonstrated substan‐ tial advantages in comparison to studies limited to applying one method at one site. Specifically, limi‐ tations and uncertainties of individual methods can be identified much more clearly. For example, geochemical weathering indices showed that the obtained OSL ages were faulty, rendering the method inapplicable in the settings under investigation (chapter 6.5.3). Furthermore, knowledge of different aspects enables a holistic understanding of the system under investigation. For instance, numerous inferences could be drawn from remote‐sensing‐based analysis of climate–topography– glacier interactions to geomorphological details at the glacier forelands studied in the field (chapters 4.5.2, 6.5.2). This holistic approach might eventually also reveal shortcomings in interpretations made based on a single dataset. For example, observations on the relationships of moraine morphol‐ ogy, tree stands, and age distributions, enabled a much more selective consideration of tree‐ring‐ based moraine dating results published in the literature (chapter 6.5.3).

Another methodological feature that arose from making a virtue of necessity was the remote‐ sensing‐first approach, which resulted from reversing the original schedule of fieldwork and remote sensing (chapter 7.1). Retrospectively, the reversal turned out to be of great benefit because the insights gained in the remote sensing phase provided fundamental knowledge for the subsequent steps. In particular, the findings regarding climate–topography–glacier interactions (summarized in chapter 7.2) helped greatly in identifying settings of high exploratory value as well as in their inter‐ pretation. Starting with remote sensing, a database –of great value on its own– was created prior to the fieldwork, allowing smart querying of data in the subsequent phases. Conversely, performing a

102 regional extrapolation subsequent to fieldwork would have been extremely challenging without this knowledge, even if more settings could have been visited in the field. Therefore, the ‘remote‐ sensing‐first’ approach is recommended from the start in similar future projects.

Given the importance the eastern Nyainqêntanglha Range has for our understanding of the mon‐ soonal dynamics and glacier fluctuations in the wider region, it remains distinctly understudied to date. Even though the studies presented here can be considered as great progress in the investigated fields, there is a considerable need for further research.

In particular, paleoclimatic data is still scarce and widely limited to tree ring‐based studies. Thus, the local datasets are mostly of very high (i.e., annual) resolution, and of high sensitivity to variations. Conversely, trees react to a multitude of influencing factors and tree ring data is, as yet, limited to the late Holocene. Further datasets and different proxy archives are needed to improve our under‐ standing of the driving forces, as well as to extend the paleoclimate reconstructions. For example, further clarification is required regarding the question of whether the EASM does or does not pene‐ trate into the study area. Several results in this study indicate that there is a slight influence and that it has decreased since the LIA (chapter 4.5.2). However, to be conclusive, these hypotheses need to be substantiated by (paleo‐)climatological evidence. Regarding the TP with its complex atmospheric circulation patterns and high level of regional climate heterogeneity (chapter 2.1), the variety of dif‐ ferent relevant factors poses great challenges for reconstructing past climates based on stable iso‐ tope measurements. Specifically, different vapor sources and possible fractionation, as well as mixing processes, need to be considered to allow robust interpretations. Consequently, the validity of paleoclimatic proxy inferences beyond a regional context must be questioned. In particular, transfer‐ ring proxy‐based climate reconstructions between regions with different dominant air masses, e.g., monsoon‐ to westerly‐dominated, EASM‐ to ISM‐dominated, etc., should be avoided unless the valid‐ ity of the transfer could be proven in detail.

The forcing of strong LIA glacier advances also demands further clarification. For example, Thompson et al. (2005) suggest that precipitation is the primary driver of glaciation in the low latitudes. Davis et al., (2005) found, however, that LIA accumulation at Dasopu glacier in the Himalayas exceeds the summer monsoon amounts in India by ∼30%. As an explanation for this phenomenon, the authors suggest higher‐than‐today early winter snowfall owing to stronger westerly influence. Such concepts challenge tree‐ring‐based paleoprecipitation reconstructions, jeopardizing the uniformitarian funda‐ ment. Specifically, the proxy records in trees are hardly adding information outside of the trees’ growth periods; if paleoprecipitation patterns comprise a distinct shift to winter snowfall, such changes may not be recorded in the proxies and may also bias the statistical approaches used for extrapolating present day observations of climate‐proxy‐correlations into the past.

In order to corroborate the late Holocene glacial chronology developed in chapter 6, further dating results are required. In particular, the failure of OSL dating needs to be compensated by the applica‐ tion of methods which are more suitable for the high‐energy environment of the study area’s glacier forelands. The precision of cosmogenic nuclide dating has increased substantially during recent years, allowing the dating of very young glacial boulders. Even though challenges in calibration re‐ main, cosmogenic nuclide dating is the most promising technique to complement tree‐ring‐based moraine dating results. 103

Comparing the glaciers at Xinpu, Gongpu and Zepu, it is evident that the capability of a glacial system to record climatic fluctuations is correlated positively to its size. Specifically, the largest glacier, Zepu, exhibits two major and several minor lobes in the LIA latero‐frontal moraine sequence (suppl. Fig. B.14), whereas Xinpu has only two lobes, and Gongpu, the smallest glacier, only a set of push‐ moraines (Fig. 6.4). In this context, it would have been of particular interest to include Zepu’s fore‐ land into the geomorphological and sedimentological field studies, which sadly was not possible due to access restrictions. Specifically, Jiao et al. (2005) report pre‐LIA late Holocene moraines at this setting. The presented morphological and chronological evidence is, however, far from conclusive. Consequently, this would be the right spot to revisit the question of pre‐LIA late Holocene advances in the eastern Nyainqêntanglha Range which had to be neglected based on the findings from Xinpu and Gongpu.

In the near future, the methods developed and data acquired in the studies summarized in this dis‐ sertation will be critical in joint studies, many in cooperation with colleagues from neighboring fields of science. First, a modelling approach will be applied to assess glacier dynamics and test whether the implications of paleoclimatological proxy data are in accordance with observed post‐LIA glacier fluctuations. Subsequently, glacier flow velocities will be measured from multi‐temporal TanDEM‐X data, facilitating geomorphological interpretation by contributing to a better understanding of the influences exerted by, for instance, debris cover, dead ice, aspect, and glacier size. If the Landsat 8 mission provides a late summer image with few clouds and snow during the next years, a follow‐up glacier mapping of the study area would be of great interest in identifying and quantifying glacier changes since 1999. The refined TRAM method has great potential for application in other study are‐ as, promising to obtain ELA data on orogen scale at an unprecedented level of detail.

104

Overall references

Aa, A.R., 1996. Topographic control of equilibrium‐line altitude depression on reconstructed “Little Ice Age” glaciers, Grovabreen, western Norway. The Holocene 6, 82–89.

Agnihotri, R., Dutta, K., Bhushan, R., Somayajulu, B.L.K., 2002. Evidence for solar forcing on the Indian monsoon during the last millennium. Earth and Planetary Science Letters 198, 521–527. doi:10.1016/S0012‐821X(02)00530‐7

Aichner, B., Herzschuh, U., Wilkes, H., Schulz, H.‐M., Wang, Y., Plessen, B., Mischke, S., Diekmann, B., Zhang, C., 2012. Ecological development of Lake Donggi Cona, north‐eastern Tibetan Plateau, since the late glacial on basis of organic geochemical proxies and non‐pollen palynomorphs. Palaeogeogra‐ phy, Palaeoclimatology, Palaeoecology 313–314, 140–149. doi:10.1016/j.palaeo.2011.10.015

Aitken, M.J., 1998. An Introduction to Optical Dating ‐ The Dating of Quaternary Sediments by the Use of Photon‐stimulated Luminescence. Oxford University Press, Oxford, New York, Tokyo.

Aitken, M.J., 1985. Thermoluminescence dating. Academic press, London.

Alexanderson, H., Murray, A.S., 2012. Problems and potential of OSL dating Weichselian and Holo‐ cene sediments in Sweden. Quaternary Science Reviews 44, 37–50. doi:10.1016/j.quascirev.2009.09.020

Allégre, C.J., Courtillot, V., Tapponnier, P., Hirn, A., Mattauer, M., Coulon, C., Jaeger, J.J., Achache, J., Schärer, U., Marcoux, J., Burg, J.P., Girardeau, J., Armijo, R., Gariépy, C., Göpel, C., Tindong, L., Xuchang, X., Chenfa, C., Guangqin, L., Baoyu, L., Jiwen, T., Naiwen, W., Guoming, C., Tonglin, H., Xibin, W., Wanming, D., Huaibin, S., Yougong, C., Ji, Z., Hongrong, Q., Peisheng, B., Songchan, W., Bixiang, W., Yaoxiu, Z., Xu, R., 1984. Structure and evolution of the Himalaya–Tibet orogenic belt. Nature 307, 17–22. doi:10.1038/307017a0

Allen, J.R.L., Thornley, D.M., 2004. Laser granulometry of Holocene estuarine silts: effects of hydro‐ gen peroxide treatment. The Holocene 14, 290–295. doi:10.1191/0959683604hl681rr

Anders, A.M., Roe, G.H., Hallet, B., Montgomery, D.R., Finnegan, N.J., Putkonen, J., 2006. Spatial pat‐ terns of precipitation and topography in the Himalaya. Geological Society of America Special Papers 398, 39–53. doi:10.1130/2006.2398(03)

Anderson, B., Mackintosh, A., 2012. Controls on mass balance sensitivity of maritime glaciers in the Southern Alps, New Zealand: The role of debris cover. Journal of Geophysical Research: Earth Surface 117, F01003. doi:10.1029/2011JF002064

Anderson, D.M., Overpeck, J.T., Gupta, A.K., 2002. Increase in the Asian Southwest Monsoon During the Past Four Centuries. Science 297, 596 – 599. doi:10.1126/science.1072881

Anderson, L.S., Roe, G.H., Anderson, R.S., 2014. The effects of interannual climate variability on the moraine record. Geology 42, 55–58. doi:10.1130/G34791.1

105

Anderson, W.T., Bernasconi, S.M., McKenzie, J.A., Saurer, M., 1998. Oxygen and carbon isotopic rec‐ ord of climatic variability in tree ring cellulose (Picea abies): An example from central Switzerland (1913–1995). Journal of Geophysical Research: Atmospheres 103, 31625–31636. doi:10.1029/1998JD200040

Andrews, J.T., 1975. Glacial Systems: An Approach to Glaciers and their Environments. Duxbury Press, North Scituate.

An, Z., Colman, S.M., Zhou, W., Li, X., Brown, E.T., Jull, A.J.T., Cai, Y., Huang, Y., Lu, X., Chang, H., Song, Y., Sun, Y., Xu, H., Liu, W., Jin, Z., Liu, X., Cheng, P., Liu, Y., Ai, L., Li, X., Liu, X., Yan, L., Shi, Z., Wang, X., Wu, F., Qiang, X., Dong, J., Lu, F., Xu, X., 2012. Interplay between the Westerlies and Asian monsoon recorded in Lake Qinghai sediments since 32 ka. Scientific Reports 2. doi:10.1038/srep00619

Argand, E., 1924. La tectonique de l’Asie, in: Conférence Faite Á Bruxelles, Le 10 Août 1922. Present‐ ed at the Congrès géologique international (XIIIe session)‐ Belgique 1922, Imprimerie Vaillant‐ Carmanne, Liége, pp. 172–372.

Armijo, R., Tapponnier, P., Han, T., 1989. Late Cenozoic right‐lateral strike‐slip faulting in southern Tibet. Journal of Geophysical Research: Solid Earth 94, 2787–2838. doi:10.1029/JB094iB03p02787

Arrell, K., Wise, S., Wood, J., Donoghue, D., 2008. Spectral filtering as a method of visualising and removing striped artefacts in digital elevation data. Earth Surface Processes and Landforms 33, 943– 961. doi:10.1002/esp.1597

Ashwell, I.Y., 1982. Glacier mapping in Iceland by the British schools’ Exploring Society. The Geo‐ graphical Journal 148, 308–316.

Bahr, D.B., Pfeffer, W.T., Sassolas, C., Meier, M.F., 1998. Response time of glaciers as a function of size and mass balance: 1. Theory. Journal of Geophysical Research: Solid Earth 103, 9777–9782. doi:10.1029/98JB00507

Balco, G., 2011. Contributions and unrealized potential contributions of cosmogenic‐nuclide expo‐ sure dating to glacier chronology, 1990–2010. Quaternary Science Reviews 30, 3–27. doi:10.1016/j.quascirev.2010.11.003

Balco, G., Stone, J.O., Lifton, N.A., Dunai, T.J., 2008. A complete and easily accessible means of calcu‐ lating surface exposure ages or erosion rates from 10Be and 26Al measurements. Quaternary Geo‐ chronology, Prospects for the New Frontiers of earth and Environmental Sciences 3, 174–195. doi:10.1016/j.quageo.2007.12.001

Ballantyne, C.K., 1997. Periglacial trimlines in the Scottish Highlands. Quaternary International 38–39, 119–136. doi:10.1016/S1040‐6182(96)00016‐X

Bamzai, A.S., Shukla, J., 1999. Relation between Eurasian Snow Cover, Snow Depth, and the Indian Summer Monsoon: An Observational Study. Journal of Climate 12, 3117–3132. doi:10.1175/1520‐ 0442(1999)012<3117:RBESCS>2.0.CO;2

Bard, E., 1998. Geochemical and geophysical implications of the radiocarbon calibration∗. Geo‐ chimica et Cosmochimica Acta 62, 2025–2038. doi:10.1016/S0016‐7037(98)00130‐6

106

Barr, I.D., Lovell, H., 2014. A review of topographic controls on moraine distribution. Geomorphology 226, 44–64. doi:10.1016/j.geomorph.2014.07.030

Barry, R.G., Chorley, R.J., 2009. Atmosphere, Weather and Climate. Routledge, Abingdon.

Bäumler, R., 2004. Pedogenic studies in eastern Nepal–new aspects about the last glaciation. Qua‐ ternary Science Reviews 23, 1777 – 1783. doi:10.1016/j.quascirev.2004.03.003

Bäumler, R., 2001. Pedogenic studies in aeolian deposits in the high mountain area of eastern Nepal. Quaternary International 76‐77, 93 – 102.

Bäumler, R., Kemp, M., Zech, W., Heuberger, H., Siebert, A., Madhikermi, D., Poudel, K., 1996. Soil weathering on glacial and glaciofluvial deposits in the Langtang Valley (Central Nepal) and its relation to the glacial history. Zeitschrift für Geomorphologie 103, 373–387.

Bäumler, R., Zech, W., Heuberger, H., Weber‐Diefenbach, K., 1991. Investigations on the intensity of weathering of soils developed from glacial and fluvioglacial deposits and their relationship with the history of the landscape in the mt. everest region. Geoderma 48, 223 – 243.

Beer, J., Raisbeck, G.M., Yiou, F., 1991. Time variations of 10Be and solar activity, in: The Sun in Time, Space Science Series. University of Arizona Press, Tuscon, p. 343.

Benedict, J.B., 1968. Recent glacial history of an alpine area in the Colorado Front Range, U.S.A. II. Dating the glacial deposits. Journal of Glaciology 7, 77–87.

Benedict, J.B., 1967. Recent glacial history of an alpine area in the Colorado Front Range, U.S.A. I. Establishing a lichen‐growth curve. Journal of Glaciology 6, 817–832.

Benn, D.I., Bolch, T., Hands, K., Gulley, J., Luckman, A., Nicholson, L.I., Quincey, D., Thompson, S., Toumi, R., Wiseman, S., 2012. Response of debris‐covered glaciers in the Mount Everest region to recent warming, and implications for outburst flood hazards. Earth‐Science Reviews 114, 156–174. doi:10.1016/j.earscirev.2012.03.008

Benn, D.I., Evans, D.J.A., 2010. Glaciers und Glaciation, 2nd ed. Hodder Education, London.

Benn, D.I., Gemmell, A.M.D., 1997. Calculating equilibrium‐line altitudes of former glaciers by the balance ratio method: a new computer spreadsheet. Glacial Geology and Geomorphology.

Benn, D.I., Lehmkuhl, F., 2000. Mass balance and equilibrium‐line altitudes of glaciers in high‐ mountain environments. Quaternary International 65‐66, 15–29. doi:10.1016/S1040‐6182(99)00034‐ 8

Benn, D.I., Owen, L.A., 1998. The role of the Indian summer monsoon and the mid‐latitude westerlies in Himalayan glaciation: review and speculative discussion. Journal of the Geological Society, London 155, 353–363. doi:10.1144/gsjgs.155.2.0353

Benn, D.I., Owen, L.A., Osmaston, H.A., Seltzer, G.O., Porter, S.C., Mark, B., 2005. Reconstruction of equilibrium‐line altitudes for tropical and sub‐tropical glaciers. Quaternary International 138–139, 8– 21. doi:10.1016/j.quaint.2005.02.003

107

Berger, G.W., 1988. Dating Quaternary events by luminescence. Geological Society of America Spe‐ cial Papers 227, 13–50. doi:10.1130/SPE227‐p13

Berger, G.W., 1986. Dating Quaternary Deposits by Luminescence ‐ Recent Advances. Geoscience Canada 13, 15–21.

Berger, G.W., 1984. Thermoluminescence dating studies of glacial silts from Ontario. Canadian Jour‐ nal of Earth Sciences 21, 1393–1399. doi:10.1139/e84‐144

Beschel, R.E., 1961. Dating rock surfaces by lichen growth and its application to glaciology and physi‐ ography (lichenometry), in: Rasch, G.O. (Ed.), Geology of the Arctic. University of Toronto Press, Toronto, pp. 1044–1062.

Beschel, R.E., 1950. Flechten als Altersmassstab rezenter Moränen. Zeitschrift für Gletscherkunde und Glazialgeologie 1, 152–161.

Bhambri, R., Bolch, T., 2009. Glacier mapping: a review with special reference to the Indian Himala‐ yas. Progress in Physical Geography 33, 672–704. doi:10.1177/0309133309348112

Bickel, L., Lüthgens, C., Lomax, J., Fiebig, M., 2015. Luminescence dating of glaciofluvial deposits linked to the penultimate glaciation in the Eastern Alps. Quaternary International, Correlations of Quaternary Fluvial, Eolian, Deltaic and Marine Sequences: SEQS 2013 357, 110–124. doi:10.1016/j.quaint.2014.10.013

Bindschadler, R., Dowdeswell, J., Hall, D., Winther, J.‐G., 2001. Glaciological applications with Land‐ sat‐7 imagery: early assessments. Remote Sensing of Environment 78, 163–179. doi:10.1016/S0034‐ 4257(01)00257‐7

Blockey, S.P.E., Housley, R.A., 2009. Calibration Commentary. Radiocarbon 51, 287–290. doi:10.2458/azu_js_rc.51.3490

Böhner, J., 2006. General climatic controls and topoclimatic variations in Central and High Asia. Bore‐ as 35, 279–295. doi:10.1111/j.1502‐3885.2006.tb01158.x

Böhner, J., 1994. Circulation and representativeness of precipitation and air temperature in the southeast of the Quinghai‐Xizang Plateau. GeoJournal 34, 55–66. doi:10.1007/BF00813970

Böhner, J., Lehmkuhl, F., 2005. Environmental change modelling for Central and High Asia: Pleisto‐ cene, present and future scenarios. Boreas 34, 220–231. doi:10.1111/j.1502‐3885.2005.tb01017.x

Bolch, T., 2007. Climate change and glacier retreat in northern Tien Shan (Kazakhstan/Kyrgyzstan) using remote sensing data. Global and Planetary Change 56, 1–12. doi:10.1016/j.gloplacha.2006.07.009

Bolch, T., Buchroithner, M., Pieczonka, T., Kunert, A., 2008. Planimetric and volumetric glacier chang‐ es in the Khumbu Himal, Nepal since 1962 using Corona, Landsat TM and ASTER data. Journal of Glac‐ iology 54, 592–600.

108

Bolch, T., Kulkarni, A., Kääb, A., Huggel, C., Paul, F., Cogley, J.G., Frey, H., Kargel, J.S., Fujita, K., Scheel, M., Bajracharya, S., Stoffel, M., 2012. The State and Fate of Himalayan Glaciers. Science 336, 310– 314. doi:10.1126/science.1215828

Bolch, T., Yao, T., Kang, S., Buchroithner, M.F., Scherer, D., Maussion, F., Huintjes, E., Schneider, C., 2010. A glacier inventory for the western Nyainqêntanglha Range and the Nam Co basin, Tibet, and glacier changes 1976–2009. The Cryosphere 4, 419–433. doi:10.5194/tc‐4‐419‐2010

Bond, G., Showers, W., Cheseby, M., Lotti, R., Almasi, P., deMenocal, P., Priore, P., Cullen, H., Hajdas, I., Bonani, G., 1997. A Pervasive Millennial‐Scale Cycle in North Atlantic Holocene and Glacial Cli‐ mates. Science 278, 1257–1266. doi:10.1126/science.278.5341.1257

Bookhagen, B., Burbank, D.W., 2006. Topography, relief, and TRMM‐derived rainfall variations along the Himalaya. Geophysical Research Letters 33, L08405. doi:10.1029/2006GL026037

Boos, W.R., Kuang, Z., 2013. Sensitivity of the South Asian monsoon to elevated and non‐elevated heating. Scientific Reports 3. doi:10.1038/srep01192

Boos, W.R., Kuang, Z., 2010. Dominant control of the South Asian monsoon by orographic insulation versus plateau heating. Nature 463, 218–222. doi:10.1038/nature08707

Booth, A.L., Chamberlain, C.P., Kidd, W.S.F., Zeitler, P.K., 2009. Constraints on the metamorphic evo‐ lution of the eastern Himalayan syntaxis from geochronologic and petrologic studies of Namche Barwa. Geological Society of America Bulletin 121, 385–407. doi:10.1130/B26041.1

Börner, N., De Baere, B., Yang, Q., Jochum, K.P., Frenzel, P., Andreae, M.O., Schwalb, A., 2013. Ostra‐ cod shell chemistry as proxy for paleoenvironmental change. Quaternary International, Proxies for Quaternary monsoon reconstruction on the Tibetan Plateau 313–314, 17–37. doi:10.1016/j.quaint.2013.09.041

Bøtter‐Jensen, L., McKeever, S.W.S., Wintle, A.G., 2003. Optically Stimulated Luminescence Dosime‐ try. Butterworth Heinemann, Amsterdam, Boston, London.

Bradley, R.S., Jones, P.D., 1993. “Little Ice Age” summer temperature variations: their nature and relevance to recent global warming trends. The Holocene 3, 367 – 376.

Bradwell, T., Armstrong, R.A., 2007. Growth rates of Rhizocarpon geographicum lichens: a review with new data from Iceland. Journal of Quaternary Science 22, 311–320. doi:10.1002/jqs.1058

Bräuning, A., 2006. Tree‐ring evidence of Little Ice Age glacier advances in southern Tibet. The Ho‐ locene 16, 369–380. doi:10.1191/0959683606hl922rp

Bräuning, A., Lehmkuhl, F., 1996. Glazialmorphologische und dendrochronologische Untersuchungen neuzeitlicher Eisrandlagen Ost‐ und Südtibets. Erdkunde 50, 341–359.

Bräuning, A., Mantwill, B., 2004. Summer temperature and summer monsoon history on the Tibetan plateau during the last 400 years recorded by tree rings. Geophysical Research Letters 31, L24205. doi:10.1029/2004GL020793

109

Briffa, K.R., Jones, P.D., Schweingruber, F.H., 1988. Summer temperature patterns over Europe: A reconstruction from 1750 A.D. based on maximum latewood density indices of conifers. Quaternary Research 30, 36–52. doi:10.1016/0033‐5894(88)90086‐5

Briffa, K.R., Jones, P.D., Schweingruber, F.H., Karlén, W., Shiyatov, S.G., 1996. Tree‐ring variables as proxy‐climate indicators: Problems with low‐frequency signals, in: Jones, P.D., Bradley, R.S., Jouzel, J. (Eds.), Climatic Variations and Forcing Mechanisms of the Last 2000 Years, NATO ASI Series. Springer, Heidelberg, pp. 9–41.

Brinkhurst, R.O., Chua, K.E., Batoosingh, E., 1969. Modifications in Sampling Procedures as Applied to Studies on the Bacteria and Tubificid Oligochaetes Inhabiting Aquatic Sediments. Journal of the Fish‐ eries Research Board of Canada 26, 2581–2593. doi:10.1139/f69‐252

Bronk Ramsey, C.B., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360. doi:10.2458/azu_js_rc.51.3494

Bronk Ramsey, C., Buck, C.E., Manning, S.W., Reimer, P., Plicht, H. van der, 2006. Developments in radiocarbon calibration for archaeology. Antiquity 80, 783–798. doi:10.1017/S0003598X00094424

Buggle, B., Glaser, B., Hambach, U., Gerasimenko, N., Marković, S., 2011. An evaluation of geochemi‐ cal weathering indices in loess–paleosol studies. Quaternary International, The Second Loessfest (2009) 240, 12–21. doi:10.1016/j.quaint.2010.07.019

Buggle, B., Glaser, B., Zöller, L., Hambach, U., Marković, S., Glaser, I., Gerasimenko, N., 2008. Geo‐ chemical characterization and origin of Southeastern and Eastern European loesses (Serbia, Romania, Ukraine). Quaternary Science Reviews 27, 1058–1075. doi:10.1016/j.quascirev.2008.01.018

Büntgen, U., Frank, D.C., Nievergelt, D., Esper, J., 2006. Summer Temperature Variations in the Euro‐ pean Alps, A.D. 755–2004. Journal of Climate 19, 5606–5623. doi:10.1175/JCLI3917.1

Burbank, D.W., Kang, J.C., 1991. Relative dating of Quaternary moraines, Rongbuk valley, Mount Ev‐ erest, Tibet: Implications for an ice sheet on the Tibetan Plateau. Quaternary Research 36, 1–18. doi:10.1016/0033‐5894(91)90013‐U

Burg, J.‐P., Nievergelt, P., Oberli, F., Seward, D., Davy, P., Maurin, J.‐C., Diao, Z., Meier, M., 1998. The Namche Barwa syntaxis: evidence for exhumation related to compressional crustal folding. Journal of Asian Earth Sciences 16, 239–252. doi:10.1016/S0743‐9547(98)00002‐6

Cai, Y., Tan, L., Cheng, H., An, Z., Edwards, R.L., Kelly, M.J., Kong, X., Wang, X., 2010. The variation of summer monsoon precipitation in central China since the last deglaciation. Earth and Planetary Sci‐ ence Letters 291, 21–31. doi:10.1016/j.epsl.2009.12.039

Cao, X., Herzschuh, U., Ni, J., Zhao, Y., Böhmer, T., 2015. Spatial and temporal distributions of major tree taxa in eastern continental Asia during the last 22,000 years. The Holocene 25, 79–91. doi:10.1177/0959683614556385

Capitanio, F.A., Replumaz, A., 2013. Subduction and slab breakoff controls on Asian indentation tec‐ tonics and Himalayan western syntaxis formation. Geochemistry, Geophysics, Geosystems 14, 3515– 3531. doi:10.1002/ggge.20171

110

Chang, C.P., 2004. East Asian Monsoon. World Scientific, New Jersey.

Chaplot, V., Darboux, F., Bourennane, H., Leguédois, S., Silvera, N., Phachomphon, K., 2006. Accuracy of interpolation techniques for the derivation of digital elevation models in relation to landform types and data density. Geomorphology 77, 126–141. doi:10.1016/j.geomorph.2005.12.010

Chemenda, A.I., Burg, J.‐P., Mattauer, M., 2000. Evolutionary model of the Himalaya–Tibet system: geopoem: based on new modelling, geological and geophysical data. Earth and Planetary Science Letters 174, 397–409. doi:10.1016/S0012‐821X(99)00277‐0

Chen, F., Yu, Z., Yang, M., Ito, E., Wang, S., Madsen, D.B., Huang, X., Zhao, Y., Sato, T., John B. Birks, H., Boomer, I., Chen, J., An, C., Wünnemann, B., 2008. Holocene moisture evolution in arid central Asia and its out‐of‐phase relationship with Asian monsoon history. Quaternary Science Reviews 27, 351–364. doi:10.1016/j.quascirev.2007.10.017

Chen, J., An, Z., Head, J., 1999. Variation of Rb/Sr Ratios in the Loess‐Paleosol Sequences of Central China during the Last 130,000 Years and Their Implications for Monsoon Paleoclimatology. Quater‐ nary Research 51, 215–219. doi:10.1006/qres.1999.2038

Chen, J., Chen, F., Feng, S., Huang, W., Liu, J., Zhou, A., 2015. Hydroclimatic changes in China and surroundings during the Medieval Climate Anomaly and Little Ice Age: spatial patterns and possible mechanisms. Quaternary Science Reviews 107, 98–111. doi:10.1016/j.quascirev.2014.10.012

Chinn, T.J.H., 1981. Use of Rock Weathering‐Rind Thickness for Holocene Absolute Age‐Dating in New Zealand. Arctic and Alpine Research 13, 33–45. doi:10.2307/1550624

Chou, C., 2003. Land–sea heating contrast in an idealized Asian summer monsoon. Climate Dynamics 21, 11–25. doi:10.1007/s00382‐003‐0315‐7

Clark, M.K., Schoenbohm, L.M., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X., Tang, W., Wang, E., Chen, L., 2004. Surface uplift, tectonics, and erosion of eastern Tibet from large‐scale drainage patterns. Tectonics 23, TC1006. doi:10.1029/2002TC001402

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

Coe, A.L., 2010. Geological Field Techniques. John Wiley & Sons, Chichester.

Cogley, J.G., Arendt, A.A., Braithwaite, R.J., Hock, R., Jansson, P., Kaser, G., Möller, M., Nicholson, L., Rasmussen, A., Zemp, M., 2011. Glossary of glacier mass balance and related terms, IHP‐VII Technical Documents in Hydrology. UNESCO‐IHP, Paris.

Colman, S.M., Yu, S.‐Y., An, Z., Shen, J., Henderson, A.C.G., 2007. Late Cenozoic climate changes in China’s western interior: a review of research on Lake Qinghai and comparison with other records. Quaternary Science Reviews 26, 2281–2300. doi:10.1016/j.quascirev.2007.05.002

Cook, E.R., Kairiukstis, L.A. (Eds.), 1990. Methods of Dendrochronology. Springer Netherlands, Dor‐ drecht.

111

Cook, E.R., Krusic, P.J., Jones, P.D., 2003. Dendroclimatic signals in long tree‐ring chronologies from the Himalayas of Nepal. International Journal of Climatology 23, 707–732. doi:10.1002/joc.911

Crowley, T.J., 2000. Causes of Climate Change Over the Past 1000 Years. Science 289, 270–277. doi:10.1126/science.289.5477.270

Crowley, T.J., Lowery, T.S., 2000. How Warm Was the Medieval Warm Period? AMBIO: A Journal of the Human Environment 29, 51–54. doi:10.1579/0044‐7447‐29.1.51

Crowley, T.J., Unterman, M.B., 2013. Technical details concerning development of a 1200 yr proxy index for global volcanism. Earth System Science Data 5, 187–197. doi:10.5194/essd‐5‐187‐2013

Cui, X., Graf, H.‐F., 2009. Recent land cover changes on the Tibetan Plateau: a review. Climatic Change 94, 47–61. doi:10.1007/s10584‐009‐9556‐8

Daansgard, W., 1964. Stable isotopes in precipitation. Tellus 16, 436–438.

Dabski, M., 2007. Testing the size‐frequency‐based lichenometric dating curve on Fláajökull moraines (SE Iceland) and quantifying lichen population dynamics with respect to stone surface aspect. Jökull 57, 21–35.

Dadson, S.J., Church, M., 2005. Postglacial topographic evolution of glaciated valleys: a stochastic landscape evolution model. Earth Surface Processes and Landforms 30, 1387–1403. doi:10.1002/esp.1199

Dahl, S.O., Bakke, J., Lie, Ø., Nesje, A., 2003. Reconstruction of former glacier equilibrium‐line alti‐ tudes based on proglacial sites: an evaluation of approaches and selection of sites. Quaternary Sci‐ ence Reviews 22, 275–287. doi:10.1016/S0277‐3791(02)00135‐X

Dallmeyer, A., Claussen, M., Wang, Y., Herzschuh, U., 2013. Spatial variability of Holocene changes in the annual precipitation pattern: a model‐data synthesis for the Asian monsoon region. Climate Dy‐ namics 40, 2919–2936. doi:10.1007/s00382‐012‐1550‐6

Daniels, F., Boyd, C.A., Saunders, D.F., 1953. Thermoluminescence as a Research Tool. Science 117, 343–349. doi:10.1126/science.117.3040.343

Dansgaard, W., Johnsen, S.J., Møller, J., Langway, C.C., 1969. One Thousand Centuries of Climatic Record from Camp Century on the Greenland Ice Sheet. Science 166, 377–380. doi:10.1126/science.166.3903.377

D’Arrigo, R.D., Jacoby, G.C., Free, R.M., 1992. Tree‐ring width and maximum latewood density at the North American tree line: parameters of climatic change. Canadian Journal of Forest Research 22, 1290–1296. doi:10.1139/x92‐171

Davi, N.K., Jacoby, G.C., Wiles, G.C., 2003. Boreal temperature variability inferred from maximum latewood density and tree‐ring width data, Wrangell Mountain region, Alaska. Quaternary Research 60, 252–262. doi:10.1016/j.yqres.2003.07.002

112

Davis, M.E., Thompson, L.G., Yao, T., Wang, N., 2005. Forcing of the Asian monsoon on the Tibetan Plateau: Evidence from high‐resolution ice core and tropical coral records. Journal of Geophysical Research: Atmospheres 110, D04101. doi:10.1029/2004JD004933

Day, M.J., 1980. Rock Hardness: Field Assessment and Geomorphic Importance. The Professional Geographer 32, 72–81. doi:10.1111/j.0033‐0124.1980.00072.x

Day, M.J., Goudie, A.S., 1977. Field assessment of rock hardness using the Schmidt Hammer. British Ceomorphological Research Group Technical Bulletin 18, 19–29.

De Geer, G., 1912. A geochronology for the last 12,000 years. Proceedings of the XIth International Geological Congress, Stockholm (1910) 1, 241–253.

Denton, G.H., Karlén, W., 1973. Lichenometry: Its Application to Holocene Moraine Studies in South‐ ern Alaska and Swedish Lapland. Arctic and Alpine Research 5, 347–372. doi:10.2307/1550128

De Vries, H., 1959. Measurement and use of natural radiocarbon, in: Abelson, P.H. (Ed.), Researches in Geochemistry. John Wiley & Sons, New York, pp. 169–89.

De Vries, H., 1958. Variation in concentration of radiocarbon with time and location on Earth. Pro‐ ceedings of the Koninklijke Nederlandse Akademie Van Wetenschappen Series B‐Palaeontology Ge‐ ology Physics Chemistry Anthropology B61, 94–102.

Dey, B., Bhanu Kumar, O.S.R.U., 1982. An Apparent Relationship between Eurasian Spring Snow Cov‐ er and the Advance Period of the Indian Summer Monsoon. Journal of Applied Meteorology 21, 1929–1932. doi:10.1175/1520‐0450(1982)021<1929:AARBES>2.0.CO;2

Ding, Y., 1992. Effects of the Qinghai‐Xizang (Tibetan) plateau on the circulation features over the plateau and its surrounding areas. Advances in Atmospheric Sciences 9, 112–130. doi:10.1007/BF02656935

DIN ISO 11277, 2002. Soil Quality ‐ Determination of particle size distribution in mineral soil material ‐ method by sieving and sedimentation.

Domrös, M., Peng, G., 1988. The Climate of China. Springer, Heidelberg, New York.

Dong, J., Wang, Y., Cheng, H., Hardt, B., Edwards, R.L., Kong, X., Wu, J., Chen, S., Liu, D., Jiang, X., Zhao, K., 2010. A high‐resolution stalagmite record of the Holocene East Asian monsoon from Mt Shennongjia, central China. The Holocene 20, 257–264. doi:10.1177/0959683609350393

Dreimanis, A., Hutt, G., Raukas, A., Whippey, P.W., 1978. Dating Methods of Pleistocene Deposits and Their Problems: I. Thermoluminescence Dating. Geoscience Canada 5, 55–60. doi:10.12789/gs.v5i2.3079

Duan, F., Wang, Y., Shen, C.‐C., Wang, Y., Cheng, H., Wu, C.‐C., Hu, H.‐M., Kong, X., Liu, D., Zhao, K., 2014. Evidence for solar cycles in a late Holocene speleothem record from Dongge Cave, China. Sci. Rep. 4. doi:10.1038/srep05159

113

Dugam, S.S., Kakade, S.B., Verma, R.K., 1997. Interannual and long‐term variability in the North At‐ lantic Oscillation and Indian Summer monsoon rainfall. Theoretical and Applied Climatology 58, 21– 29. doi:10.1007/BF00867429

Duller, G. a. T., 2004. Luminescence dating of quaternary sediments: recent advances. Journal of Quaternary Science 19, 183–192. doi:10.1002/jqs.809

Duller, G.A.T., 2003. Distinguishing quartz and feldspar in single grain luminescence measurements. Radiation Measurements 37, 161–165. doi:10.1016/S1350‐4487(02)00170‐1

Dunai, T.J., 2010. Cosmogenic Nuclides ‐ Principles, Concepts and Applications in the Earth Surface Sciences, Geochemistry and environmental chemist. Cambridge University Press, Cambridge.

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. doi:10.1016/j.epsl.2005.01.036

Eddy, J.A., 1976. The Maunder Minimum. Science 192, 1189–1202. doi:10.1126/science.192.4245.1189

Ehhalt, D., Knott, K., Nagel, J.F., Vogel, J.C., 1963. Deuterium and oxygen 18 in rain water. Journal of Geophysical Research 68, 3775–3780. doi:10.1029/JZ068i013p03775

Ehlers, J., Gibbard, P.L., 2004a. Quaternary Glaciations ‐ Extent and Chronology ‐ Part I: Europe. De‐ velopments in Quaternary Science. Elsevier, Amsterdam, San Diego.

Ehlers, J., Gibbard, P.L., 2004b. Quaternary Glaciations ‐ Extent and Chronology: Part II: North Ameri‐ ca. Developments in Quaternary Science. Elsevier, Amsterdam, San Diego.

Ehlers, J., Gibbard, P.L., 2004c. Quaternary Glaciations ‐ Extent and Chronology ‐ Part III: South Amer‐ ica, Asia, Africa, Australia, Antarctica. Developments in Quaternary Science. Elsevier, Amsterdam, San Diego.

Eicher, U., Siegenthaler, U., 1976. Palynological and oxygen isotope investigations on Late‐Glacial sediment cores from Swiss lakes. Boreas 5, 109–117. doi:10.1111/j.1502‐3885.1976.tb00335.x

Erikstad, L., Sollid, J.L., 1986. Neoglaciation in South Norway using lichenometric methods. Norsk Geografisk Tidsskrift ‐ Norwegian Journal of Geography 40, 85–105. doi:10.1080/00291958608552159

Esper, J., Cook, E.R., Schweingruber, F.H., 2002. Low‐Frequency Signals in Long Tree‐Ring Chronolo‐ gies for Reconstructing Past Temperature Variability. Science 295, 2250–2253. doi:10.1126/science.1066208

Evans, I.S., 2006. Local aspect asymmetry of mountain glaciation: a global survey of consistency of favoured directions for glacier numbers and altitudes. Geomorphology 73, 166–184. doi:10.1016/j.geomorph.2005.07.009

Evans, I.S., 1990. Climatic effects on glacier distribution across the southern coast mountains, B.C., Canada. Annals of Glaciology 14, 58–64.

114

Evans, I.S., Cox, N.J., 2005. Global variations of local asymmetry in glacier altitude: separation of northsouth and eastwest components. Journal of Glaciology 51, 469–482. doi:10.3189/172756505781829205

Faccenna, C., Becker, T.W., Conrad, C.P., Husson, L., 2013. Mountain building and mantle dynamics. Tectonics 32, 80–93. doi:10.1029/2012TC003176

Fairchild, I.J., Baker, A., 2012. Speleothem Science: From Process to Past Environments. John Wiley & Sons, Hoboken, NJ.

Ferrat, M., Weiss, D.J., Dong, S., Large, D.J., Spiro, B., Sun, Y., Gallagher, K., 2012a. Lead atmospheric deposition rates and isotopic trends in Asian dust during the last 9.5 kyr recorded in an ombrotrophic peat bog on the eastern Qinghai–Tibetan Plateau. Geochimica et Cosmochimica Acta, Environmental Records of Anthropogenic Impacts 82, 4–22. doi:10.1016/j.gca.2010.10.031

Ferrat, M., Weiss, D.J., Spiro, B., Large, D., 2012b. The inorganic geochemistry of a peat deposit on the eastern Qinghai‐Tibetan Plateau and insights into changing atmospheric circulation in central Asia during the Holocene. Geochimica et Cosmochimica Acta 91, 7–31. doi:10.1016/j.gca.2012.05.028

Feulner, G., 2011a. Comment on “Strong signature of the active Sun in 100 years of terrestrial insola‐ tion data” by W. Weber. Annalen der Physik 523, 946–950. doi:10.1002/andp.201100179

Feulner, G., 2011b. Are the most recent estimates for Maunder Minimum solar irradiance in agree‐ ment with temperature reconstructions? Geophysical Research Letters 38, L16706. doi:10.1029/2011GL048529

Fiebig, M., Ellwanger, D., Doppler, G., 2011. Pleistocene glaciations of southern Germany, in: Ehlers, J., Gibbard, P.L., Hughes, P.D. (Eds.), Developments in Quaternary Sciences, Quaternary Glaciations ‐ Extent and Chronology, A Closer Look. Elsevier, Amsterdam, pp. 163–173.

Fielding, E., Isacks, B., Barazangi, M., Duncan, C., 1994. How flat is Tibet? Geology 22, 163 – 167.

Fielding, E.J., 2000. Morphotectonic evolution of the Himalayas and Tibetan Plateau, in: Summerfield, M.A. (Ed.), Geomorphology and Global Tectonics. John Wiley & Sons, Chichester, pp. 201–222.

Finnegan, N.J., Hallet, B., Montgomery, D.R., Zeitler, P.K., Stone, J.O., Anders, A.M., Yuping, L., 2008. Coupling of rock uplift and river incision in the Namche Barwa–Gyala Peri massif, Tibet. Geological Society of America Bulletin 120, 142–155. doi:10.1130/B26224.1

Fischer, M.J., Treble, P.C., 2008. Calibrating climate‐δ18O regression models for the interpretation of high‐resolution speleothem δ18O time series. Journal of Geophysical Research: Atmospheres 113, D17103. doi:10.1029/2007JD009694

Flint, R.F., 1971. Glacial and Quaternary Geology. Wiley, New York.

Flohn, H., 1968. Contributions to a Meteorology of the Tibetan Highlands. Tech. Rep. 130, Colorado State Univ.

115

Flohn, H., 1953. Hochgebirge und allgemeine Zirkulation. II. Die Gebirge als Wärmequellen. Archiv für Meteorologie, Geophysik und Bioklimatologie A 5, 265–279.

Fort, M., 2014. Natural hazards versus climate change and their potential impacts in the dry, north‐ ern Himalayas: focus on the upper Kali Gandaki (Mustang District, Nepal). Environmental Earth Sci‐ ences 1–14. doi:10.1007/s12665‐014‐3087‐y

Friedman, I., Machta, L., Soller, R., 1962. Water‐vapor exchange between a water droplet and its environment. Journal of Geophysical Research 67, 2761–2766. doi:10.1029/JZ067i007p02761

Fritts, H.C., 2001. Tree Rings and Climate. The Blackburn Press, Caldwell, N.J.

Fritts, H.C., Blasing, T.J., Hayden, B.P., Kutzbach, J.E., 1970. Multivariate Techniques for Specifying Tree‐Growth and Climate Relationships and for Reconstructing Anomalies in Paleoclimate. Laborato‐ ry of Tree‐Ring Research Archives. The University of Arizona.

Fu, C., Fletcher, J.O., 1985. The Relationship between Tibet‐Tropical Ocean Thermal Contrast and Interannual Variability of Indian Monsoon Rainfall. Journal of Climate and Applied Meteorology 24, 841–847. doi:10.1175/1520‐0450(1985)024<0841:TRBTTO>2.0.CO;2

Fuchs, M., Owen, L.A., 2008. Luminescence dating of glacial and associated sediments: review, rec‐ ommendations and future directions. Boreas 37, 636–659. doi:10.1111/j.1502‐3885.2008.00052.x

Fu, C., Jiang, Z., Guan, Z., He, J., Xu, Z., 2008. Climate of China and East Asian Monsoon, in: Fu, C., Jiang, Z., Guan, Z., He, J., Xu, Z. (Eds.), Regional Climate Studies of China, Regional Climate Studies. Springer, Heidelberg, pp. 1–48.

Furbish, D.J., Andrews, J.T., 1984. The use of hypsometry to indicate long‐term stability and response of valley glaciers to changes in mass transfer. Journal of Glaciology 30, 199–211.

Gaar, D., Lowick, S.E., Preusser, F., 2014. Performance of different luminescence approaches for the dating of known‐age glaciofluvial deposits from northern Switzerland. Geochronometria 41, 65–80. doi:10.2478/s13386‐013‐0139‐0

Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., Olley, J.M., 1999. Optical Dating of Single and Multiple Grains of Quartz from Jinmium Rock Shelter, Northern Australia: Part I, Experimental Design and Statistical Models*. Archaeometry 41, 339–364. doi:10.1111/j.1475‐4754.1999.tb00987.x

Gallet, S., Jahn, B., Torii, M., 1996. Geochemical characterization of the Luochuan loess‐paleosol se‐ quence, China, and paleoclimatic implications. Chemical Geology 133, 67–88. doi:10.1016/S0009‐ 2541(96)00070‐8

Gardelle, J., Berthier, E., Arnaud, Y., Kääb, A., 2013. Region‐wide glacier mass balances over the Pa‐ mir‐Karakoram‐Himalaya during 1999–2011. The Cryosphere 7, 1263–1286. doi:10.5194/tc‐7‐1263‐ 2013

Gardner, A.S., Moholdt, G., Cogley, J.G., Wouters, B., Arendt, A.A., Wahr, J., Berthier, E., Hock, R., Pfeffer, W.T., Kaser, G., Ligtenberg, S.R.M., Bolch, T., Sharp, M.J., Hagen, J.O., Broeke, M.R. van den, Paul, F., 2013. A Reconciled Estimate of Glacier Contributions to Sea Level Rise: 2003 to 2009. Science 340, 852–857. doi:10.1126/science.1234532 116

Gardner, J.S., Hewitt, K., 1990. A surge of Bualtar Glacier, Karakorum Range, Pakistan: a possible landslide trigger. Journal of Glaciology 36, 159–162.

Gauthier, G., Burke, A.L., 2011. The effects of surface weathering on the geochemical analysis of ar‐ chaeological lithic samples using non‐destructive polarized energy dispersive XRF. Geoarchaeology 26, 269–291. doi:10.1002/gea.20346

Gellatly, A.F., Chinn, T.J.H., Röthlisberger, F., 1988. Holocene glacier variations in New Zealand: A review. Quaternary Science Reviews 7, 227–242. doi:10.1016/0277‐3791(88)90008‐X

Gerlitz, L., Conrad, O., Böhner, J., 2015. Large‐scale atmospheric forcing and topographic modifica‐ tion of precipitation rates over High Asia – a neural‐network‐based approach. Earth System Dynamics 6, 61–81. doi:10.5194/esd‐6‐61‐2015

Gindl, W., Grabner, M., Wimmer, R., 2000. The influence of temperature on latewood lignin content in treeline Norway spruce compared with maximum density and ring width. Trees 14, 409–414. doi:10.1007/s004680000057

Gloaguen, R., Ratschbacher, L., 2011. Growth and collapse of the Tibetan Plateau: introduction. Geo‐ logical Society, London, Special Publications 353, 1–8. doi:10.1144/SP353.1

Godfrey‐Smith, D.I., Huntley, D.J., Chen, W.‐H., 1988. Optical dating studies of quartz and feldspar sediment extracts. Quaternary Science Reviews 7, 373–380. doi:10.1016/0277‐3791(88)90032‐7

Gong, D.‐Y., Wang, S.‐W., Zhu, J.‐H., 2001. East Asian Winter Monsoon and Arctic Oscillation. Geo‐ physical Research Letters 28, 2073–2076. doi:10.1029/2000GL012311

Gosse, J.C., Phillips, F.M., 2001. Terrestrial in situ cosmogenic nuclides: theory and application. Qua‐ ternary Science Reviews 20, 1475–1560. doi:10.1016/S0277‐3791(00)00171‐2

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

Goudie, A.S., 2006. The Schmidt Hammer in geomorphological research. Progress in Physical Geogra‐ phy 30, 703 – 718.

Grießinger, J., Bräuning, A., Helle, G., Thomas, A., Schleser, G., 2011. Late Holocene Asian summer monsoon variability reflected by δ18O in tree‐rings from Tibetan junipers. Geophysical Research Let‐ ters 38, 1–5. doi:10.1029/2010GL045988

Gross, G., Kerschner, H., Patzelt, G., 1977. Methodische Untersuchungen über die Schneegrenze in alpinen Gletschergebieten. Zeitschrift für Gletscherkunde und Glazialgeologie 12, 223–251.

Grove, J.M., 2004. Little Ice Ages, 2nd ed, Routledge Studies in Physical Geography and Environment. Routledge, London.

Grove, J.M., Switsur, R., 1994. Glacial geological evidence for the medieval warm period. Climatic Change 26, 143–169. doi:10.1007/BF01092411

117

Gruber, S., Haeberli, W., 2007. Permafrost in steep bedrock slopes and its temperature‐related de‐ stabilization following climate change. Journal of Geophysical Research: Earth Surface 112, F02S18. doi:10.1029/2006JF000547

Guggenberger, G., Bäumler, R., Zech, W., 1998. Weathering of soils developed in eolian material overlying glacial deposits in Eastern Nepal. Soil Science 163, 325–337.

Günther, F., Aichner, B., Siegwolf, R., Xu, B., Yao, T., Gleixner, G., 2013. A synthesis of hydrogen iso‐ tope variability and its hydrological significance at the Qinghai–Tibetan Plateau. Quaternary Interna‐ tional, Proxies for Quaternary monsoon reconstruction on the Tibetan Plateau 313–314, 3–16. doi:10.1016/j.quaint.2013.07.013

Günther, F., Mügler, I., Mäusbacher, R., Daut, G., Leopold, K., Gerstmann, U.C., Xu, B., Yao, T., Gleixner, G., 2011. Response of δD values of sedimentary n‐alkanes to variations in source water isotope signals and climate proxies at lake Nam Co, Tibetan Plateau. Quaternary International, Qua‐ ternary Paleoenvironmental Change and Landscape Development in Tibet and the Bordering Moun‐ tains 236, 82–90. doi:10.1016/j.quaint.2010.12.006

Gupta, A.K., 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, 354–357. doi:10.1038/nature01340

Gupta, A.K., Das, M., Anderson, D.M., 2005. Solar influence on the Indian summer monsoon during the Holocene. Geophysical Research Letters 32, L17703, 1–4. doi:10.1029/2005GL022685

Haeberli, W., 2004. Glaciers and ice caps: historical background and strategies of world‐wide moni‐ toring, in: Bamber, J., Payne, A.J. (Eds.), Mass Balance of the Cryosphere. Cambridge University Press, Cambridge, pp. 559–578.

Hagg, W., Braun, L.N., Weber, M., Becht, M., 2006. Runoff modelling in glacierized Central Asian catchments for present‐day and future climate. Nordic Hydrology 93–105.

Hahn, D.G., Manabe, S., 1975. The Role of Mountains in the South Asian Monsoon Circulation. Jour‐ nal of the Atmospheric Sciences 32, 1515 – 1541.

Hahn, D.G., Shukla, J., 1976. An Apparent Relationship between Eurasian Snow Cover and Indian Monsoon Rainfall. Journal of the Atmospheric Sciences 33, 2461–2462. doi:10.1175/1520‐ 0469(1976)033<2461:AARBES>2.0.CO;2

Hajdas, I., 2008. Radiocarbon dating and its applications in Quaternary studies. Eiszeitalter und Ge‐ genwart Quaternary Science Journal 57, 2–24. doi:10.3285/eg.57.1‐2.1

Hamann, C., Embleton, C., 1988. Morphometrische Analyse der kare in den Salzburger Zentralalpen. Salzburger Geographische Arbeiten 17, 57–74.

Harrison, T.M., Copeland, P., Kidd, W.S.F., Yin, A., 1992. Raising Tibet. Science 255, 1663–1670. doi:10.1126/science.255.5052.1663

118

Harrison, W.D., Elsberg, D.H., Echelmeyer, K.A., Krimmel, R.M., 2001. On the characterization of glac‐ ier response by a single time‐scale. Journal of Glaciology 47, 659–664. doi:10.3189/172756501781831837

Hasegawa, S., Dahal, R.K., Yamanaka, M., Bhandary, N.P., Yatabe, R., Inagaki, H., 2008. Causes of large‐scale landslides in the Lesser Himalaya of central Nepal. Environmental Geology 57, 1423–1434. doi:10.1007/s00254‐008‐1420‐z

Hawkings, F.F., 1985. Equilibrium‐line altitudes and paleoenviroment in the Merchants Bay area, Baffin Island, N.W.T., Canada. Journal of Glaciology 31, 205–213.

Hendy, C.H., Wilson, A.T., 1968. Palaeoclimatic Data from Speleothems. Nature 219, 48–51. doi:10.1038/219048a0

Herb, C., Zhang, W., Koutsodendris, A., Appel, E., Fang, X., Pross, J., 2013. Environmental implications of the magnetic record in Pleistocene lacustrine sediments of the Qaidam Basin, NE Tibetan Plateau. Quaternary International, Proxies for Quaternary monsoon reconstruction on the Tibetan Plateau 313–314, 218–229. doi:10.1016/j.quaint.2013.06.015

Herzschuh, U., Borkowski, J., Schewe, J., Mischke, S., Tian, F., 2014. Moisture‐advection feedback supports strong early‐to‐mid Holocene monsoon climate on the eastern Tibetan Plateau as inferred from a pollen‐based reconstruction. Palaeogeography, Palaeoclimatology, Palaeoecology 402, 44–54. doi:10.1016/j.palaeo.2014.02.022

Herzschuh, U., Kramer, A., Mischke, S., Zhang, C., 2009. Quantitative climate and vegetation trends since the late glacial on the northeastern Tibetan Plateau deduced from Koucha Lake pollen spectra. Quaternary Research 71, 162–171. doi:10.1016/j.yqres.2008.09.003

Herzschuh, U., Mischke, S., Meyer, H., Plessen, B., Zhang, C., 2010. Lake nutrient variability inferred from elemental (C, N, S) and isotopic (δ13C, δ15N) analyses of aquatic plant macrofossils. Quaternary Science Reviews 29, 2161–2172. doi:10.1016/j.quascirev.2010.05.011

Herzschuh, U., Winter, K., Wünnemann, B., Li, S., 2006. A general cooling trend on the central Tibetan Plateau throughout the Holocene recorded by the Lake Zigetang pollen spectra. Quaternary Interna‐ tional, Quaternary palaeoenvironmental change in Tibet and the bordering mountains Quaternary palaeoenvironmental change in Tibet and the bordering mountains 154–155, 113–121. doi:10.1016/j.quaint.2006.02.005

Hewitt, K., 2009. Rock avalanches that travel onto glaciers and related developments, Karakoram Himalaya, Inner Asia. Geomorphology, Dating, triggering, modelling, and hazard assessment of large landslides 103, 66–79. doi:10.1016/j.geomorph.2007.10.017

Hewitt, K., 2005. The Karakoram Anomaly? Glacier Expansion and the “Elevation Effect,” Karakoram Himalaya. Mountain Research and Development 25, 332–340. doi:10.1659/0276‐ 4741(2005)025[0332:TKAGEA]2.0.CO;2

Hewitt, K., 1999. Quaternary Moraines vs Catastrophic Rock Avalanches in the Karakoram Himalaya, Northern Pakistan. Quaternary Research 51, 220–237. doi:10.1006/qres.1999.2033

119

Heyman, J., 2014. Paleoglaciation of the Tibetan Plateau and surrounding mountains based on expo‐ sure ages and ELA depression estimates. Quaternary Science Reviews 91, 30–41. doi:10.1016/j.quascirev.2014.03.018

Heyman, J., Stroeven, A.P., Harbor, J.M., Caffee, M.W., 2011. Too young or too old: Evaluating cos‐ mogenic exposure dating based on an analysis of compiled boulder exposure ages. Earth and Plane‐ tary Science Letters 302, 71–80. doi:10.1016/j.epsl.2010.11.040

Hind, A., Moberg, A., 2013. Past millennial solar forcing magnitude. Climate Dynamics 41, 2527– 2537. doi:10.1007/s00382‐012‐1526‐6

Hochreuther, P., Loibl, D., Wernicke, J., Zhu, H., Grießinger, J., Bräuning, A., 2015. Ages of major Little Ice Age glacier fluctuations on the southeast Tibetan Plateau derived from tree‐ring‐based moraine dating. Palaeogeography, Palaeoclimatology, Palaeoecology 422, 1–10. doi:10.1016/j.palaeo.2015.01.002

Hodges, K.V., 2000. Tectonics of the Himalaya and southern Tibet from two perspectives. Geological Society of America Bulletin 112, 324–350. doi:10.1130/0016‐7606(2000)112<324:TOTHAS>2.0.CO;2

Hoelzle, M., Chinn, T., Stumm, D., Paul, F., Zemp, M., Haeberli, W., 2007. The application of glacier inventory data for estimating past climate change effects on mountain glaciers: A comparison be‐ tween the European Alps and the Southern Alps of New Zealand. Climate Change Impacts on Moun‐ tain Glaciers and Permafrost 56, 69 – 82.

Hoelzle, M., Haeberli, W., Dischl, M., Peschke, W., 2003. Secular glacier mass balances derived from cumulative glacier length changes. Global and Planetary Change 36, 295 – 306. doi:10.1016/S0921‐ 8181(02)00223‐0

Höfer, H.V., 1879. Gletscher und Eiszeitstudien. Sitzungsberichte der Akademie der Wissenschaften Wien, math.‐phys. Klasse I 79, 331–367.

Holmes, R.L., 1983. Computer‐Assisted Quality Control in Tree‐Ring Dating and Measurement. Tree‐ Ring Bulletin 43, 69–78.

Holzhauser, H., Magny, M., Zumbuühl, H.J., 2005. Glacier and lake‐level variations in west‐central Europe over the last 3500 years. The Holocene 15, 789–801. doi:10.1191/0959683605hl853ra

Holzhauser, H., Zumbuehl, H.J., 1996. On the history of the Lower Grindelwald Glacier during the last 2800 years ‐ palaeosols, fossil wood and historical pictorial records ‐ new results. Zeitschrift für Geo‐ morphologie 104, 95–127.

Hormes, A., Müller, B.U., Schlüchter, C., 2001. The Alps with little ice: evidence for eight Holocene phases of reduced glacier extent in the Central Swiss Alps. The Holocene 11, 255–265. doi:10.1191/095968301675275728

Hua, Q., 2013. Atmospheric Radiocarbon for the Period 1950–2010. Radiocarbon 55, 2059–2072. doi:10.2458/azu_js_rc.v55i2.16177

120

Hu, C., Henderson, G.M., Huang, J., Xie, S., Sun, Y., Johnson, K.R., 2008. Quantification of Holocene Asian monsoon rainfall from spatially separated cave records. Earth and Planetary Science Letters 266, 221–232. doi:10.1016/j.epsl.2007.10.015

Hughes, M.K., Diaz, H.F., 1994. Was there a “medieval warm period”, and if so, where and when? Climatic Change 26, 109–142. doi:10.1007/BF01092410

Hülle, D., Hilgers, A., Radtke, U., Stolz, C., Hempelmann, N., Grunert, J., Felauer, T., Lehmkuhl, F., 2010. OSL dating of sediments from the Gobi Desert, Southern Mongolia. Quaternary Geochronolo‐ gy, 12th International Conference on Luminescence and Electron Spin Resonance Dating (LED 2008) 5, 107–113. doi:10.1016/j.quageo.2009.06.002

Huntley, D.J., Baril, M.R., 1997. The K content of the K‐feldspars being measured in optical dating or in thermoluminescence dating. Ancient TL 15, 11–13.

Hutter, K., 1983. Theoretical Glaciology ‐ Material Science of Ice and the Mechanics of Glaciers | Springer, Mathematical Approaches to Geophysics. Springer, New York.

IJmker, J., Stauch, G., Dietze, E., Hartmann, K., Diekmann, B., Lockot, G., Opitz, S., Wünnemann, B., Lehmkuhl, F., 2012a. Characterisation of transport processes and sedimentary deposits by statistical end‐member mixing analysis of terrestrial sediments in the Donggi Cona lake catchment, NE Tibetan Plateau. Sedimentary Geology 281, 166–179. doi:10.1016/j.sedgeo.2012.09.006

IJmker, J., Stauch, G., Hartmann, K., Diekmann, B., Dietze, E., Opitz, S., Wünnemann, B., Lehmkuhl, F., 2012b. Environmental conditions in the Donggi Cona lake catchment, NE Tibetan Plateau, based on factor analysis of geochemical data. Journal of Asian Earth Sciences, Asian Climate & Tectonics Asian Climate & Tectonics 44, 176–188. doi:10.1016/j.jseaes.2011.04.021

Immerzeel, W.W., van Beek, L.P.H., Bierkens, M.F.P., 2010. Climate change will affect the Asian water towers. Science 328, 1382–1385. doi:10.1126/science.1183188

IPCC, 2013. Summary for policymakers ‐ contribution of working group I to the fifth assessment re‐ port of the Intergovernmental Panel on Climate Change, in: Stocker, T.F., Qin, D., Plattner, G.‐K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M. (Eds.), Climate Change 2013: The Physical Science Basis. Cambridge University Press, Cambridge, New York, p. 27.

ISO 13320–1, 1999. Particle size analysis ‐ Laser Diffraction Methods, part 1: general principles, An‐ nex A: theoretical background of laser diffraction.

Iwata, S., Jiao, K., 1993. Fluctuations of the Zepu Glacier in late Holocene Epoch, the eastern Nyainqentanglha Mountains, in: Yao, T., Ageta, Y. (Eds.), Glaciation, Climate and Environment in Qinghai–Tibet Plateau. Science Press, Beijing, pp. 130–138.

Jacobsen, K., 2010. Comparison of ASTER GDEMs with SRTM height models, in: Remote Sensing for Science, Education and Culture. Presented at the 30th EARSeL Symposium, Paris.

Jacob, T., Wahr, J., Pfeffer, W.T., Swenson, S., 2012. Recent contributions of glaciers and ice caps to sea level rise. Nature 482, 514–518. doi:10.1038/nature10847

121

Jarvis, A., Reuter, H.I., Nelson, A., Guevara, E., 2008. Hole‐filled SRTM for the globe Version 4, availa‐ ble from the CGIAR‐CSI SRTM 90m Database [WWW Document]. URL http:/srtm.csi.cgiar.org (ac‐ cessed 2.14.14).

Jenness, J., 2006. Topographic Position Index (TPI) v. 1.3a [WWW Document]. URL http://www.jennessent.com/arcview/tpi.htm

Jiao, K., Iwata, S., Yao, T., Jing, Z., Li, Z., 2005. Variation of Zepu Glacier and Environmental Change in the Eastern Nyainqêtanglha Range since 3.2 ka BP. Journal of Glaciology and Geocryology 27, 736– 739.

Jiao, K., Zhang, Z., 1988. Glacier Inventory of China, VII: Qinghai‐Xizang Plateau Interior Area. Science Press, Beijing.

Jóhannesson, T., Raymond, C.F., Waddington, E.D., 1989. A Simple Method for Determining the Re‐ sponse Time of Glaciers, in: Oerlemans, J. (Ed.), Glacier Fluctuations and Climatic Change, Glaciology and Quaternary Geology. Springer Netherlands, Dordrecht, pp. 343–352.

Johnsen, S.J., Dahl‐Jensen, D., Dansgaard, W., Gundestrup, N., 1995. Greenland palaeotemperatures derived from GRIP bore hole temperature and ice core isotope profiles. Tellus B 47, 624–629. doi:10.1034/j.1600‐0889.47.issue5.9.x

Jouzel, J., 2013. A brief history of ice core science over the last 50 yr. Clim. Past 9, 2525–2547. doi:10.5194/cp‐9‐2525‐2013

Jouzel, J., Lorius, C., Petit, J.R., Genthon, C., Barkov, N.I., Kotlyakov, V.M., Petrov, V.M., 1987. Vostok ice core: a continuous isotope temperature record over the last climatic cycle (160,000 years). Nature 329, 403–408. doi:10.1038/329403a0

Ju, J., Slingo, J., 1995. The Asian summer monsoon and ENSO. Quarterly Journal of the Royal Meteor‐ ological Society 121, 1133–1168. doi:10.1002/qj.49712152509

Kääb, A., 2008. Remote sensing of permafrost‐related problems and hazards. Permafrost and Perigla‐ cial Processes 19, 107–136. doi:10.1002/ppp.619

Kääb, A., Berthier, E., Nuth, C., Gardelle, J., Arnaud, Y., 2012. Contrasting patterns of early twenty‐ first‐century glacier mass change in the Himalayas. Nature 488, 495–498. doi:10.1038/nature11324

Kang, S., Mayewski, P.A., Qin, D., Yan, Y., Hou, S., Zhang, D., Ren, J., Kruetz, K., 2002. Glaciochemical records from a Mt. Everest ice core: relationship to atmospheric circulation over Asia. Atmospheric Environment 36, 3351–3361. doi:10.1016/S1352‐2310(02)00325‐4

Kanner, L.C., Buenning, N.H., Stott, L.D., Timmermann, A., Noone, D., 2014. The role of soil processes in δ18O terrestrial climate proxies. Global Biogeochemical Cycles 28, 2013GB004742. doi:10.1002/2013GB004742

Kapp, P., Guynn, J.H., 2004. Indian punch rifts Tibet. Geology 32, 993 – 996.

Kargel, J.S., Leonard, G.J., Bishop, M.P., Kääb, A., Raup, B.H. (Eds.), 2014. Global Land Ice Measure‐ ments from Space, 2014 edition. ed. Springer, New York.

122

Karlén, W., 1984. Dendrochronology, Mass Balance and Glacier Front Fluctuations in Northern Swe‐ den, in: Mörner, N.‐A., Karlén, W. (Eds.), Climatic Changes on a Yearly to Millennial Basis. Springer Netherlands, Amsterdam, pp. 263–271.

Kaser, G., Osmaston, H., 2002. Tropical Glaciers. Cambridge University Press, Cambridge.

Kasper, T., Haberzettl, T., Doberschütz, S., Daut, G., Wang, J., Zhu, L., Nowaczyk, N., Mäusbacher, R., 2012. Indian Ocean Summer Monsoon (IOSM)‐dynamics within the past 4 ka recorded in the sedi‐ ments of Lake Nam Co, central Tibetan Plateau (China). Quaternary Science Reviews 39, 73–85. doi:10.1016/j.quascirev.2012.02.011

Kattelmann, R., 2003. Glacial Lake Outburst Floods in the Nepal Himalaya: A Manageable Hazard? Natural Hazards 28, 145–154. doi:10.1023/A:1021130101283

Kayastha, R.B., Harrison, S.P., 2008. Changes of the equilibrium line altitude since the Little Ice Age in the Nepalese Himalayas. Annals of Glaciology 48, 93–99.

Kienast, F., Schweingruber, F.H., Bräker, O.U., Schär, E., 1987. Tree‐ring studies on conifers along ecological gradients and the potential of single‐year analyses. Canadian Journal of Forest Research 17, 683–696. doi:10.1139/x87‐111

King, G.E., Robinson, R.A.J., Finch, A.A., 2014a. Towards successful OSL sampling strategies in glacial environments: deciphering the influence of depositional processes on bleaching of modern glacial sediments from Jostedalen, Southern Norway. Quaternary Science Reviews 89, 94–107. doi:10.1016/j.quascirev.2014.02.001

King, G.E., Sanderson, D.C.W., Robinson, R.A.J., Finch, A.A., 2014b. Understanding processes of sedi‐ ment bleaching in glacial settings using a portable OSL reader. Boreas 43, 955–972. doi:10.1111/bor.12078

Kirkbride, M.P., Winkler, S., 2012. Correlation of Late Quaternary moraines: impact of climate varia‐ bility, glacier response, and chronological resolution. Quaternary Science Reviews 46, 1–29. doi:10.1016/j.quascirev.2012.04.002

Klasen, N., Fiebig, M., Preusser, F., Reitner, J.M., Radtke, U., 2007. Luminescence dating of proglacial sediments from the Eastern Alps. Quaternary International, From the Swiss Alps to the Crimean Mountains ‐ Alpine Quaternary stratigraphy in a European context 164–165, 21–32. doi:10.1016/j.quaint.2006.12.003

Klein, A.G., Isacks, B.L., 1999. Spectral mixture analysis of Landsat thematic mapper images applied to the detection of the transient snowline on tropical Andean glaciers. Global and Planetary Change 22, 139–154. doi:10.1016/S0921‐8181(99)00032‐6

Klingbjer, P., Brown, I.A., Holmlund, P., 2005. Identification of Climate Controls on the Dynamic Be‐ haviour of the Subarctic Glacier Salajekna, Northern Scandinavia. Geografiska Annaler: Series A, Phy‐ sical Geography 87, 215–229. doi:10.1111/j.0435‐3676.2005.00254.x

123

Klinge, M., 2001. Glazialgeomorphologische Untersuchungen im Mongolischen Altai als Beitrag zur jungquartären Landschafts‐ und Klimageschichte der Westmongolei, Aachener Geographische Arbei‐ ten. Aachen.

Klinge, M., Böhner, J., Lehmkuhl, F., 2003. Climate Pattern, Snow‐ and Timberlines in the Altai Moun‐ tains, Central Asia. Erdkunde 57, 296–308.

Klinge, M., Lehmkuhl, F., 2004. Pleistocene glaciations in southern and eastern Tibet, in: Ehlers, J., Gibbard, P.L. (Eds.), Developments in Quaternary Science. Elsevier, pp. 361 – 369.

Klinge, M., Lehmkuhl, F., 2003. Paleosols and aeolian mantles in southern and eastern Tibet ‐ Results and implications for late Quaternary climate change. Berliner Paläobiologische Abhandlungen 2, 62– 64.

Koch, J., 2014. Holocene Glacier Fluctuations, in: Singh, V.P., Singh, P., Haritashya, U.K. (Eds.), Ency‐ clopedia of Snow, Ice and Glaciers, Encyclopedia of Earth Sciences Series. Springer Netherlands, Dor‐ drecht, pp. 525–529.

Koch, J., 2009. Improving age estimates for late Holocene glacial landforms using dendrochronology – Some examples from Garibaldi Provincial Park, British Columbia. Quaternary Geochronology 4, 130– 139. doi:10.1016/j.quageo.2008.11.002

Kohn, M.J., 2014. Himalayan Metamorphism and Its Tectonic Implications. Annual Review of Earth and Planetary Sciences 42, 381–419. doi:10.1146/annurev‐earth‐060313‐055005

Kornfeld, D., Eckert, S., Appel, E., Ratschbacher, L., Sonntag, B.‐L., Pfänder, J.A., Ding, L., Liu, D., 2014. Cenozoic clockwise rotation of the Tengchong block, southeastern Tibetan Plateau: A paleomagnetic and geochronologic study. Tectonophysics 628, 105–122. doi:10.1016/j.tecto.2014.04.032

Korup, O., Montgomery, D.R., 2008. Tibetan plateau river incision inhibited by glacial stabilization of the Tsangpo gorge 455, 786–790.

Korup, O., Montgomery, D.R., Hewitt, K., 2010. Glacier and landslide feedbacks to topographic relief in the Himalayan syntaxes. Proceedings of the National Academy of Sciences 107, 5317–5322. doi:10.1073/pnas.0907531107

Kotlia, B.S., Ahmad, S.M., Zhao, J.‐X., Raza, W., Collerson, K.D., Joshi, L.M., Sanwal, J., 2012. Climatic fluctuations during the LIA and post‐LIA in the Kumaun Lesser Himalaya, India: Evidence from a 400 y old stalagmite record. Quaternary International, Late Quaternary morphodynamics in East Asia 263, 129–138. doi:10.1016/j.quaint.2012.01.025

Krinner, G., Genthon, C., Jouzel, J., 1997. GCM analysis of local influences on ice core δ signals. Geo‐ physical Research Letters 24, 2825–2828. doi:10.1029/97GL52891

Kuhle, M., 1986. Schneegrenzberechnung und typologische Klassifikation von Gletschern anhand spezifischer Reliefparameter. Petermanns Geographische Mitteilungen 86, 41–51.

Kulkarni, A.V., Bahuguna, I.M., Rathore, B.P., Singh, S.K., Randhawa, S.S., Sood, R.K., Dhar, S., 2007. Glacial retreat in Himalaya using Indian remote sensing satellite data. Current Science 92, 69–74.

124

Kutzbach, G., 1987. Monsoons. John Wiley & Sons, Chichester.

LaMarche, V.C., 1974. Paleoclimatic Inferences from Long Tree‐Ring Records Intersite comparison shows climatic anomalies that may be linked to features of the general circulation. Science 183, 1043–1048. doi:10.1126/science.183.4129.1043

LaMarche, V.C., Fritts, H.C., 1971. Tree Rings, Glacial Advance, And Climate in The Alps. Zeitschrift fur Gletscherkunde und Glazialgeologie 7, 125–131.

Lambert, F., Delmonte, B., Petit, J.R., Bigler, M., Kaufmann, P.R., Hutterli, M.A., Stocker, T.F., Ruth, U., Steffensen, J.P., Maggi, V., 2008. Dust‐climate couplings over the past 800,000 years from the EPICA Dome C ice core. Nature 452, 616–619. doi:10.1038/nature06763

Lamoureux, S.F., England, J.H., Sharp, M.J., Bush, A.B.G., 2001. A varve record of increased “Little Ice Age” rainfall associated with volcanic activity, Arctic Archipelago, Canada. The Holocene 11, 243–249. doi:10.1191/095968301668776315

Lang, K.A., Huntington, K.W., 2014. Antecedence of the Yarlung–Siang–Brahmaputra River, eastern Himalaya. Earth and Planetary Science Letters 397, 145–158. doi:10.1016/j.epsl.2014.04.026

Large, D.J., Spiro, B., Ferrat, M., Shopland, M., Kylander, M., Gallagher, K., Li, X., Shen, C., Possnert, G., Zhang, G., Darling, W.G., Weiss, D., 2009. The influence of climate, hydrology and permafrost on Holocene peat accumulation at 3500 m on the eastern Qinghai–Tibetan Plateau. Quaternary Science Reviews 28, 3303–3314. doi:10.1016/j.quascirev.2009.09.006

Larocque, S.J., Smith, D.J., 2005. “Little Ice Age” proxy glacier mass balance records reconstructed from tree rings in the Mt Waddington area, British Columbia Coast Mountains, Canada. The Holocene 15, 748–757. doi:10.1191/0959683605hl848rp

Lau, K.‐M., Kim, K.‐M., Yang, S., 2000. Dynamical and Boundary Forcing Characteristics of Regional Components of the Asian Summer Monsoon. Journal of Climate 13, 2461–2482. doi:10.1175/1520‐ 0442(2000)013<2461:DABFCO>2.0.CO;2

Lawrence, D.B., 1950. Estimating dates of recent glacier advances and recession rates by studying tree growth layers. Eos, Transactions American Geophysical Union 31, 243–248. doi:10.1029/TR031i002p00243

Lawrence Edwards, R., Chen, J.H., Wasserburg, G.J., 1987. 238U, 234U, 230Th, 232Th systematics and the precise measurement of time over the past 500,000 years. Earth and Planetary Science Letters 81, 175–192. doi:10.1016/0012‐821X(87)90154‐3

Lean, J., Rind, D., 1999. Evaluating sun‐climate relationships since the Little Ice Age. Journal of At‐ mospheric and Solar‐Terrestrial Physics 61, 25 – 36.

Leber, D., Holawe, F., Häusler, H., 1995. Climatic classification of the Tibet Autonomous Region using multivariate statistical methods. GeoJournal 37, 451–472.

Leech, M.L., Singh, S., Jain, A.K., Klemperer, S.L., Manickavasagam, R.M., 2005. The onset of India– Asia continental collision: Early, steep subduction required by the timing of UHP metamorphism in

125 the western Himalaya. Earth and Planetary Science Letters 234, 83–97. doi:10.1016/j.epsl.2005.02.038

Lee, H.‐Y., Chung, S.‐L., Wang, J.‐R., Wen, D.‐J., Lo, C.‐H., Yang, T.F., Zhang, Y., Xie, Y., Lee, T.‐Y., Wu, G., Ji, J., 2003. Miocene Jiali faulting and its implications for Tibetan tectonic evolution. Earth and Planetary Science Letters 205, 185–194. doi:10.1016/S0012‐821X(02)01040‐3

Lehmkuhl, F., 2012. Holocene glaciers in the Mongolian Altai: an example from the Turgen– Kharkhiraa Mountains. Journal of Asian Earth Sciences 52, 12–20. doi:10.1016/j.jseaes.2011.11.027

Lehmkuhl, F., 1998. Extent and spatial distribution of Pleistocene glaciations in eastern Tibet. Qua‐ ternary International 45‐46, 123 – 134. doi:10.1016/S1040‐6182(97)00010‐4

Lehmkuhl, F., 1997. Late pleistocene, Late‐glacial and Holocene glacier advances on the Tibetan Plat‐ eau. Quaternary International 38‐39, 77 – 83. doi:10.1016/S1040‐6182(96)00025‐0

Lehmkuhl, F., 1989. Geomorphologische Höhenstufen in den Alpen unter besonderer Berücksichti‐ gung des nivalen Formenschatzes, Göttinger Geographische Abhandlungen. Goltze, Göttingen.

Lehmkuhl, F., Haselein, F., 2000. Quaternary paleoenvironmental change on the Tibetan Plateau and adjacent areas (Western China and Western Mongolia). Quaternary International 65‐66, 121 – 145.

Lehmkuhl, F., Hilgers, A., Fries, S., Hülle, D., Schlütz, F., Shumilovskikh, L., Felauer, T., Protze, J., 2011. Holocene geomorphological processes and soil development as indicator for environmental change around Karakorum, Upper Orkhon Valley (Central Mongolia). Catena 87, 31–44. doi:10.1016/j.catena.2011.05.005

Lehmkuhl, F., Klinge, M., Lang, A., 2002. Late Quaternary glacier advances, lake level fluctuations and aeolian sedimentation in Southern Tibet. Zeitschrift für Geomorphologie Supplement 126, 183–218.

Lehmkuhl, F., Owen, L.A., Derbyshire, E., 1998. Late Quaternary glacial history of northeast Tibet. Quaternary Proceedings 6, 121–142.

Lehmkuhl, F., Schulte, P., Zhao, H., Hülle, D., Protze, J., Stauch, G., 2014. Timing and spatial distribu‐ tion of loess and loess‐like sediments in the mountain areas of the northeastern Tibetan Plateau. Catena, Loess and dust dynamics, environments, landforms, and pedogenesis: a tribute to Edward Derbyshire 117, 23–33. doi:10.1016/j.catena.2013.06.008

Leonard, K.C., Fountain, A.G., 2003. Map‐based methods for estimating glacier equilibrium‐line alti‐ tudes. Journal of Glaciology 49, 329–336.

Leuenberger, M., Siegenthaler, U., Langway, C., 1992. Carbon isotope composition of atmospheric CO2 during the last ice age from an Antarctic ice core. Nature 357, 488–490. doi:10.1038/357488a0

Leysinger Vieli, G.J.‐M.C., Gudmundsson, G.H., 2004. On estimating length fluctuations of glaciers caused by changes in climatic forcing. Journal of Geophysical Research: Earth Surface 109, F01007. doi:10.1029/2003JF000027

Liang, E.Y., Shao, X.M., Xu, Y., 2009. Tree‐ring evidence of recent abnormal warming on the southeast Tibetan Plateau. Theoretical and Applied Climatology 98, 9–18. doi:10.1007/s00704‐008‐0085‐6

126

Lian, O.B., Roberts, R.G., 2006. Dating the Quaternary: progress in luminescence dating of sediments. Quaternary Science Reviews, Dating the Quaternary: progress in luminescence dating of sediments Dating the Quaternary: progress in luminescence dating of sediments 25, 2449–2468. doi:10.1016/j.quascirev.2005.11.013

Libby, W.F., 1955. Radiocarbon Dating, Auflage: 2nd Revised edition. ed. University of Chicago Press, Chicago.

Lichtenecker, N., 1938. Die gegenwärtige und die eiszeitliche Schneegrenze in den Ostalpen, in: Ver‐ handlungen Der III Internationalen Quartär‐Conferenz, 1936. Vienna, pp. 141–147.

Li, C., Yanai, M., 1996. The Onset and Interannual Variability of the Asian Summer Monsoon in Rela‐ tion to Land‐Sea Thermal Contrast. Journal of Climate 9, 358 – 375.

Liebke, U., Appel, E., Ding, L., Neumann, U., Antolin, B., Xu, Q., 2010. Position of the Lhasa terrane prior to India–Asia collision derived from palaeomagnetic inclinations of 53 Ma old dykes of the Lin‐ zhou Basin: constraints on the age of collision and post‐collisional shortening within the Tibetan Plat‐ eau. Geophysical Journal International 182, 1199–1215. doi:10.1111/j.1365‐246X.2010.04698.x

Li, J., Zheng, B., Yan, X., Xie, Y., 1986. Glaciers of Xizang (Tibet). Science Press, Beijing.

Linderholm, H.W., Ou, T., Jeong, J.‐H., Folland, C.K., Gong, D., Liu, H., Liu, Y., Chen, D., 2011. Interan‐ nual teleconnections between the summer North Atlantic Oscillation and the East Asian summer monsoon. Journal of Geophysical Research: Atmospheres 116, D13107. doi:10.1029/2010JD015235

Lipp, J., Trimborn, P., Fritz, P., Moser, H., Becker, B., Frenzel, B., 1991. Stable isotopes in tree ring cellulose and climatic change. Tellus B 43, 322–330. doi:10.1034/j.1600‐0889.1991.t01‐2‐00005.x

Liu, B., Xu, M., Henderson, M., Qi, Y., 2005. Observed trends of precipitation amount, frequency, and intensity in China, 1960–2000. Journal of Geophysical Research: Atmospheres 110, D08103. doi:10.1029/2004JD004864

Liu, C., Shi, Y., Huang, M., Mi, D., 2008. Glaciers and their distribution in China, in: Shi, Y. (Ed.), Glaci‐ ers and Related Environments in China. Science Press, Beijing, pp. 16–94.

Liu, T., Ding, Z., 1998. Chinese Loess and the Paleomonsoon. Annual Review of Earth and Planetary Sciences 26, 111–145. doi:10.1146/annurev.earth.26.1.111

Liu, X., Dong, H., Yang, X., Herzschuh, U., Zhang, E., Stuut, J.‐B.W., Wang, Y., 2009. Late Holocene forcing of the Asian winter and summer monsoon as evidenced by proxy records from the northern Qinghai–Tibetan Plateau. Earth and Planetary Science Letters 280, 276–284. doi:10.1016/j.epsl.2009.01.041

Liu, X., Zeng, X., Leavitt, S.W., Wang, W., An, W., Xu, G., Sun, W., Wang, Y., Qin, D., Ren, J., 2013. A 400‐year tree‐ring δ18O chronology for the southeastern Tibetan Plateau: implications for inferring variations of the regional hydroclimate. Global and Planetary Change 104, 23–33. doi:10.1016/j.gloplacha.2013.02.005

127

Liu, Y., Siebel, W., Theye, T., Massonne, H.‐J., 2011. Isotopic and structural constraints on the late Miocene to Pliocene evolution of the Namche Barwa area, eastern Himalayan syntaxis, SE Tibet. Gondwana Research 19, 894–909.

Li, X., Cheng, G., Jin, H., Kang, E., Che, T., Jin, R., Wu, L., Nan, Z., Wang, J., Shen, Y., 2008. Cryospheric change in China. Global and Planetary Change 62, 210 – 218.

Lliboutry, L., 1964. Traité de glaciologie. Masson et Cie, Paris.

Loeppert, R.H., Suarez, D.L., 1996. Carbonate and Gypsum, in: Bigham, J.M. (Ed.), Methods of Soil Analysis: Part 3, Chemical Methods, SSSA Book Series. SSSA and ASoA, Madison, Wisconsin, USA.

Loibl, D., Lehmkuhl, F., 2014. Glaciers and equilibrium line altitudes of the eastern Nyainqêntanglha Range, SE Tibet. Journal of Maps 0, 1–14. doi:10.1080/17445647.2014.933451

Loibl, D., Lehmkuhl, F., 2013. High‐resolution geomorphological map of a low mountain range near Aachen, Germany. Journal of Maps 9, 245–253. doi:10.1080/17445647.2013.771291

Loibl, D., Lehmkuhl, F., Grießinger, J., 2014. Reconstructing glacier retreat since the Little Ice Age in SE Tibet by glacier mapping and equilibrium line altitude calculation. Geomorphology 214, 22–39. doi:10.1016/j.geomorph.2014.03.018

Louis, H., 1955. Schneegrenze und Schneegrenzbestimmung. Geographisches Taschenbuch 55, 414– 418.

Lowe, J.J., Walker, M.J.C., 1997. Reconstructing quaternary environments, 2nd ed. Longman, Harlow.

Lowell, T.V., 1995. The application of radiocarbon age estimates to the dating of glacial sequences: An example from the Miami sublobe, Ohio, U.S.A. Quaternary Science Reviews 14, 85–99. doi:10.1016/0277‐3791(94)00113‐P

Luckman, B.H., 2000. The Little Ice Age in the Canadian Rockies. Geomorphology 32, 357 – 384. doi:10.1016/S0169‐555X(99)00104‐X

Lu, H., Wang, X., Ma, H., Tan, H., Vandenberghe, J., Miao, X., Li, Z., Sun, Y., An, Z., Cao, G., 2004. The Plateau Monsoon variation during the past 130 kyr revealed by loess deposit at northeast Qinghai– Tibet (China). Global and Planetary Change, Human Dimensions and Natural Processes in Environ‐ mental Change 41, 207–214. doi:10.1016/j.gloplacha.2004.01.006

Lukas, S., Spencer, J.Q.G., Robinson, R.A.J., Benn, D.I., 2007. Problems associated with luminescence dating of Late Quaternary glacial sediments in the NW Scottish Highlands. Quaternary Geochronolo‐ gy 2, 243–248. doi:10.1016/j.quageo.2006.04.007

Luo, W., Wang, S., 2009. Transmission of δ 13C signals and its paleoclimatic implications in Liangfeng Cave system of Guizhou Province, SW China. Environmental Earth Sciences 59, 655–661. doi:10.1007/s12665‐009‐0062‐0

Luo, W., Wang, S., Xie, X., 2013. A comparative study on the stable isotopes from precipitation to speleothem in four caves of Guizhou, China. Chemie der Erde ‐ Geochemistry 73, 205–215. doi:10.1016/j.chemer.2012.05.002

128

Maher, B.A., MengYu, H., Roberts, H.M., Wintle, A.G., 2003. Holocene loess accumulation and soil development at the western edge of the Chinese Loess Plateau: implications for magnetic proxies of palaeorainfall. Quaternary Science Reviews 22, 445–451. doi:10.1016/S0277‐3791(02)00188‐9

Maisch, M., 2000. The longterm signal of climate change in the Swiss Alps: glacier retreat since the end of the Little Ice Age and future ice decay scenarios. Geografia Fisica e Dinamica Quaternaria 23, 139–150.

Maisch, M., 1987. Zur Gletschergeschichte des alpinen Spätglazials: Analyse und Interpretation von Schneegrenzdaten. Geographica Helvetica 42, 63–71.

Maisch, M., 1981. Glazialmorphologische und gletschergeschichtliche Untersuchungen im Gebiet zwischen Landwasser‐ und Albulatal (Kt. Graubünden, Schweiz), Physische Geographie. Zürich.

Mall, R.K., Gupta, A., Singh, R., Singh, R.S., Rathore, L.S., 2006. Water resources and climate change: An Indian perspective. Current Science 90, 1610–1626.

Manabe, S., Terpstra, T.B., 1974. The Effects of Mountains on the General Circulation of the Atmos‐ phere as Identified by Numerical Experiments. Journal of the Atmospheric Sciences 31, 3–42. doi:10.1175/1520‐0469(1974)031<0003:TEOMOT>2.0.CO;2

Mangerud, J., Andersen, S.T., Berglund, B.E., Donner, J.J., 1974. Quaternary stratigraphy of Norden, a proposal for terminology and classification. Boreas 3, 109–126. doi:10.1111/j.1502‐ 3885.1974.tb00669.x

Mann, M.E., 2002. Little Ice Age, in: MacCracken, M., Perry, J.S. (Eds.), Encyclopedia of Global Envi‐ ronmental Change, Volume 1, The Earth System: Physical and Chemical Dimensions of Global Envi‐ ronmental Change. Wiley, Chichester, pp. 504–509.

Mann, M.E., Jones, P.D., 2003. Global surface temperatures over the past two millennia. Geophys. Res. Lett. 30, 1820.

Mann, M.E., Zhang, Z., Rutherford, S., Bradley, R.S., Hughes, M.K., Shindell, D., Ammann, C., Faluvegi, G., Ni, F., 2009. Global Signatures and Dynamical Origins of the Little Ice Age and Medieval Climate Anomaly. Science 326, 1256–1260. doi:10.1126/science.1177303

Mason, B.H., Moore, C.B., 1982. Principles of geochemistry. Wiley.

Matthews, J.A., 1984. Limitations of 14C Dates from Buried Soils in Reconstructing Glacier Variations and Holocene Climate, in: Mörner, N.‐A., Karlén, W. (Eds.), Climatic Changes on a Yearly to Millennial Basis. Springer Netherlands, Dordrecht, pp. 281–290.

Matthews, J.A., 1977. Glacier and climatic fluctuations inferred from tree‐growth variations over the last 250 years, central southern Norway. Boreas 6, 1–24. doi:10.1111/j.1502‐3885.1977.tb00692.x

Matthews, J.A., Owen, G., 2010. Schmidt hammer exposure‐age dating: developing linear age‐ calibration curves using Holocene bedrock surfaces from the Jotunheimen–Jostedalsbreen regions of southern Norway. Boreas 39, 105–115. doi:10.1111/j.1502‐3885.2009.00107.x

129

Matthews, J.A., Shakesby, R.A., 1984. The status of the “Little Ice Age” in southern Norway: relative‐ age dating of Neoglacial moraines with Schmidt hammer and lichenometry. Boreas 13, 333–346. doi:10.1111/j.1502‐3885.1984.tb01128.x

Matthews, J.A., Winkler, S., 2011. Schmidt‐hammer exposure‐age dating (SHD): application to early Holocene moraines and a reappraisal of the reliability of terrestrial cosmogenic‐nuclide dating (TCND) at Austanbotnbreen, Jotunheimen, Norway. Boreas 40, 256 – 270.

Maussion, F., Scherer, D., Mölg, T., Collier, E., Curio, J., Finkelnburg, R., 2014. Precipitation seasonali‐ ty and variability over the Tibetan Plateau as resolved by the High Asia Reanalysis. Journal of Climate 27, 1910–1927. doi:10.1175/JCLI‐D‐13‐00282.1

Mayewski, P.A., Lyons, W.B., Spencer, M.J., Twickler, M.S., Buck, C.F., Whitlow, S., 1990. An ice‐core record of atmospheric response to anthropogenic sulphate and nitrate. Nature 346, 554–556. doi:10.1038/346554a0

Mayewski, P.A., Meeker, L.D., Morrison, M.C., Twickler, M.S., Whitlow, S.I., Ferland, K.K., Meese, D.A., Legrand, M.R., Steffensen, J.P., 1993. Greenland ice core “signal” characteristics: An expanded view of climate change. Journal of Geophysical Research: Atmospheres 98, 12839–12847. doi:10.1029/93JD01085

Mayewski, P.A., Rohling, E.E., Curt Stager, J., Karlén, W., Maasch, K.A., David Meeker, L., Meyerson, E.A., Gasse, F., van Kreveld, S., Holmgren, K., Lee‐Thorp, J., Rosqvist, G., Rack, F., Staubwasser, M., Schneider, R.R., Steig, E.J., 2004. Holocene climate variability. Quaternary Research 62, 243 – 255.

McCrea, J.M., 1950. On the Isotopic Chemistry of Carbonates and a Paleotemperature Scale. The Journal of Chemical Physics 18, 849–857. doi:10.1063/1.1747785

McDermott, F., 2004. Palaeo‐climate reconstruction from stable isotope variations in speleothems: a review. Quaternary Science Reviews, Isotopes in Quaternary Paleoenvironmental reconstruction 23, 901–918. doi:10.1016/j.quascirev.2003.06.021

Mckinzey, K.M., Orwin, J.F., Bradwell, T., 2004. Re‐Dating the Moraines at Skálafellsjökull and Heina‐ bergsjökull using different Lichenometric Methods: Implications for the Timing of the Icelandic Little Ice Age Maximum. Geografiska Annaler: Series A, Physical Geography 86, 319–335. doi:10.1111/j.0435‐3676.2004.00235.x

Meehl, G.A., 1994. Influence of the Land Surface in the Asian Summer Monsoon: External Conditions versus Internal Feedbacks. Journal of Climate 7, 1033–1049. doi:10.1175/1520‐ 0442(1994)007<1033:IOTLSI>2.0.CO;2

Meierding, T.C., 1982. Late Pleistocene glacial equilibrium‐line altitudes in the Colorado Front Range: a comparison of methods. Quaternary Research 18, 289–310. doi:10.1016/0033‐5894(82)90076‐X

Meier, M.F., 1984. Contribution of Small Glaciers to Global Sea Level. Science 226, 1418 – 1421.

Meier, M.F., Dyurgerov, M.B., Rick, U.K., O’Neel, S., Pfeffer, W.T., Anderson, R.S., Anderson, S.P., Glazovsky, A.F., 2007. Glaciers Dominate Eustatic Sea‐Level Rise in the 21st Century. Science 317, 1064–1067. doi:10.1126/science.1143906

130

Mellars, P., 2006. A new radiocarbon revolution and the dispersal of modern humans in Eurasia. Na‐ ture 439, 931–935. doi:10.1038/nature04521

Menzies, I.R., 1951. Some observations on the glaciology of the Ruwenzori Range. Journal of Glaciol‐ ogy 1, 511–512.

Mi, D., Xie, Z., 2001. Glacier Inventory of China, VI: The Ganga and Indus River Drainage Basin. Carto‐ graphic Publishing House, Xian.

Miehe, G., 1994. Der Himalaya, eine multizonale Gebirgsregion, in: Walter, H., Breckle, S.‐W. (Eds.), Spezielle Ökologie Der Gemäßigten Und Arktischen Zonen Euro‐Nordasiens. Fischer, Stuttgart, pp. 181–230.

Miller, G.H., Bradley, R.S., Andrews, J.T., 1975. The Glaciation Level and Lowest Equilibrium Line Alti‐ tude in the High Canadian Arctic: Maps and Climatic Interpretation. Arctic and Alpine Research 7, 155–168. doi:10.2307/1550318

Miller, G.H., Geirsdóttir, Á., Zhong, Y., Larsen, D.J., Otto‐Bliesner, B.L., Holland, M.M., Bailey, D.A., Refsnider, K.A., Lehman, S.J., Southon, J.R., Anderson, C., Björnsson, H., Thordarson, T., 2012. Abrupt onset of the Little Ice Age triggered by volcanism and sustained by sea‐ice/ocean feedbacks. Geo‐ physical Research Letters 39, L02708. doi:10.1029/2011GL050168

Milly, P.C.D., Wetherald, R.T., Dunne, K.A., Delworth, T.L., 2002. Increasing risk of great floods in a changing climate. Nature 415, 514–517. doi:10.1038/415514a

Mischke, S., Weynell, M., Zhang, C., Wiechert, U., 2013. Spatial variability of 14C reservoir effects in Tibetan Plateau lakes. Quaternary International, Proxies for Quaternary monsoon reconstruction on the Tibetan Plateau 313–314, 147–155. doi:10.1016/j.quaint.2013.01.030

Mischke, S., Zhang, C., 2010. Holocene cold events on the Tibetan Plateau. Global and Planetary Change 72, 155–163. doi:10.1016/j.gloplacha.2010.02.001

Mischke, S., Zhang, C., Börner, A., Herzschuh, U., 2010. Lateglacial and Holocene variation in aeolian sediment flux over the northeastern Tibetan Plateau recorded by laminated sediments of a saline meromictic lake. Journal of Quaternary Science 25, 162–177. doi:10.1002/jqs.1288

Moberg, A., 2013. Comparisons of simulated and observed Northern Hemisphere temperature varia‐ tions during the past millennium – selected lessons learned and problems encountered. Tellus B 65. doi:10.3402/tellusb.v65i0.19921

Mölg, T., Maussion, F., Scherer, D., 2013. Mid‐latitude westerlies as a driver of glacier variability in monsoonal High Asia. Nature Climate Change 4, 68–73. doi:10.1038/nclimate2055

Mölg, T., Maussion, F., Yang, W., Scherer, D., 2012. The footprint of Asian monsoon dynamics in the mass and energy balance of a Tibetan glacier. The Cryosphere 6, 1445–1461. doi:10.5194/tc‐6‐1445‐ 2012

Molnar, P., Boos, W.R., Battisti, D.S., 2010. Orographic controls on climate and paleoclimate of Asia: thermal and mechanical roles for the Tibetan Plateau. Annual Review of Earth and Planetary Sciences 38, 77–102. doi:10.1146/annurev‐earth‐040809‐152456 131

Molnar, P., England, P., Martinod, J., 1993. Mantle dynamics, uplift of the Tibetan Plateau, and the Indian Monsoon. Reviews of Geophysics 31, 357–396. doi:10.1029/93RG02030

Morgenroth, G., Kerscher, H., Kretschmer, W., Klein, M., Reichel, M., Tully, T., Wrzosok, I., 2000. Im‐ proved sample preparation techniques at the Erlangen AMS‐facility. Nuclear Instruments and Meth‐ ods in Physics Research Section B: Beam Interactions with Materials and Atoms, 8th International Conference on Accelerator Mass Spectrometry 172, 416–423. doi:10.1016/S0168‐583X(00)00277‐9

Morrill, C., Overpeck, J.T., Cole, J.E., Liu, K., Shen, C., Tang, L., 2006. Holocene variations in the Asian monsoon inferred from the geochemistry of lake sediments in central Tibet. Quaternary Research 65, 232 – 243. doi:10.1016/j.yqres.2005.02.014

Mukhopadhyay, B., Dasgupta, S., 2015. Earthquake swarms near eastern Himalayan Syntaxis along Jiali Fault in Tibet: A seismotectonic appraisal. Geoscience Frontiers. doi:10.1016/j.gsf.2014.12.009

Müller, F., Caflish, T., Müller, G., 1977. Instructions for the Compilation and Assemblage of Data for a World Glacier Inventory. Department of Geography, Swiss Federal Institute of Technology (ETH), Zürich.

Murakami, T., 1987. Effects of the Tibetan Plateau, in: Chang, C.‐P., Krishnamurti, T.N. (Eds.), Mon‐ soon Meteorology, Oxford Monographs on Geology and Geophysics. Oxford University Press, Oxford, pp. 235–270.

Murari, M.K., Owen, L.A., Dortch, J.M., Caffee, M.W., Dietsch, C., Fuchs, M., Haneberg, W.C., Sharma, M.C., Townsend‐Small, A., 2014. Timing and climatic drivers for glaciation across monsoon‐ influenced regions of the Himalayan–Tibetan orogen. Quaternary Science Reviews 88, 159–182. doi:10.1016/j.quascirev.2014.01.013

Murphy, M.A., Yin, A., Harrison, T.M., Dürr, S.B., Z, C., Ryerson, F.J., Kidd, W.S.F., X, W., X, Z., 1997. Did the Indo‐Asian collision alone create the Tibetan plateau? Geology 25, 719–722. doi:10.1130/0091‐7613(1997)025<0719:DTIACA>2.3.CO;2

Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved single‐aliquot regenerative‐dose protocol. Radiation Measurements 32, 57–73. doi:10.1016/S1350‐4487(99)00253‐ X

Muscheler, R., Joos, F., Beer, J., Müller, S.A., Vonmoos, M., Snowball, I., 2007. Solar activity during the last 1000 yr inferred from radionuclide records. Quaternary Science Reviews 26, 82–97. doi:10.1016/j.quascirev.2006.07.012

Neckel, N., Kropáček, J., Bolch, T., Hochschild, V., 2014. Glacier mass changes on the Tibetan Plateau 2003–2009 derived from ICESat laser altimetry measurements. Environmental Research Letters 9, 014009. doi:10.1088/1748‐9326/9/1/014009

Nesbitt, H.W., Markovics, G., price, R.C., 1980. Chemical processes affecting alkalis and alkaline earths during continental weathering. Geochimica et Cosmochimica Acta 44, 1659–1666. doi:10.1016/0016‐7037(80)90218‐5

132

Nesbitt, H.W., Young, G.M., 1984. Prediction of some weathering trends of plutonic and volcanic rocks based on thermodynamic and kinetic considerations. Geochimica et Cosmochimica Acta 48, 1523–1534. doi:10.1016/0016‐7037(84)90408‐3

Nesje, A., Kvamme, M., 1991. Holocene glacier and climate variations in western Norway: Evidence for early Holocene glacier demise and multiple Neoglacial events. Geology 19, 610–612. doi:10.1130/0091‐7613(1991)019<0610:HGACVI>2.3.CO;2

Nicholls, R.J., Cazenave, A., 2010. Sea‐Level Rise and Its Impact on Coastal Zones. Science 328, 1517– 1520. doi:10.1126/science.1185782

Nicholson, L., Benn, D.I., 2006. Calculating ice melt beneath a debris layer using meteorological data. Journal of Glaciology 52, 463–470. doi:10.3189/172756506781828584

Oerlemans, J., 2001. Glaciers and Climate Change. Balkema, Leiden.

Ohmura, A., Kasser, P., Funk, M., 1992. Climate at the equilibrium line of glaciers. Journal of Glaciolo‐ gy 38, 397–411.

Opitz, S., Wünnemann, B., Aichner, B., Dietze, E., Hartmann, K., Herzschuh, U., IJmker, J., Lehmkuhl, F., Li, S., Mischke, S., Plotzki, A., Stauch, G., Diekmann, B., 2012. Late Glacial and Holocene develop‐ ment of Lake Donggi Cona, north‐eastern Tibetan Plateau, inferred from sedimentological analysis. Palaeogeography, Palaeoclimatology, Palaeoecology 337–338, 159–176. doi:10.1016/j.palaeo.2012.04.013

Osmaston, H., 2005. Estimates of glacier equilibrium line altitudes by the Area×Altitude, the Ar‐ ea×Altitude Balance Ratio and the Area×Altitude Balance Index methods and their validation. Qua‐ ternary International 138–139, 22–31. doi:10.1016/j.quaint.2005.02.004

Osmaston, H.A., 1989. Glaciers, glaciations and equilibrium line altitudes on the Rwenzori, in: Ma‐ haney, W.C. (Ed.), Quaternary and Environmental Research on East African Mountains. Balkema, Rotterdam, pp. 31–104.

Østrem, G., 1975. ERTS data in glaciology – an effort to monitor glacier mass balance from satellite imagery. Journal of Glaciology 15, 403–415.

Østrem, G., 1973. The Transient Snowline and Glacier Mass Balance in Southern British Columbia and Alberta, Canada. Geografiska Annaler. Series A, Physical Geography 55, 93–106. doi:10.2307/520877

Østrem, G., Stanley, A., 1969. Glacier mass‐balance measurements: a manual for field and office work. Canadian Department of energy, mines and resources.

Owen, L.A., 2009. Latest Pleistocene and Holocene glacier fluctuations in the Himalaya and Tibet. Quaternary Science Reviews, Holocene and Latest Pleistocene Alpine Glacier Fluctuations: A Global Perspective 28, 2150–2164. doi:10.1016/j.quascirev.2008.10.020

Owen, L.A., Dortch, J.M., 2014. Nature and timing of Quaternary glaciation in the Himalayan–Tibetan orogen. Quaternary Science Reviews 88, 14–54. doi:10.1016/j.quascirev.2013.11.016

133

Özer, M., Orhan, M., Işik, N.S., 2010. Effect of Particle Optical Properties on Size Distribution of Soils Obtained by Laser Diffraction. Environmental & Engineering Geoscience 16, 163–173. doi:10.2113/gseegeosci.16.2.163

Palais, J.M., Germani, M.S., Zielinski, G.A., 1992. Inter‐hemispheric Transport of Volcanic Ash from a 1259 A.D. Volcanic Eruption to the Greenland and Antarctic Ice Sheets. Geophysical Research Letters 19, 801–804. doi:10.1029/92GL00240

Paul, F., 2010. Towards a global glacier inventory from satellite data ‐ recent efforts and challenges to complete the global glacier inventory with modern geoinformatic techniques. Geographica Helvetica 65, 103–105.

Paul, F., Barry, R.G., Cogley, J.G., Frey, H., Haeberli, W., Ohmura, A., Ommanney, C.S.L., Raup, B., Ri‐ vera, A., Zemp, M., 2010. Guidelines for the compilation of glacier inventory data from digital sources [WWW Document]. URL http://globglacier.ch/docs/guidelines_inventory.pdf

Pelto, M., 2011. Utility of late summer transient snowline migration rate on Taku Glacier, Alaska. The Cryosphere 5, 1127–1133. doi:10.5194/tc‐5‐1127‐2011

Penck, A., Brückner, E., 1909. Die Alpen im Eiszeitalter. Tauchnitz, Leipzig.

Peng, B., Bao, H., Pu, L., 2000. Geo‐Ecology of Mts Namjabarwa Region, in: Zheng, D., Zhang, Q., Wu, S. (Eds.), Mountain Geoecology and Sustainable Development of the Tibetan Plateau. Kluwer Aca‐ demic Publishers, Dordrecht, pp. 265–282.

Petit, J.R., Mourner, L., Jouzel, J., Korotkevich, Y.S., Kotlyakov, V.I., Lorius, C., 1990. Palaeoclimatolog‐ ical and chronological implications of the Vostok core dust record. Nature 343, 56–58. doi:10.1038/343056a0

Pfeffer, W.T., Arendt, A.A., Bliss, A., Bolch, T., Cogley, J.G., Gardner, A.S., Hagen, J.‐O., Hock, R., Kaser, G., Kienholz, C., Miles, E.S., Moholdt, G., Mölg, N., Paul, F., Radić, V., Rastner, P., Raup, B.H., Rich, J., Sharp, M.J., 2014. The Randolph Glacier Inventory: a globally complete inventory of glaciers. Journal of Glaciology 60, 537–552. doi:10.3189/2014JoG13J176

Pfeffer, W.T., Harper, J.T., O’Neel, S., 2008. Kinematic Constraints on Glacier Contributions to 21st‐ Century Sea‐Level Rise. Science 321, 1340–1343. doi:10.1126/science.1159099

Porter, S.C., 2001. Chinese loess record of monsoon climate during the last glacial–interglacial cycle. Earth‐Science Reviews, Recent research on loess and palaeosols, pure and applied 54, 115–128. doi:10.1016/S0012‐8252(01)00043‐5

Porter, S.C., 1975. Equilibrium‐line altitudes of late Quaternary glaciers in the Southern Alps, New Zealand. Quaternary Research 5, 27–47. doi:10.1016/0033‐5894(75)90047‐2

Porter, S.C., Carson III, R.J., 1971. Problems of interpreting radiocarbon dates from dead‐ice terrain, with an example from the Puget Lowland of Washington. Quaternary Research 1, 410–414. doi:10.1016/0033‐5894(71)90074‐3

Prell, W.L., Kutzbach, J.E., 1992. Sensitivity of the Indian monsoon to forcing parameters and implica‐ tions for its evolution. Nature 360, 647. 134

Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contributions to dose rates for luminescence and ESR dating: Large depths and long‐term time variations. Radiation Measurements 23, 497–500. doi:10.1016/1350‐4487(94)90086‐8

Preusser, F., Andersen, B.G., Denton, G.H., Schlüchter, C., 2005. Luminescence chronology of Late Pleistocene glacial deposits in North Westland, New Zealand. Quaternary Science Reviews 24, 2207– 2227. doi:10.1016/j.quascirev.2004.12.005

Preusser, F., Degering, D., Fuchs, M., Hilgers, A., Kadereit, A., Klasen, N., Krbetschel, M., Richter, D., Spencer, J.Q.G., 2008. Luminescence dating: basics, methods and applications. Eiszeitalter und Gegenwart Quaternary Science Journal 57, 95–149. doi:10.3285/eg.57.1‐2.5

Preusser, F., Ramseyer, K., Schlüchter, C., 2006. Characterisation of low OSL intensity quartz from the New Zealand Alps. Radiation Measurements, Proceedings of the 11th International Conference on Luminescence and Electron‐Spin Resonance Dating (LED 2005) Proceedings of the 11th International Conference on Luminescence and Electron‐Spin Resonance Dating (LED 2005) 41, 871–877. doi:10.1016/j.radmeas.2006.04.019

Pye, K., Blott, S.J., 2004. Particle size analysis of sediments, soils and related particulate materials for forensic purposes using laser granulometry. Forensic Science International 144, 19–27. doi:10.1016/j.forsciint.2004.02.028

Qin, D., Hou, S., Zhang, D., Ren, J., Kang, S., Mayewski, P.A., Wake, C.P., 2002. Preliminary results from the chemical records of an 80.4 m ice core recovered from East Rongbuk Glacier, Qomolangma (Mount Everest), Himalaya. Annals of Glaciology 35, 278–284. doi:10.3189/172756402781816799

Qiu, J., 2008. China: The third pole. Nature News 454, 393–396. doi:10.1038/454393a

Quanru, G., Guitang, P., Zheng, L., Chen, Z., Fisher, R.D., Sun, Z., Ou, C., Dong, H., Wang, X., Li, S., Lou, X., Fu, H., 2006. The Eastern Himalayan syntaxis: major tectonic domains, ophiolitic mélanges and geologic evolution. Journal of Asian Earth Sciences 27, 265–285. doi:10.1016/j.jseaes.2005.03.009

Quincey, D.J., Richardson, S.D., Luckman, A., Lucas, R.M., Reynolds, J.M., Hambrey, M.J., Glasser, N.F., 2007. Early recognition of glacial lake hazards in the Himalaya using remote sensing datasets. Global and Planetary Change, Climate Change Impacts on Mountain Glaciers and Permafrost 56, 137– 152. doi:10.1016/j.gloplacha.2006.07.013

Rabatel, A., Jomelli, V., Naveau, P., Francou, B., Grancher, D., 2005. Dating of Little Ice Age glacier fluctuations in the tropical Andes: Charquini glaciers, Bolivia, 16°S. Comptes Rendus Geoscience 337, 1311–1322. doi:10.1016/j.crte.2005.07.009

Racoviteanu, A., Williams, M.W., 2012. Decision Tree and Texture Analysis for Mapping Debris‐ Covered Glaciers in the Kangchenjunga Area, Eastern Himalaya. Remote Sensing 4, 3078–3109. doi:10.3390/rs4103078

Ram, M., Koenig, G., 1997. Continuous dust concentration profile of pre‐Holocene ice from the Greenland Ice Sheet Project 2 ice core: Dust stadials, interstadials, and the Eemian. Journal of Geo‐ physical Research: Oceans 102, 26641–26648. doi:10.1029/96JC03548

135

Rampino, M.R., Self, S., 1982. Historic eruptions of Tambora (1815), Krakatau (1883), and Agung (1963), their stratospheric aerosols, and climatic impact. Quaternary Research 18, 127–143. doi:10.1016/0033‐5894(82)90065‐5

Raper, S.C.B., Braithwaite, R.J., 2009. Glacier volume response time and its links to climate and to‐ pography based on a conceptual model of glacier hypsometry. The Cryosphere 3, 183–194. doi:10.5194/tc‐3‐183‐2009

Rasmussen, L.A., 2013. Meteorological controls on glacier mass balance in High Asia. Annals of Glaci‐ ology 54, 352–359. doi:10.3189/2013AoG63A353

Rasmussen, L.A., Andreassen, L.M., Baumann, S., Conway, H., 2010. “Little Ice Age” precipitation in Jotunheimen, southern Norway. The Holocene 20, 1039–1045. doi:10.1177/0959683610369510

Rathjens, C., 1982. Geographie des Hochgebirges ‐ Der Naturraum. Vieweg+Teubner, Wiesbaden.

Raup, B., Kääb, A., Kargel, J.S., Bishop, M.P., Hamilton, G., Lee, E., Paul, F., Rau, F., Soltesz, D., Khalsa, S.J.S., Beedle, M., Helm, C., 2007a. Remote sensing and GIS technology in the Global Land Ice Meas‐ urements from Space (GLIMS) Project. Computers und Geosciences 33, 104 – 125.

Raup, B., Racoviteanu, A., Khalsa, S.J.S., Helm, C., Armstrong, R., Arnaud, Y., 2007b. The GLIMS geo‐ spatial glacier database: A new tool for studying glacier change. Climate Change Impacts on Moun‐ tain Glaciers and Permafrost 56, 101–110.

Raymo, M.E., Ruddiman, W.F., 1992. Tectonic forcing of late Cenozoic climate. Nature 359, 117–122. doi:10.1038/359117a0

Reger, R.D., Sturmann, A.G., Begét, J.E., 1993. Dating Holocene moraines of Black Rapids Glacier, Delta River valley, central Alaska Range. Short Notes on Alaskan Geology 113, 51–59.

Reimann, T., Tsukamoto, S., Naumann, M., Frechen, M., 2011. The potential of using K‐rich feldspars for optical dating of young coastal sediments – A test case from Darss‐Zingst peninsula (southern Baltic Sea coast). Quaternary Geochronology 6, 207–222. doi:10.1016/j.quageo.2010.10.001

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., 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., Plicht, J. van der, Weyhenmeyer, C.E., 2011. IntCal09 and Marine09 Radiocarbon Age Calibration Curves, 0‐50,000 Years cal BP. Radiocarbon 51, 1111– 1150. doi:10.2458/azu_js_rc.51.3569

Replumaz, A., Capitanio, F.A., Guillot, S., Negredo, A.M., Villaseñor, A., 2014. The coupling of Indian subduction and Asian continental tectonics. Gondwana Research 26, 608–626. doi:10.1016/j.gr.2014.04.003

Rhodes, E.J., 2011. Optically Stimulated Luminescence Dating of Sediments over the Past 200,000 Years. Annual Review of Earth and Planetary Sciences 39, 461–488. doi:10.1146/annurev‐earth‐ 040610‐133425

136

Rhodes, E.J., Pownall, L., 1994. Zeroing of the OSL signal in quartz from young glaciofluvial sediments. Radiation Measurements 23, 581–585. doi:10.1016/1350‐4487(94)90103‐1

Richards, B.W.M., 2000. Luminescence dating of Quaternary sediments in the Himalaya and High Asia: a practical guide to its use and limitations for constraining the timing of glaciation. Quaternary International 65‐66, 49 – 61.

Richardson, C., Holmlund, P., 1996. Glacial cirque formation in northern Scandinavia. Annals of Glaci‐ ology 22, 102–106.

Richardson, S.D., Reynolds, J.M., 2000. An overview of glacial hazards in the Himalayas. Quaternary International 65–66, 31–47. doi:10.1016/S1040‐6182(99)00035‐X

Richmond, G.M., 1965. Glaciation of the southern Rocky Mountains, in: Wright, H.E., Frey, D.G. (Eds.), The Quaternary of the United States. Princeton University Press, Princeton, pp. 217–230.

Rinn, F., 2003. TSAP‐Win. Time series analysis and presentation for dendrochronology and related applications ‐ Version 0.53 for Microsoft Windows. User Reference.

Robock, A., 1979. The “Little Ice Age”: Northern Hemisphere Average Observations and Model Calcu‐ lations. Science 206, 1402–1404. doi:10.1126/science.206.4425.1402

Rodbell, D.T., 1992. Lichenometric and Radiocarbon Dating of Holocene Glaciation, Cordillera Blanca, Perú. The Holocene 2, 19–29. doi:10.1177/095968369200200103

Roden, J.S., Lin, G., Ehleringer, J.R., 2000. A mechanistic model for interpretation of hydrogen and oxygen isotope ratios in tree‐ring cellulose. Geochimica et Cosmochimica Acta 64, 21–35. doi:10.1016/S0016‐7037(99)00195‐7

Rohling, E.J., Liu, Q.S., Roberts, A.P., Stanford, J.D., Rasmussen, S.O., Langen, P.L., Siddall, M., 2009. Controls on the East Asian monsoon during the last glacial cycle, based on comparison between Hulu Cave and polar ice‐core records. Quaternary Science Reviews 28, 3291–3302. doi:10.1016/j.quascirev.2009.09.007

Rother, H., Lehmkuhl, F., Fink, D., Nottebaum, V., 2014. Surface exposure dating reveals MIS‐3 glacial maximum in the Khangai Mountains of Mongolia. Quaternary Research 82, 297–308. doi:10.1016/j.yqres.2014.04.006

Rott, H., 2009. Advances in interferometric synthetic aperture radar (InSAR) in earth system science. Progress in Physical Geography 33, 769–791. doi:10.1177/0309133309350263

Rowan, A.V., Brocklehurst, S.H., Schultz, D.M., Plummer, M.A., Anderson, L.S., Glasser, N.F., 2014. Late Quaternary glacier sensitivity to temperature and precipitation distribution in the Southern Alps of New Zealand. Journal of Geophysical Research: Earth Surface 119, 2013JF003009. doi:10.1002/2013JF003009

Rowley, D.B., 1996. Age of initiation of collision between India and Asia: A review of stratigraphic data. Earth and Planetary Science Letters 145, 1–13. doi:10.1016/S0012‐821X(96)00201‐4

137

Royden, L.H., Burchfiel, B.C., Hilst, R.D. van der, 2008. The Geological Evolution of the Tibetan Plat‐ eau. Science 321, 1054–1058. doi:10.1126/science.1155371

Savoskul, O.S., 1997. Modern and Little Ice Age glaciers in humid and arid areas of the Tien Shan, central Asia: two different patterns of fluctuation. Annals of Glaciology 24, 142–147.

Schaefer, J.M., Denton, G.H., Kaplan, M., Putnam, A., Finkel, R.C., Barrell, D.J.A., Andersen, B.G., Schwartz, R., Mackintosh, A., Chinn, T., Schlüchter, C., 2009. High‐Frequency Holocene Glacier Fluc‐ tuations in New Zealand Differ from the Northern Signature. Science 324, 622–625. doi:10.1126/science.1169312

Scherler, D., Bookhagen, B., Strecker, M.R., 2011a. Spatially variable response of Himalayan glaciers to climate change affected by debris cover. Nature Geoscience 4, 156–159. doi:10.1038/ngeo1068

Scherler, D., Bookhagen, B., Strecker, M.R., 2011b. Hillslope‐glacier coupling: the interplay of topog‐ raphy and glacial dynamics in High Asia. J. Geophys. Res. 116, F02019.

Schimmelpfennig, I., Schaefer, J.M., Akçar, N., Koffman, T., Ivy‐Ochs, S., Schwartz, R., Finkel, R.C., Zimmerman, S., Schlüchter, C., 2014. A chronology of Holocene and Little Ice Age glacier culminations of the Steingletscher, Central Alps, Switzerland, based on high‐sensitivity beryllium‐10 moraine da‐ ting. Earth and Planetary Science Letters 393, 220–230. doi:10.1016/j.epsl.2014.02.046

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. doi:10.1016/j.quascirev.2009.01.009

Schrijver, C.J., Livingston, W.C., Woods, T.N., Mewaldt, R.A., 2011. The minimal solar activity in 2008– 2009 and its implications for long‐term climate modeling. Geophysical Research Letters 38, L06701. doi:10.1029/2011GL046658

Seward, D., Burg, J.‐P., 2008. Growth of the Namche Barwa syntaxis and associated evolution of the Tsangpo Gorge: constraints from structural and thermochronological data. Tectonophysics 451, 282– 289. doi:10.1016/j.tecto.2007.11.057

Shakesby, R.A., Matthews, J.A., Owen, G., 2006. The Schmidt hammer as a relative‐age dating tool and its potential for calibrated‐age dating in Holocene glaciated environments. Quaternary Science Reviews 25, 2846 – 2867.

Shang, Y., Yang, Z., Li, L., Liu, D., Liao, Q., Wang, Y., 2003. A super‐large landslide in Tibet in 2000: background, occurrence, disaster, and origin. Geomorphology 54, 225–243. doi:10.1016/S0169‐ 555X(02)00358‐6

Shao, X., Wang, Y., Cheng, H., Kong, X., Wu, J., Edwards, R.L., 2006. Long‐term trend and abrupt events of the Holocene Asian monsoon inferred from a stalagmite δ10O record from Shennongjia in Central China. Chinese Science Bulletin 51, 221–228. doi:10.1007/s11434‐005‐0882‐6

Shapiro, A.I., Schmutz, W., Rozanov, E., Schoell, M., Haberreiter, M., Shapiro, A.V., Nyeki, S., 2011. A new approach to the long‐term reconstruction of the solar irradiance leads to large historical solar forcing. Astronomy & Astrophysics 529, A67. doi:10.1051/0004‐6361/201016173

138

Sheldon, N.D., Tabor, N.J., 2009. Quantitative paleoenvironmental and paleoclimatic reconstruction using paleosols. Earth‐Science Reviews 95, 1–52. doi:10.1016/j.earscirev.2009.03.004

Shen, C., Wang, W.‐C., Hao, Z., Gong, W., 2007. Exceptional drought events over eastern China during the last five centuries. Climatic Change 85, 453–471. doi:10.1007/s10584‐007‐9283‐y

Shi, C., Daux, V., Zhang, Q.‐B., Risi, C., Hou, S.‐G., Stievenard, M., Pierre, M., Li, Z., Masson‐Delmotte, V., 2012. Reconstruction of southeast Tibetan Plateau summer climate using tree ring δ18O: mois‐ ture variability over the past two centuries. Climate of the Past 8, 205–213. doi:10.5194/cp‐8‐205‐ 2012

Shi, F., Yang, B., Mairesse, A., Gunten, L. von, Li, J., Bruning, A., Yang, F., Xiao, X., 2013. Northern Hemisphere temperature reconstruction during the last millennium using multiple annual proxies. Clim Res 56, 231–244. doi:10.3354/cr01156

Shi, Y., 1992. Glaciers and glacial geomorphology in China. Zeitschrift für Geomorphologie Supple‐ ment 86, 51–63.

Shi, Y., Huang, M., Yao, Y., He, Y., 2008. Glaciers and related environments in China. Science Press, Beijing.

Shi, Y., Li, J., 1981. Glaciological Research of the Qinghai‐Xizang Plateau in China, in: Geological and Ecological Studies of Qinghai‐Xizang Plateau. Proceedings of Symposium on Qinghai‐Xizang (Tibet) Plateau (Beijing, China). Science Press, Beijing, pp. 1589–1598.

Shi, Y., Liu, C., Kang, E., 2010. The Glacier Inventory of China. Annals of Glaciology 50, 1–4. doi:10.3189/172756410790595831

Shi, Y., Liu, C., Wang, Z., 2006. Catalogue of Glaciers in China. Shanghai Press of Science & Technolo‐ gy, Shanghai.

Shukla, J., Paolino, D.A., 1983. The Southern Oscillation and Long‐Range Forecasting of the Summer Monsoon Rainfall over India. Monthly Weather Review 111, 1830–1837. doi:10.1175/1520‐ 0493(1983)111<1830:TSOALR>2.0.CO;2

Siegenthaler, U., Oeschger, H., 1980. Correlation of 18O in precipitation with temperature and alti‐ tude. Nature 285, 314–317. doi:10.1038/285314a0

Sigafoos, R.S., Hendricks, E.L., 1969. The time interval between stabilization of alpine glacial deposits and establishment of tree seedlings, in: Geological Survey Research 1969, U.S. Geological Survey Professional Paper. U.S. Government Printing Office, Washington, pp. 659–663.

Sigafoos, R.S., Hendricks, E.L., 1961. Botanical Evidence of the Modern History of Nisqually Glacier, Washington (USGS Numbered Series No. 387‐A), Professional Paper. U.S. Geological Survey.

Singh, J.S., Singh, S.P., 1987. Forest vegetation of the Himalaya. The Botanical Review 53, 80–192. doi:10.1007/BF02858183

139

Sinha, A., Berkelhammer, M., Stott, L., Mudelsee, M., Cheng, H., Biswas, J., 2011. The leading mode of Indian Summer Monsoon precipitation variability during the last millennium. Geophysical Re‐ search Letters 38, L15703. doi:10.1029/2011GL047713

Sissons, J.B., 1967. Glacial stages and radiocarbon dates in Scotland. Scottish Journal of Geology 3, 375–381. doi:10.1144/sjg03030375

Sissons, J.B., Sutherland, D.G., 1976. Climatic inferences from former glaciers in the south‐east Grampian Highlands. Journal of Glaciology 17, 325–346.

Smiraglia, C., 1997. Holocene variations of the Yanzigou Glacier (Gongga Shan Massif, Da Xueshan, China). Geografica Fisica e Dinamica Quaternaria 20, 339–351.

Solanki, S.K., Usoskin, I.G., Kromer, B., Schüssler, M., Beer, J., 2004. Unusual activity of the Sun during recent decades compared to the previous 11,000 years. Nature 431, 1084–1087. doi:10.1038/nature02995

Solomina, O.N., Bradley, R.S., Hodgson, D.A., Ivy‐Ochs, S., Jomelli, V., Mackintosh, A.N., Nesje, A., Owen, L.A., Wanner, H., Wiles, G.C., Young, N.E., 2015. Holocene glacier fluctuations. Quaternary Science Reviews 111, 9–34. doi:10.1016/j.quascirev.2014.11.018

Sorg, A., Bolch, T., Stoffel, M., Solomina, O., Beniston, M., 2012. Climate change impacts on glaciers and runoff in Tien Shan (Central Asia). Nature Climate Change 2, 725–731. doi:10.1038/nclimate1592

Spectro, 2007. Analysis of trace elements in geological materials, soils and sludges prepared as pressed pellets.

Spencer, J.Q., Owen, L.A., 2004. Optically stimulated luminescence dating of Late Quaternary glacio‐ genic sediments in the upper Hunza valley: validating the timing of glaciation and assessing dating methods. Quaternary Science Reviews 23, 175–191. doi:10.1016/S0277‐3791(03)00220‐8

Stauch, G., IJmker, J., Pötsch, S., Zhao, H., Hilgers, A., Diekmann, B., Dietze, E., Hartmann, K., Opitz, S., Wünnemann, B., Lehmkuhl, F., 2012. Aeolian sediments on the north‐eastern Tibetan Plateau. Qua‐ ternary Science Reviews 57, 71–84. doi:10.1016/j.quascirev.2012.10.001

Steinhilber, F., Beer, J., Fröhlich, C., 2009. Total solar irradiance during the Holocene. Geophysical Research Letters 36, L19704. doi:10.1029/2009GL040142

Stokes, M.A., Smiley, T.L., 1968. An introduction to tree‐ring dating. University of Chicago Press, Chi‐ cago.

Stothers, R.B., 1984. The Great Tambora Eruption in 1815 and Its Aftermath. Science 224, 1191– 1198. doi:10.1126/science.224.4654.1191

Strahler, A., Strahler, A.H., 2004. Physical Geography: Science and Systems of the Human Environ‐ ment. John Wiley & Sons, New York.

Stroup, J.S., Kelly, M.A., Lowell, T.V., Applegate, P.J., Howley, J.A., 2014. Late Holocene fluctuations of Qori Kalis outlet glacier, Quelccaya Ice Cap, Peruvian Andes. Geology 42, 347–350. doi:10.1130/G35245.1

140

Stuiver, M., Polach, H.A., 1977. Discussion: reporting of 14C data. Radiocarbon 19, 355–363.

Stuiver, M., Quay, P.D., 1980. Changes in Atmospheric Carbon‐14 Attributed to a Variable Sun. Sci‐ ence 207, 11–19. doi:10.1126/science.207.4426.11

Suess, S.T., 1979. The solar wind during the Maunder minimum. Planetary and Space Science 27, 1001–1013. doi:10.1016/0032‐0633(79)90030‐8

Summerfield, M.A., 1991. Global Geomorphology. Longman, Harlow.

Su, Z., Shi, Y., 2002. Response of monsoonal temperate glaciers to global warming since the Little Ice Age. Quaternary International 97‐98, 123 – 131.

Symkatz‐Kloss, W., Roy, P.D., 2010. Evaporite mineralogy and major element geochemistry as tools for palaeoclimatic investigations in arid regions: A synthesis. Boletín de la Sociedad Geológica Mexi‐ cana 62, 379–390.

Taft, L., Wiechert, U., Riedel, F., Weynell, M., Zhang, H., 2012. Sub‐seasonal oxygen and carbon iso‐ tope 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. doi:10.1016/j.quascirev.2011.12.006

Taft, L., Wiechert, U., Zhang, H., Lei, G., Mischke, S., Plessen, B., Weynell, M., Winkler, A., Riedel, F., 2013. Oxygen and carbon isotope patterns archived in shells of the aquatic gastropod Radix: Hydro‐ logic and climatic signals across the Tibetan Plateau in sub‐monthly resolution. Quaternary Interna‐ tional, The Baikal‐Hokkaido Archaeology Project: Environmental archives, proxies and reconstruction approaches 290–291, 282–298. doi:10.1016/j.quaint.2012.10.031

Tao, S., Ding, Y., 1981. Observational Evidence of the Influence of the Qinghai‐Xizang (Tibet) Plateau on the Occurrence of Heavy Rain and Severe Convective Storms in China. Bulletin of the American Meteorological Society 62, 23–30. doi:10.1175/1520‐0477(1981)062<0023:OEOTIO>2.0.CO;2

Tapponnier, P., Zhiqin, X., Roger, F., Meyer, B., Arnaud, N., Wittlinger, G., Jingsui, Y., 2001. Oblique Stepwise Rise and Growth of the Tibet Plateau. Science 294, 1671–1677. doi:10.1126/science.105978

Taylor, K.C., Hammer, C.U., Alley, R.B., Clausen, H.B., Dahl‐Jensen, D., Gow, A.J., Gundestrup, N.S., Kipfstuh, J., Moore, J.C., Waddington, E.D., 1993. Electrical conductivity measurements from the GISP2 and GRIP Greenland ice cores. Nature 366, 549–552. doi:10.1038/366549a0

Thompson, L.G., Davis, M.E., Mosley‐Thompson, E., Lin, P.‐N., Henderson, K.A., Mashiotta, T.A., 2005. Tropical ice core records: evidence for asynchronous glaciation on Milankovitch timescales. Journal of Quaternary Science 20, 723 – 733. doi:10.1002/jqs.972

Thompson, L.G., Mosley‐Thompson, E., Davis, M.E., Bolzan, J.F., Dai, J., Klein, L., Yao, T., Wu, X., Xie, Z., Gundestrup, N., 1989. Holocene—Late Pleistocene Climatic Ice Core Records from Qinghai‐Tibetan Plateau. Science 246, 474 – 477.

Thompson, L.G., Mosley‐Thompson, E., Davis, M.E., Lin, P.N., Dai, J., Bolzan, J.F., Yao, T., 1995. A 1000 year climatic ice‐core record from the Guliya ice cap, China: its relationship to global climate variabil‐ ity. Annals of Glaciology 21, 175–181. 141

Thompson, L.G., Yao, T., Davis, M.E., Mosley‐Thompson, E., Mashiotta, T.A., Lin, P.‐N., Mikhalenko, V.N., Zagorodnov, V.S., 2006. Holocene climate variability archived in the Puruogangri ice cap on the central Tibetan Plateau. Annals of Glaciology 43, 61–69. doi:10.3189/172756406781812357

Thompson, L.G., Yao, T., Mosley‐Thompson, E., Davis, M.E., Henderson, K.A., Lin, P.‐N., 2000. A high‐ resolution millennial record of the South Asian Monsoon from Himalayan ice cores. Science 289, 1916–1919. doi:10.1126/science.289.5486.1916

Thompson, P., Schwarcz, H.P., Ford, D.C., 1974. Continental Pleistocene Climatic Variations from Spe‐ leothem Age and Isotopic Data. Science 184, 893–895. doi:10.1126/science.184.4139.893

Thrasher, I.M., Mauz, B., Chiverrell, R.C., Lang, A., 2009. Luminescence dating of glaciofluvial depos‐ its: A review. Earth‐Science Reviews 97, 133–146. doi:10.1016/j.earscirev.2009.09.001

Torsnes, I., Rye, N., Nesje, A., 1993. Modern and Little Ice Age equilibrium‐line altitudes on outlet valley glaciers from Jostedalsbreen, western Norway: an evaluation of different approaches to their calculation. Arctic and Alpine Research 25, 106. doi:10.2307/1551546

Tu, J.‐Y., Ji, J.‐Q., Sun, D.‐X., Gong, J.‐F., Zhong, D.‐L., Han, B.‐F., 2015. Thermal structure, rock exhu‐ mation, and glacial erosion of the Namche Barwa Peak, constraints from thermochronological data. Journal of Asian Earth Sciences 105, 223–233. doi:10.1016/j.jseaes.2015.03.035

Turney, C.S.M., Roberts, R.G., Jacobs, Z., 2006. Archaeology: Progress and pitfalls in radiocarbon da‐ ting. Nature 443, E3–E3. doi:10.1038/nature05214

Usoskin, I.G., Solanki, S.K., Schüssler, M., Mursula, K., Alanko, K., 2003. Millennium‐Scale Sunspot Number Reconstruction: Evidence for an Unusually Active Sun since the 1940s. Physical Review Let‐ ters 91, 211101. doi:10.1103/PhysRevLett.91.211101

Van der Burght, L., Stoffel, M., Bigler, C., 2012. Analysis and modelling of tree succession on a recent rockslide deposit. Plant Ecology 213, 35–46. doi:10.1007/s11258‐011‐0004‐2

Villalba, R., Leiva, J.C., Rubulls, S., Suarez, J., Lenzano, L., 1990. Climate, Tree‐Ring, and Glacial Fluctu‐ ations in the Rio Frias Valley, Rio Negro, Argentina. Arctic and Alpine Research 22, 215. doi:10.2307/1551585

Visser, P.C., 1938. Wissenschaftliche Ergebnisse der Niederländischen Expeditionen in den Karakoram und die angrenzenden Gebiete in den Jahren 1922‐1935, in: Band 2: Glaziologie. Brill, Leiden.

Vonmoos, M., Beer, J., Muscheler, R., 2006. Large variations in Holocene solar activity: Constraints from 10Be in the Greenland Ice Core Project ice core. Journal of Geophysical Research: Space Physics 111, A10105. doi:10.1029/2005JA011500

Von Rad, U., Schaaf, M., Michels, K.H., Schulz, H., Berger, W.H., Sirocko, F., 1999. A 5000‐yr Record of Climate Change in Varved Sediments from the Oxygen Minimum Zone off Pakistan, Northeastern Arabian Sea. Quaternary Research 51, 39–53. doi:10.1006/qres.1998.2016

Wager, L.R., 1937. The Arun River Drainage Pattern and the Rise of the Himalaya. The Geographical Journal 89, 239–250. doi:10.2307/1785796

142

Wager, L.R., 1933. The Rise of the Himalaya. Nature 132, 28.

Wagnon, P., Vincent, C., Arnaud, Y., Berthier, E., Vuillermoz, E., Gruber, S., Ménégoz, M., Gilbert, A., Dumont, M., Shea, J.M., Stumm, D., Pokhrel, B.K., 2013. Seasonal and annual mass balances of Mera and Pokalde glaciers (Nepal Himalaya) since 2007. The Cryosphere 7, 1769–1786. doi:10.5194/tc‐7‐ 1769‐2013

Walker, M.J.C., Berkelhammer, M., Björck, S., Cwynar, L.C., Fisher, D.A., Long, A.J., Lowe, J.J., Newnham, R.M., Rasmussen, S.O., Weiss, H., 2012. Formal subdivision of the Holocene Series/Epoch: a Discussion Paper by a Working Group of INTIMATE (Integration of ice‐core, marine and terrestrial records) and the Subcommission on Quaternary Stratigraphy (International Commission on Stratigra‐ phy). Journal of Quaternary Science 27, 649–659. doi:10.1002/jqs.2565

Wallinga, J., 2002. Optically stimulated luminescence dating of fluvial deposits: a review. Boreas 31, 303–322. doi:10.1111/j.1502‐3885.2002.tb01076.x

Walter, H., Breckle, S.‐W., 2004. Spezielle Ökologie der tropischen und subtropischen Zonen, 3rd ed. Elsevier, München.

Walter, H., Breckle, S.‐W., 1994. Spezielle Ökologie der gemäßigten und arktischen Zonen Euro‐ Nordasiens, 2nd ed. Fischer, Stuttgart.

Walter, H., Breckle, S.‐W., 1991. Ökologische Grundlagen in globaler Sicht, 2nd ed. Fischer, Stuttgart.

Wang, F.‐B., Fan, C.Y., 1987. Climatic changes in the Qinghai‐Xizang (Tibetan) region of China during the Holocene. Quaternary Research 28, 50–60. doi:10.1016/0033‐5894(87)90032‐9

Wang, H., Hong, Y., Lin, Q., Hong, B., Zhu, Y., Wang, Y., Xu, H., 2010. Response of humification degree to monsoon climate during the Holocene from the Hongyuan peat bog, eastern Tibetan Plateau. Pal‐ aeogeography, Palaeoclimatology, Palaeoecology 286, 171–177. doi:10.1016/j.palaeo.2009.12.015

Wang, J., Yang, B., Qin, C., Kang, S., He, M., Wang, Z., 2014. Tree‐ring inferred annual mean tempera‐ ture variations on the southeastern Tibetan Plateau during the last millennium and their relation‐ ships with the Atlantic Multidecadal Oscillation. Climate Dynamics 43, 627–640. doi:10.1007/s00382‐ 013‐1802‐0

Wang, L.‐C., Behling, H., Lee, T.‐Q., Li, H.‐C., Huh, C.‐A., Shiau, L.‐J., Chen, S.‐H., Wu, J.‐T., 2013. In‐ creased precipitation during the Little Ice Age in northern Taiwan inferred from diatoms and geo‐ chemistry in a sediment core from a subalpine lake. Journal of Paleolimnology 49, 619–631. doi:10.1007/s10933‐013‐9679‐9

Wang, P., Scherler, D., Liu‐Zeng, J., Mey, J., Avouac, J.‐P., Zhang, Y., Shi, D., 2014. Tectonic control of Yarlung Tsangpo Gorge revealed by a buried canyon in Southern Tibet. Science 346, 978–981. doi:10.1126/science.1259041

Wang, Y., Cheng, H., Edwards, R.L., He, Y., 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.

143

Wang, Y., Cheng, H., Edwards, R.L., An, Z.S., Wu, J.Y., Shen, C.‐C., Dorale, J.A., 2001. A High‐Resolution Absolute‐Dated Late Pleistocene Monsoon Record from Hulu Cave, China. Science 294, 2345–2348. doi:10.1126/science.1064618

Wang, Y., Cheng, H., Edwards, R.L., Kong, X., Shao, X., Chen, S., Wu, J., Jiang, X., Wang, X., An, Z., 2008. Millennial‐ and orbital‐scale changes in the East Asian monsoon over the past 224,000 years. Nature 451, 1090–1093. doi:10.1038/nature06692

Wang, Y., Liu, X., 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 103, 135–153. doi:10.1016/j.earscirev.2010.09.004

Wang, Y., Herzschuh, U., Shumilovskikh, L.S., Mischke, S., Birks, H.J.B., Wischnewski, J., Böhner, J., Schlütz, F., Lehmkuhl, F., Diekmann, B., Wünnemann, B., Zhang, C., 2014. Quantitative reconstruction of precipitation changes on the NE Tibetan Plateau since the Last Glacial Maximum – extending the concept of pollen source area to pollen‐based climate reconstructions from large lakes. Climate of the Past 10, 21–39. doi:10.5194/cp‐10‐21‐2014

Wanner, H., Beer, J., Bütikofer, J., Crowley, T.J., Cubasch, U., Flückiger, J., Goosse, H., Grosjean, M., Joos, F., Kaplan, J.O., Küttel, M., Müller, S.A., Prentice, I.C., Solomina, O., Stocker, T.F., Tarasov, P., Wagner, M., Widmann, M., 2008. Mid‐ to Late Holocene climate change: an overview. Quaternary Science Reviews 27, 1791 – 1828.

Wanner, H., Solomina, O., Grosjean, M., Ritz, S.P., Jetel, M., 2011. Structure and origin of Holocene cold events. Quaternary Science Reviews 30, 3109–3123. doi:10.1016/j.quascirev.2011.07.010

Warren, C.R., 1991. Terminal environment, topographic control and fluctuations of West Greenland glaciers. Boreas 20, 1–15. doi:10.1111/j.1502‐3885.1991.tb00453.x

Watson, D.F., Philip, G.M., 1985. A refinement of inverse distance weighted interpolation. Geopro‐ cessing 2, 315–327.

Weber, W., 2010. Strong signature of the active Sun in 100 years of terrestrial insolation data. An‐ nalen der Physik 522, 372–381. doi:10.1002/andp.201000019

Webster, P.J., 1987. The elementary monsoon, in: Fein, J.S. (Ed.), Monsoons. Wiley, New York, pp. 3– 32.

Webster, P.J., Yang, S., 1992. Monsoon and Enso: Selectively Interactive Systems. Quarterly Journal of the Royal Meteorological Society 118, 877–926. doi:10.1002/qj.49711850705

Weischet, W., Endlicher, W., 2000. Regionale Klimatologie, Tl.2, Die Alte Welt. Teubner, Stuttgart.

Wen, D.‐R., Liu, D., Chung, S.‐L., Chu, M.‐F., Ji, J., Zhang, Q., Song, B., Lee, T.‐Y., Yeh, M.‐W., Lo, C.‐H., 2008. Zircon SHRIMP U–Pb ages of the Gangdese Batholith and implications for Neotethyan subduc‐ tion in southern Tibet. Chemical Geology 252, 191–201. doi:10.1016/j.chemgeo.2008.03.003

Wernicke, J., Grießinger, J., Hochreuther, P., Bräuning, A., 2015. Variability of summer humidity dur‐ ing the past 800 years on the eastern Tibetan Plateau inferred from δ18O of tree‐ring cellulose. Cli‐ mate of the Past 11, 327–337. doi:10.5194/cp‐11‐327‐2015 144

Wiles, G.C., Barclay, D.J., Calkin, P.E., 1999. Tree‐ring‐dated“Little Ice Age” histories of maritime glac‐ iers from western Prince William Sound, Alaska. The Holocene 9, 163–173. doi:10.1191/095968399671927145

Wiles, G.C., Barclay, D.J., Young, N.E., 2010. A Review of Lichenometric Dating of Glacial Moraines in Alaska. Geografiska Annaler: Series A, Physical Geography 92, 101–109. doi:10.1111/j.1468‐ 0459.2010.00380.x

Wilhelm, B., Arnaud, F., Sabatier, P., Crouzet, C., Brisset, E., Chaumillon, E., Disnar, J.‐R., Guiter, F., Malet, E., Reyss, J.‐L., Tachikawa, K., Bard, E., Delannoy, J.‐J., 2012. 1400 years of extreme precipita‐ tion patterns over the Mediterranean French Alps and possible forcing mechanisms. Quaternary Re‐ search 78, 1–12. doi:10.1016/j.yqres.2012.03.003

Williams, V.S., 1983. Present and Former Equilibrium‐Line Altitudes near Mount Everest, Nepal and Tibet. Arctic and Alpine Research 15, 201–211. doi:10.2307/1550922

Wilmking, M., Hallinger, M., Van Bogaert, R., Kyncl, T., Babst, F., Hahne, W., Juday, G.P., de Luis, M., Novak, K., Völlm, C., 2012. Continuously missing outer rings in woody plants at their distributional margins. Dendrochronologia 30, 213–222. doi:10.1016/j.dendro.2011.10.001

Winkler, S., 2011. Application of Schmidt‐hammer exposure‐age dating (SHD) to improve numerical age dating of Holocene moraines. Geophysical Research Abstracts 13.

Winkler, S., 2005. The Schmidt hammer as a relative‐age dating technique: Potential and limitations of its application on Holocene moraines in Mt Cook National Park, Southern Alps, New Zealand. New Zealand Journal of Geology and Geophysics 48, 105–116. doi:10.1080/00288306.2005.9515102

Winkler, S., Matthews, J.A., 2010. Holocene glacier chronologies: Are “high‐resolution” global and inter‐hemispheric comparisons possible? The Holocene 20, 1137 – 1147. doi:10.1177/0959683610369511

Wischnewski, J., Kramer, A., Kong, Z., Mackay, A.W., Simpson, G.L., Mischke, S., Herzschuh, U., 2011a. Terrestrial and aquatic responses to climate change and human impact on the southeastern Tibetan Plateau during the past two centuries. Global Change Biology 17, 3376–3391. doi:10.1111/j.1365‐ 2486.2011.02474.x

Wischnewski, J., Mackay, A.W., Appleby, P.G., Mischke, S., Herzschuh, U., 2011b. Modest diatom responses to regional warming on the southeast Tibetan Plateau during the last two centuries. Jour‐ nal of Paleolimnology 46, 215–227. doi:10.1007/s10933‐011‐9533‐x

Wischnewski, J., Mischke, S., Wang, Y., Herzschuh, U., 2011c. Reconstructing climate variability on the northeastern Tibetan Plateau since the last Lateglacial – a multi‐proxy, dual‐site approach com‐ paring terrestrial and aquatic signals. Quaternary Science Reviews 30, 82–97. doi:10.1016/j.quascirev.2010.10.001

Wright Jr., H.H., 1990. An improved Hongve sampler for surface sediments. Journal of Paleolimnology 4, 91–92. doi:10.1007/BF00208302

145

Wu, B., Huang, R., 1999. Effects of the Extremes in the North Atlantic Oscillation on East Asia Winter Monsoon. Scientia Atmospherica Sinica 6.

Wünnemann, B., 2014. Salt lakes in China: Challenges for reconstructing ecological and climate varia‐ tions. Acta Geologica Sinica 88, 199–200.

Wu, S., Yin, Y., Zheng, D., Yang, Q., 2006. Moisture conditions and climate trends in China during the period 1971–2000. International Journal of Climatology 26, 193–206. doi:10.1002/joc.1245

Wu, X.D., 1995. Dendroclimatic studies in China, in: Bradley, R.S., Jones, P.D. (Eds.), Climate since A. D. 1500. Routledge, New York, pp. 432–445.

Xu, P., Zhu, H., Shao, X., Yin, Z., 2012. Tree ring‐dated fluctuation history of Midui glacier since the little ice age in the southeastern Tibetan plateau. Science China Earth Sciences 55, 521–529. doi:10.1007/s11430‐011‐4338‐3

Xu, X., Yi, C., 2014. Little Ice Age on the Tibetan Plateau and its bordering mountains: Evidence from moraine chronologies. Global and Planetary Change 116, 41–53. doi:10.1016/j.gloplacha.2014.02.003

Yamanaka, T., 2002. Footprint analysis using event‐based isotope data for identifying source area of precipitated water. Journal of Geophysical Research 107. doi:10.1029/2001JD001187

Yan, D., Wünnemann, B., 2014. Late Quaternary water depth changes in Hala Lake, northeastern Tibetan Plateau, derived from ostracod assemblages and sediment properties in multiple sediment records. Quaternary Science Reviews 95, 95–114. doi:10.1016/j.quascirev.2014.04.030

Yang, B., Bräuning, A., Dong, Z., Zhang, Z., Keqing, J., 2008. Late Holocene monsoonal temperate glac‐ ier fluctuations on the Tibetan Plateau. Global and Planetary Change 60, 126–140. doi:10.1016/j.gloplacha.2006.07.035

Yang, B., Bräuning, A., Shi, Y., 2003. Late Holocene temperature fluctuations on the Tibetan Plateau. Quaternary Science Reviews 22, 2335–2344. doi:10.1016/S0277‐3791(03)00132‐X

Yang, B., Kang, X., Bräuning, A., Liu, J., Qin, C., Liu, J., 2010. A 622‐year regional temperature history of southeast Tibet derived from tree rings. The Holocene 20, 181–190. doi:10.1177/0959683609350388

Yang, W., Guo, X., Yao, T., Yang, K., Zhao, L., Li, S., Zhu, M., 2011. Summertime surface energy budget and ablation modeling in the ablation zone of a maritime Tibetan glacier. J. Geophys. Res. 116, D14116. doi:10.1029/2010JD015183

Yang, W., Yao, T., Guo, X., Zhu, M., Li, S., Kattel, D.B., 2013. Mass balance of a maritime glacier on the southeast Tibetan Plateau and its climatic sensitivity. Journal of Geophysical Research: Atmospheres 118, 9579–9594. doi:10.1002/jgrd.50760

Yang, W., Yao, T., Xu, B., Ma, L., Wang, Z., Wan, M., 2010a. Characteristics of recent temperate glaci‐ er fluctuations in the Parlung Zangbo River basin, southeast Tibetan Plateau. Chinese Science Bulletin 55, 2097–2102. doi:10.1007/s11434‐010‐3214‐4

146

Yang, W., Yao, T., Xu, B., Wu, G., Ma, L., Xin, X., 2008. Quick ice mass loss and abrupt retreat of the maritime glaciers in the Kangri Karpo Mountains, southeast Tibetan Plateau. Chinese Science Bulletin 53, 2547–2551. doi:10.1007/s11434‐008‐0288‐3

Yang, W., Yao, T., Xu, B., Zhou, H., 2010b. Influence of supraglacial debris on summer ablation and mass balance in the 24K glacier, southeast Tibetan plateau. Geografiska Annaler: Series A, Physical Geography 92, 353–360. doi:10.1111/j.1468‐0459.2010.00400.x

Yao, T., Li, Z., Yang, W., Guo, X., Zhu, L., Kang, S., Wu, Y., Yu, W., 2010. Glacial distribution and mass balance in the Yarlung Zangbo River and its influence on lakes. Chinese Science Bulletin 55, 2072– 2078. doi:10.1007/s11434‐010‐3213‐5

Yao, T., Pu, J., Lu, A., Wang, Y., Yu, W., 2007. Recent Glacial Retreat and Its Impact on Hydrological Processes on the Tibetan Plateau, China, and Surrounding Regions. Arctic, Antarctic, and Alpine Re‐ search 39, 642–650. doi:10.1657/1523‐0430(07‐510)[YAO]2.0.CO;2

Yao, T., Thompson, L., Yang, W., Yu, W., Gao, Y., Guo, X., Yang, X., Duan, K., Zhao, H., Xu, B., Pu, J., Lu, A., Xiang, Y., Kattel, D.B., Joswiak, D., 2012. Different glacier status with atmospheric circulations in Tibetan Plateau and surroundings. Nature Climate Change 2, 663–667. doi:10.1038/nclimate1580

Yao, T., Xie, Z., Yang, Q., Thompson, L., 1991. Temperature and precipitation fluctuations since 1600 AD provided by the Dunde Ice Cap, China. Glaciers‐Ocean‐Atmosphere Interactions 208, 61–70.

Ye, D., 1981. Some Characteristics of the Summer Circulation Over the Qinghai‐Xizang (Tibet) Plateau and Its Neighborhood. Bulletin of the American Meteorological Society 62, 14–19. doi:10.1175/1520‐ 0477(1981)062<0014:SCOTSC>2.0.CO;2

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. doi:10.1002/jqs.1228

Yin, A., Harrison, T.M., 2000. Geologic Evolution of the Himalayan‐Tibetan Orogen. Annual Review of Earth and Planetary Sciences 28, 211–280. doi:10.1146/annurev.earth.28.1.211

Yin, J.‐J., Yuan, D.‐X., Li, H.‐C., Cheng, H., Li, T.‐Y., Edwards, R.L., Lin, Y.‐S., Qin, J.‐M., Tang, W., Zhao, Z.‐Y., Mii, H.‐S., 2014. Variation in the Asian monsoon intensity and dry–wet conditions since the Little Ice Age in central China revealed by an aragonite stalagmite. Climate of the Past 10, 1803– 1816. doi:10.5194/cp‐10‐1803‐2014

Young, J.A., 1987. Physics of monsoons: the current view, in: Fein, J.S. (Ed.), Monsoons. Wiley, New York, pp. 211–243.

Zan, J., Fang, X., Yang, S., Yan, M., 2013. Evolution of the arid climate in High Asia since ∼1 Ma: Evi‐ dence from loess deposits on the surface and rims of the Tibetan Plateau. Quaternary International, Proxies for Quaternary monsoon reconstruction on the Tibetan Plateau 313–314, 210–217. doi:10.1016/j.quaint.2013.03.012

147

Zech, M., Tuthorn, M., Zech, R., Schlütz, F., Zech, W., Glaser, B., 2014. A 16‐ka δ18O record of lacus‐ trine sugar biomarkers from the High Himalaya reflects Indian Summer Monsoon variability. Journal of Paleolimnology 51, 241–251. doi:10.1007/s10933‐013‐9744‐4

Zech, M., Zech, R., Zech, W., Glaser, B., Brodowski, S., Amelung, W., 2008. Characterisation and pal‐ aeoclimate of a loess‐like permafrost palaeosol sequence in NE Siberia. Geoderma 143, 281–295. doi:10.1016/j.geoderma.2007.11.012

Zhang, K.‐J., Zhang, Y.‐X., Tang, X.‐C., Xia, B., 2012. Late Mesozoic tectonic evolution and growth of the Tibetan plateau prior to the Indo‐Asian collision. Earth‐Science Reviews 114, 236–249. doi:10.1016/j.earscirev.2012.06.001

Zhang, P., Cheng, H., Edwards, R.L., Chen, F., Wang, Y., Yang, X., Liu, J., Tan, M., Wang, X., Liu, J., An, C., Dai, Z., Zhou, J., Zhang, D., Jia, J., Jin, L., Johnson, K.R., 2008. A Test of Climate, Sun, and Culture Relationships from an 1810‐Year Chinese Cave Record. Science 322, 940–942. doi:10.1126/science.1163965

Zhang, Y., Li, T., Wang, B., 2004. Decadal Change of the Spring Snow Depth over the Tibetan Plateau: The Associated Circulation and Influence on the East Asian Summer Monsoon. Journal of Climate 17, 2780–2793. doi:10.1175/1520‐0442(2004)017<2780:DCOTSS>2.0.CO;2

Zhang, Z.M., Dong, X., Santosh, M., Zhao, G.C., 2014. Metamorphism and tectonic evolution of the Lhasa terrane, Central Tibet. Gondwana Research 25, 170–189. doi:10.1016/j.gr.2012.08.024

Zhao, H., Moore, G.W.K., 2004. On the relationship between Tibetan snow cover, the Tibetan plateau monsoon and the Indian summer monsoon. Geophysical Research Letters 31, L14204. doi:10.1029/2004GL020040

Zhao, L., Ping, C.‐L., Yang, D., Cheng, G., Ding, Y., Liu, S., 2004. Changes of climate and seasonally fro‐ zen ground over the past 30 years in Qinghai‐Xizang (Tibetan) Plateau, China. Global and Planetary Change 43, 19 – 31.

Zheng, B., Zhao, X., Li, T., Wang, C., 1999. Features and fluctuation of the Melang Glacier in the Main‐ ri Mountain. Journal of Glaciology and Geocryology 21, 145–150.

Zheng, Y., Zhou, W., Meyers, P.A., Xie, S., 2007. Lipid biomarkers in the Zoigê‐Hongyuan peat deposit: Indicators of Holocene climate changes in West China. Organic Geochemistry 38, 1927–1940. doi:10.1016/j.orggeochem.2007.06.012

Zhisheng, A., Kutzbach, J.E., Prell, W.L., Porter, S.C., 2001. Evolution of Asian monsoons and phased uplift of the Himalaya–Tibetan plateau since Late Miocene times. Nature 411, 62–66. doi:10.1038/35075035

Zhou, S., Li, J., Zhao, J., Wang, J., Zheng, J., 2011. Chapter 70 ‐ Quaternary Glaciations: Extent and Chronology in China, in: Jürgen Ehlers, P.L.G. and P.D.H. (Ed.), Developments in Quaternary Sciences, Quaternary Glaciations ‐ Extent and Chronology A Closer Look. Elsevier, pp. 981–1002.

Zhou, S., Xu, L., Colgan, P., Mickelson, D., Wang, X., Wang, J., Zhong, W., 2007. Cosmogenic 10Be dating of Guxiang and Baiyu Glaciations. Chinese Science Bulletin Y1 52, 1387–1393.

148

Zhou, S.Z., Chen, F.H., Pan, B.T., Cao, J.X., Li, J.J., Derbyshire, E., 1991. Environmental change during the Holocene in western China on a millennial timescale. The Holocene 1, 151–156. doi:10.1177/095968369100100207

Zhu, H.‐F., Shao, X.‐M., Yin, Z.‐Y., Xu, P., Xu, Y., Tian, H., 2011. August temperature variability in the southeastern Tibetan Plateau since AD 1385 inferred from tree rings. Palaeogeography, Palaeoclima‐ tology, Palaeoecology 305, 84–92. doi:10.1016/j.palaeo.2011.02.017

Zhu, H., Xu, P., Shao, X., Luo, H., 2013. Little Ice Age glacier fluctuations reconstructed for the south‐ eastern Tibetan Plateau using tree rings. Quaternary International 283, 134–138. doi:10.1016/j.quaint.2012.04.011

Zimmerman, D., Pavlik, C., Ruggles, A., Armstrong, M.P., 1999. An Experimental Comparison of Ordi‐ nary and Universal Kriging and Inverse Distance Weighting. Mathematical Geology 31, 375–390. doi:10.1023/A:1007586507433

149

Appendices

Appendix A: Tables Table A.1: OSL details

150

Appendix B: Figures

Fig. B.1: OSL 1 sampling site (yellow cap marks the sample).

151

Fig. B.2: OSL 2 sampling site (yellow cap marks the sample).

152

Fig. B.3: OSL 3 sampling site (yellow cap marks the sample).

153

Fig. B.4: OSL 4 sampling site (yellow cap marks the sample).

154

Fig. B.5: OSL 5 sampling site (yellow cap marks the sample).

155

Fig. B.6: OSL 6 sampling site (yellow cap marks the sample).

156

Fig. B.7: OSL 7 sampling site (yellow cap marks the sample).

157

Fig. B.8: OSL 8 sampling site (yellow cap marks the sample).

158

Fig. B.9: OSL 9 sampling site (yellow cap marks the sample).

159

Fig. B.10: Geochemical ratios. N refers to Xinpu glacier, S to Gongpu glacier. 160

Fig. B.11: 14C calibration curve.

161

Fig. B.12: Tree trunk from inner part of true right moraine at Xinpu. This was used for 14C dating. Note the good state of the tree and its horizontal position with the top pointing toward the moraine center.

162

Fig. B.13: Tree growing on debris‐covered ice close to the glacier snout of Xinpu glacier.

163

Fig. B.14: Geomorphological map of Zepu glacier (legend and scale are equal to Fig. 4). Please note that this was mapped solely from remote sensing data (Landsat ETM+) and therefore has a noticeably lower level of detail than the maps of Xinpu and Gongpu.

164

Acknowledgements

A large number of people have helped and contributed to this thesis in various ways.

Foremost, I thank my supervisor Frank Lehmkuhl who initiated and coordinated the project with our collaborators in Erlangen and China. I am very grateful to Frank for giving me the opportunity to de‐ velop and explore my research interests during the past four years. In particular, Frank was, in spite of being overburdened by other work, always there to discuss research, providing numerous im‐ portant hints and insights from his impressive stock of knowledge. Of main importance, Frank has offered a nice and friendly working climate and let me work on the topics I have found most interest‐ ing, giving me the best opportunities for this thesis.

Many thanks to the colleagues from Erlangen –Achim Bräuning, Jussi Grießinger and Philipp Hochreu‐ ther– for great collaborative fieldwork, fruitful discussions, and lots of fun.

Thanks to many of my former and present colleagues RWTH Aachen University, in particular from both the physical geography departments, for the pleasant atmosphere, the enriching discussions and the great support throughout the years. Special thanks go to Georg Stauch, Veit Nottebaum and Philipp Schulte for great debates and numerous thunderstorms.

Our collaborators from the Chinese Academy of Science, especially Zhu Haifeng, did a great job in organizing and conducting joint fieldwork in spite of numerous bureaucratic obstacles.

Thanks to my family and friends for supporting me in such an outstanding way during this tough time.

My very personal thanks and wishes are dedicated to my wife Christa and my daughters Greta and Emilie for bringing so much brightness into my life.

Acknowledgements for “Reconstructing glacier retreat since the Little Ice Age in SE Tibet by glacier mapping and equilibrium line altitude calculation” (Chapter 4)

This research has been conducted within the scope of the project ‘Monsoonal variations and climate change during the late Holocene derived from tree rings and glacier fluctuations’ that was estab‐ lished by the DFG as part of the SPP 1372: Tibetan Plateau—Formation, Climate, Ecosystems. We thank the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) for funding the project. We gratefully acknowledge assistance from Zhou Haifeng for logistic support and access to the field sites. We thank Wolfgang Römer, Georg Stauch, Christian Zeeden, Veit Nottebaum, Kai Gil‐ lis, and Philipp Hochreuther for helpful comments and discussions. The editor Richard Marston and two anonymous reviewers helped greatly to improve the manuscript.

165

Acknowledgements for “Glaciers and equilibrium line altitudes of the eastern Nyainqêntanglha Range, SE Tibet” (Chapter 5)

This research has been conducted within the scope of the project ‘Monsoonal variations and climate change during the late Holocene derived from tree rings and glacier fluctuations’ which was estab‐ lished by the DFG as part of the SPP 1372: Tibetan Plateau—Formation, Climate, Ecosystems. Landsat data is available from the U.S. Geological Survey. ASTER GDEM is a product of METI and NASA. We thank Wolfgang Römer, Georg Stauch, Christian Zeeden, Veit Nottebaum, and Philipp Hochreuther for helpful comments and discussions. The editor Chris R. Stokes and the reviewers Ian S. Evans, Jakob Heyman, and Matti Peljo helped greatly to improve the manuscript and the map.

Acknowledgements for “Toward a late Holocene glacial chronology for the eastern Nyainqêntan‐ glha Range, southeastern Tibet” (Chapter 6)

This research has been conducted within the scope of the project ‘Monsoonal variations and climate change during the late Holocene derived from tree rings and glacier fluctuations’ which was estab‐ lished by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) as part of the SPP 1372: Tibetan Plateau—Formation, Climate, Ecosystems. We thank DFG for funding of the pro‐ ject (grants Br 1895‐23, Le 730‐28). Zhu Haifeng was supported by National Natural Science Founda‐ tion of China (grant 41271206). We gratefully acknowledge assistance from the Institute of Tibetan Plateau Research at the Chinese Academy of Sciences for logistical support and access to the field sites. Three Tibetan bear hunters did a great job guiding us on the three‐day walk to Xinpu glacier and back. We thank Marianne Dohms and team for support in the sedimentological laboratory. We thank Wolfgang Römer, Georg Stauch and Veit Nottebaum for comments and discussions. The con‐ structive reviewer comments of Jakob Heyman and Stefan Winkler helped greatly to improve the manuscript.

166

Abstract

The eastern Nyainqêntanglha Range is located at the southeastern fringe of the Tibetan Plateau, representing one of the major pathways of monsoonal moisture from the foreland to the plateau. Owing to the combination of a rugged high mountain topography and abundant moisture, more than 8000 km² of the eastern Nyainqêntanglha Range are covered by monsoonal temperate glaciers. Pre‐ vious studies have shown that these glaciers are particularly sensitive to climate change. The eastern Nyainqêntanglha Range’s glaciers thus provide an important archive of past and present climate dy‐ namics. The interplay of glaciers, topography and climate in this region is, however, hardly con‐ strained to date. Furthermore, a consistent late Holocene glacial chronology, combining dating re‐ sults with geomorphological and sedimentological evidence, is still lacking. The research synthesized in this dissertation aims to improve the understanding of spatial and temporal patterns of late Holo‐ cene glacier fluctuations in the eastern Nyainqêntanglha Range and to identify relevant climatic forc‐ ing mechanisms. In this context, focus was put on the Little Ice Age (LIA) because its maximum glacier advance represents the last major turning point from an advancing to a retreating glacier regime, implying a transient steady‐state. Aiming to achieve a consistent and well‐constrained overall result, the study was conducted by application of a multi‐proxy approach including glaciological, geomor‐ phological, sedimentological and dendrochronological methods, aided by optically stimulated lumi‐ nescence and radiocarbon dating. Starting with remote sensing, 1964 glaciers were mapped from a Landsat ETM+ scene and subsequently parameterized by DEM‐supported measurements. Analysis of the geomorphological evidence provided the basis to investigate patterns of modern glaciation and post‐LIA changes quantitatively, which revealed substantial retreat. An evaluation of different meth‐ ods to calculate equilibrium line altitudes (ELAs) showed that an optimized toe‐to‐ridge altitude method was more suitable than other methods in settings with complex topography, a lack of mass balance measurements, and adequate hypsographic data. The results of these studies provided in‐ sights into the complex topography–climate–glacier coupling and revealed the underlying patterns. These included a combination of distinct insolation effects, channeling of moisture by the valley sys‐ tem, foehn effects in lee positions, orographic precipitation in luv positions, and large‐scale hygric and thermal gradients. Additionally, analysis of glacier length and ELA changes showed that the sen‐ sitivity of the glacier’s reaction to climatic changes is positively correlated to both size of the glacier and the grade of continentality. Field‐work‐based work focused on the creation of a consistent mor‐ phosequence. The results yielded similar configurations and numbers of moraines at all studied set‐ tings, implying that similar climatic events affected the whole region since the LIA. The geomorpho‐ logical settings of the glacier forelands were, however, remarkably different, showing the substantial impact of the local topography–climate–glacier configurations. A critical evaluation of different rela‐ tive and numerical dating methods provided constraints for a conceptual chronosequence. They sug‐ gest that the LIA comprised several distinct glacier advances, particularly an early advance before ∼ 1500 CE and a succession to the maximum advance from the mid‐17th to mid‐18th century. After the LIA maximum, continued retreat that was only interrupted by short phases of stability followed, as evidenced by 2–3 recessional moraines in the investigated settings.

167

Kurzfassung

Der östliche Nyainqêntanglha stellt den südöstlichen Teil des das Tibet Plateau umrahmenden Rand‐ gebirges dar. Die Region ist mit ihren tiefen Tälern einer der wichtigsten Pfade für den Transport monsunaler Luftmassen aus dem Vorland auf das Plateau. Mehr als 8000 km² des Gebirgszugs sind infolge des Aufeinandertreffens dieser feuchten Luftmassen mit extremem Relief von monsunal‐ temperierten Gletschern bedeckt, die, wie vorherige Studien gezeigt haben, sehr sensitiv auf Klima‐ veränderungen reagieren und dadurch ein wichtiges Archiv für aktuelle und frühere Klimadynamik darstellen. Die komplexen Interaktionen von Gletschern, Topographie und Klima in der Region sind jedoch bis heute kaum erforscht. Darüber hinaus fehlt bis heute eine konsistente Glazialchronologie für das späte Holozän, die die heterogenen Datierungsergebnisse und Befunde geomorphologischer und sedimentologischer Analysen zusammenbringt. Die in dieser Dissertation zusammengeführten Forschungsarbeiten hatten zum Ziel, Fortschritte in diesen Bereichen zu erzielen; einerseits durch neue Erkenntnisse bezüglich der räumlichen und zeitlichen Muster der spätholozänen Gletscher‐ schwankungen im östlichen Nyainqêntanglha Gebirge, andererseits durch die Identifikationen der relevanten klimatischen und topographischen Steuerungsfaktoren. Um dies zu erreichen wurde der Fokus auf die sog. ‚Kleine Eiszeit‘ gelegt, da der Maximalstand dieser Phase die letzte große Umkehr von einer Vorstoß‐ zu einer Rückzugsphase repräsentiert, womit von einer temporär ausgeglichenen Gletschermassenbilanz sowie einem temporären Gleichgewicht zwischen Gletschergeometrie und Klima ausgegangen werden kann. Mit dem Ziel, ein konsistentes Ergebnis unter Einbeziehung mög‐ lichst vieler relevanter Faktoren zu erreichen, wurden die Studien unter Anwendung eines Multi‐ Proxy‐Ansatzes durchgeführt. Dieser vereinte glaziologische, geomorphologische, sedimentologische und dendrochronologische Methoden, flankiert durch nummerische Datierungen mittels 14C und optisch stimulierter Lumineszenz. Zu Beginn wurde in einer Fernerkundungsstudie ein 1964 Gletscher umfassendes Inventar aus ‚Landsat ETM+‘‐Daten kartiert und durch DGM‐basierte Messungen para‐ metrisiert. Auf Grundlage der Analyse geomorphologischer Fallstudien konnten die Muster der re‐ zenten Vereisung und des Rückzugs seit der ‚Kleinen Eiszeit‘ untersucht werden, die generell zu‐ nächst einen substanziellen Rückzug anzeigen. Anschließend wurden verschiedene Methoden zur Berechnung der Höhe der Gleichgewichtslinie evaluiert; hierbei zeigte sich, dass in dem von extre‐ men Relief geprägten Untersuchungsgebiet, in dem darüber hinaus zuverlässige Massenbilanz‐ und Hypsometriedaten fehlen, eine eigene Weiterentwicklung der ‚toe‐to‐ridge‘‐ Methode am besten zur flächigen Datenerhebung geeignet ist. Die Ergebnisse dieser Studien liefern Einblicke in die komplexe Relief‐Klima‐Gletscher‐Interaktion und decken zugrunde liegende Muster auf; dies sind insbesondere ein deutlicher Einfluss expositionsabhängiger Strahlungsunterschiede, Kanalisierung monsunaler Feuchtigkeit durch das Talsystem, Föhneffekte in Leelagen, Steigungsniederschläge in Luvlagen sowie ausgeprägte großräumige Niederschlags‐ und Temperaturgradienten. Darüber hinaus zeigt die Ana‐ lyse der räumlichen Muster von Daten zur Änderung von Gletscherlängen und Höhen der Gleichge‐ wichtslinien eine positive Korrelation der Sensitivität zur Größe der Gletscher und der Kontinentalität des jeweiligen Standorts. Der Fokus der Geländearbeiten lag in der Erstellung einer konsistenten Morphosequenz. Die Ergebnisse zeigen, dass an allen untersuchten Gletschern ähnliche Anzahlen und Anordnungen von Moränen vorliegen, was eine Steuerung durch gleiche, regional übergeordne‐

168 te Klimasignale impliziert. Andererseits weisen die einzelnen Gletschervorfelder aber auch deutliche Unterschiede hinsichtlich ihrer Detailausprägung auf, was die große Bedeutung der stark topogra‐ phiegesteuerten, lokalklimatischen Faktoren hervorhebt. Eine kritische Evaluierung verschiedener Datierungsmethoden lieferte das Fundament für die Erstellung einer konzeptionellen Chronose‐ quenz, die die Beobachtungen von verschiedenen Standorten in gute Übereinstimmung mit Palä‐ oklimadaten und den zuvor festgestellten räumlichen Mustern bringt. Diese zeigt mehrere Phasen mit ausgeprägten Gletschervorstößen während der ‚Kleinen Eiszeit‘ im östlichen Nyainqêntanglha Gebirge: eine frühe Phase vor ∼ 1500 CE, eine Sukzession mehrerer Vorstöße bis zum Maximalstand von der Mitte des 17ten bis zur Mitte des 18ten Jahrhunderts sowie mehrere Rückzugsstände, die auf zwischengeschaltete Stabilitätsphasen während des anschließenden, deutlichen Rückzugs hin‐ deuten.

169

30°30'N 30°N 30°N

km 100 Quzong Ts. 96°E 96°E

0 5 10 20 5200

Bomi 5000 Authors: David Loibl and Frank Lehmkuhl DEM Cartographical basis: SRTM Projection: UTM zone 46 Spheroid: WGS 1984 Prime Meridian: Greenwich (0,0) 96°E

100

96°E

100 4600

5200 95°E 95°E

Bomi 4600 100

5000 Nyingchi

4800

30°N

5200 5000

Bodui Ts. 95°30'E

95°30'E

180

160

140

4800 120

Yarlung Ts.

4800 100 80

5000 60 ELA Little Ice Age - 1999 [m]

Δ Post-’Little Ice Age’ ELA change (inset) ELA Age’ Post-’Little Ice 40

Parlung Ts.

5200 5600

Parlung Ts.

5500 5400

5000 95°E 5000

Yarlung Ts. 95°E

5000

5000 5300

5200 5000 5100

Yigong Ts. 5000

5000

4900

4800

4700

5200 4600

ELA [m a.s.l.] Calculated 1999 ELA map (main) 4500 km

94°30'E 94°30'E 5000

Nyingchi

5200 5400

5400 2,5 Cities Major rivers Major ridges 0 5 10 15 20

Common elements

30°30'N 30°N

Calculated equilibrium-line altitudes (ELAs) and post-‘Little Ice Age‘ ELA change Age‘ ELA Calculated equilibrium-line altitudes (ELAs) and post-‘Little Ice

30°30'N 30°N 29°30'N 30°N ) 0 5 3 6 (

i r

g Parlung Ts. n a K

e e n

g o Gone Kangri (6350) G n a

Salween Bomi a

h l Quzong Ts. )

g 0 7 3

n 6 ( a

i r a w r g a

t n B a Parlung Tsangpo a K n h ) u c 2 p 8 m ê e 7 ) a Zepu Kangri (6370) Z Zepu 7 0

Namcha BarwaN (7782) ( Namcha Barwa (7782)

9 q ) Brahmaputra 95°E 4 0 6 ( 0

n a 3 i 6 ) g (

1 g 5 u

n a 96°E 1 95°E g o 7 n K Kongga (6490) y (

o i l r u e Pulongu (6300) Pulongu P

N P

a l a y

n 96°E Gyala Peri (7151) Gyala Peri G Gyala Peri (7151) Gyala Peri Nyingchi r

Yigong Tsangpo e

a E ) 0 7 8 6 (

g n a n km e N Nenang (6870) China

Yarlung Tsangpo Bomi 10 India

020406080

30°N Summits 6000

Bodui Ts. 95°30'E 4000

95°30'E 2000

Elevation [m a.s.l.] Topographic overview map (inset) Topographic 0

Yarlung Ts.

Parlung Ts.

Parlung Ts. ELA map coverage ELA calculation Glacier used for ELA 95°E 95°E

Yigong Ts. Valley glacier Valley glacier network Valley

Yarlung Ts. Plateau glacier Slope glacierCirque glacier Complex valley glacier Glacier types map (main) 94°30'E 94°30'E km Nyingchi 510 Cities Landsat ETM+ scene coverage Major rivers 2,5 01520

Common elements

30°30'N 30°N 29°30'N 1999 glacier distribution Glaciers and equilibrium line altitudes of the eastern Nyainqêntanglha Range, SE Tibet Glaciers and equilibrium line altitudes of the eastern Nyainqêntanglha Range, SE