Empirical Data

This chapter presents the empirical field data. Ready to use data (maps) can be found in Figs.1 and 3. Readers wishing to modify details are referred to the following detailed sections.

Other readers can proceed directly to the maps and from that to the discussion in sections


An important point is, that extending the data from CLIMAP the Tibet Plateau was during the LGP and possibly during older neogene glacials covered by a large inland-ice that extended 2.4 Mio km2. Paragraph two presents these. These data are part of boundary conditions atmospheric and environmental simulation models, such as CSM from NCAR can use. If all components of the earth system are modeled, these data are part of the milestones the models have to meet.

Photo 1 and 2: Striations on phyllites (stick 145 cm) and some far-travelled erratic boulders (ft) (granite) in 3760 m asl indicating the LGP-glaciation in the very arid Chapursan Valley (Fig.1, No.62). Analogue-photos M. Kuhle, 24/8/2006

2.1 Overview of Existing Knowledge

A synopsis of older results and views on the Pleistocene glacier cover in Tibet has been provided by von Wissmann's compilation (1959). The glacier cover of Tibet is echoed in recent Chinese literature by Shi and Wang (1979), and has also been reproduced by the LGM reconstrcution of CLIMAP (Cline 1981). These authors speak of a 10% to maximum 20% ice cover of the mountains and plateaus of Tibet. Contrasting to the CLIMAP reconstruction Loczy (1893), v. Handel-Mazzetti (1927), Dainelli and Marinelli (1928), Norin (1932), de Terra (1932) and others (such as Kuhle 1987b, 1988e pp416/417) describe ancient ice margin sites scattered throughout the high regions of Asia. Other early researchers, such as Tafel (1914), Prinz (1927), Trinkler (1932) and Zabirov (1955) (cf. Kuhle 1988e, pp.416/417), making more or less direct use of the data they obtained by observation, reconstructed larger glacier areas. Despite their observations above authors inferred only a few hundred meters of depression of the equilibrium line altitiude (ELA).

The older literature implies a much larger glacial ice-cover during the LGM and possibly also during older Pleistocene and Neogene times than inferred by CLIMAP. According to the author's calculations, the above work represents ELA-depressions of more than 1000 m and indicate, locally, significantly more glacier cover than the von Wissmann's (1959) scheme had acknowledged. However, above-mentioned authors, interpreting their own findings did not draw the necessary conclusions (ice-extension, ELA-depressions).

The author has been fortunate in being able to carry out 39 expeditions and research visits since 1973, some of which extended to seven months, with the purpose of reconstructing the extent of glaciers in Asia during glacial periods. Two of these were to the arid East Zagros Mountains, one to the Sayan Mountains in South Siberia, the others to Tibet and its flanking mountain systems (Fig.1). The location and large number of areas under investigation permit reconstructions of the glacier areas to be made for entire Tibet. Reconstructions are supported by data from some earlier authors. They modify reconstructions with negligible ice cover as published by CLIMAP (Cline, 1981) and related authors (see Ehlers and Gibbard 2003, p.1562).

Apart from the Tien Shan, the glaciation of Tibet during the last ice age is given as approximately 2.4 million km2 (Fig.3a), and is, based on field-data, estimated to include a central thickness of about 2.7 km (Fig.3b). There was, therefore, inland ice with a central dome of about 7000 m asl in Tibet, the details of which are to be demonstrated below. Breaking up on the edges, ice discharged through the surrounding mountains as steep outlet glaciers (Kuhle 2004) (cf. among others Hughes 1998).

2.2. Evidence of a Large Inland-Ice on the Tibet Plateau

Twenty areas covered by till and erratics, or by erratics alone, ground moraines and central areas of roches moutonnées, give evidence of a former glacier cover (Figs. 1 and 3).

In addition a 150 m deep cored drillsite at 36°48'N/99°04'E documents 6-15 glacier advances and retreats (see below).

Figure 1: Index map of sites referred to in the text. The localities, quoted in the text as evidence of inland-ice, have numbers.

Drilling in the Chaka basin, 3170 m asl.

Till (brown) Grave/ (coarse) Gravel (fine and coarse) Gravel (very coarse) Till

Till (brown) Gravel (coarse) Gravel (fine and coarse) Gravel (coarse) Till with gravel Gravel-rich till Till (brown)

Gravel in loamy ground mass Tilt (brown) Gravel

Till ( with many glaciated clasts. brown) Gravel in fine ground mass (glacial day ?) Gravel

Till (with some few coarse glaciated clasts, brown ) Gravel with fine material Till (brown)

Till with coarse glaciated clasts (reddish - brown, soggy) Till with coarse glaciated clasts (reddish ■ brown)"1 Till (yefio ish - brown)

nterfingered with limnic sediments of the lake of Chaka

Figure 2: 150 m deep drilling (11 cm diameter core) 0,5 km N of the settlement Chaka (Fig.1, No.49) originating from a position on a ground moraine (till); the surface elevation is 3170 m asl. Till layers with different coloured matrix (brown, reddish-brown to terra fusca coloured, yellowish-brown) are interbedded at least 5 times with gravel and interposed limnic sediments. The gravels consist of red and white mica (biotite-rich) granites and syenite-like rocks; they are partly rounded, rounded at the edges and seldom angular; but also facetted gravels, derived from a moraine, have been found. Each change of fabric represents a period of occupation by a glacier tongue. In the meantime the site has been covered by glaciofluvial gravels which accompanied the glacial advances and retreats and also by lacustrine deposits which developed in lakes dammed up by the terminal moraines.

2.2.1. 150 m Deep CoredDrillsite

A 150 m deep core was drilled in the foreland of the Kukunor Shan (Qinghai Nanshan; Chaka Basin, Fig.1, No.49; 3170 m asl, 36°48N/99°04'E; Kuhle 1998 Fig.15, No.VIII). It exhibits alternating deposits of stratified advance scree, ground moraine and limnic sediments deposited in terminal basins. The interpretation of the core renders likely the passage of 6 to 15 Pleistocene glaciations in the foreland and, in consequence, in the interior of Tibet (Kuhle 1987b, p.261 Fig.6).

2.2.2. Northern Edge of the Central Tibetan Plateau

On the northern edge of the Central Tibetan Plateau (Fig.1, No.1) large quantities of erratic granite blocks have been deposited on ridges up to heights of 5300 m asl (Kuhle 1993, Photo 2). They lie on outcropping slatey rocks up to 700 m above the thalweg. These materials, such as basal or lodgement till, have to be regarded as glacier deposits. Thus an ice thickness of 700 m is needed to explain the deposition of the erratic blocks. At the edge of the plateau, individual outlet glaciers branched off and may have flowed down the adjacent, steeply descending valleys. This datum, from the Kuenlun Shankou (35°39N/94°05'E), is supplemented by information further to the west, where (35°38N/93°52'E) decameter-thick masses of granite blocks, having been transported from the granite barriers in Central Tibet (i.e. from the south/west), overlie outcropping crystalline schists. The granite blocks have been superimposed upon the currently unglaciated marginal chains of hills or mountains in Tibet in such considerable quantities, that they form own erratic hills (Kuhle 1987c, p.188 Fig.7). This can only be explained as result of an extensive glaciation with inland ice.

A northern inland ice margin not only explains these erratic hills, but validates them as typical for moraines which have been deposited on the plateau face. By considering the steep valley gradient, from the former glacier feeding areas to the wastage areas, the ice flow velocity must have increased sharply and thus resulted in a reduction of the glacier's cross-section. This variable ice-flow velocity in the glacier, in turn, was the reason for the breaking-up of the ice margin into separate outlet glacier tongues. The space between these bodies of ice allowed the above-mentioned moraines to accumulate.

For reasons of topography or ice-flow dynamics mentioned above, the moraines extended far above the glacial snow line. During the Late Glacial melt-down of the inland ice (i.e. the upward advance of the snow line) this accumulation received additional, and probably more substantial, supplies of moraine.

2.2.3. South of the Kuenlun Shankou

To the south of the Kuenlun Shankou 10 to 30 km wide plateau-areas are covered with boulder clay containing large, sometimes rounded, erratic blocks. Further south, this boulder clay facies thins and makes way for a slightly undulating landform with more fine-grained ground moraine (Fig.1, No.2) that is decimeter- to several meters thick. 1-8% matrix-supported gravels, with sharp to rounded edges, are embedded in the groundmass. Similar ground moraines the author observed in the extensive inland ice areas of Canada. Neither the slightly undulating morphology nor the deposits of far-travelled, sometimes centimeter long clasts and pebble-sized stones, scattered in loamy matrix, can be interpreted as fluviatile. In accordance with the North American pattern, they are to be understood as a ground moraine plain.

2.2.4. Bayan Har and Animachin Massif

Further east (Fig. 1, No.3) wide-spread granite erratics do occur from 5 km N of the Bayan Har Pass (34°10'N/97°42'E) to beyond the northern edge of the Yen Yougo Basin. They lie right in the from W to E running valley of Yeh Matan, S of Madoi (Mato) (34°41N/98°03'E), and on the NW edge of Chaling Hu (Ngorin Lake 35°04N/97°42'E; see also Kuhle 1982c, pp.74-76 Tab.1; 1987b, pp.302-311 Fig.9; Kuhle 1997; 2003 Fig.15).

These are granite erratics with large crystals of sanidine, which have been transported from the 5160 m high threshold of Bayan Har in Central Tibet over a distance of at least 70140 km. At present this threshold is completely non-glaciated. Bedrock granite can be observed in places. The erratic blocks, up to the size of a room, are facetted and belong to typical ground moraine with loamy groundmass. They have been deposited above of reddish and brown sandstones and crystalline schists. The groundmass, containing polymict blocks, shows fist- to head-sized striated quartzite boulders, locally broken out. The ground moraine covers polish depressions of meter- and polish thresholds of decimeter thickness.

This is proven by glacially streamlined hills. Evidence can also be given by the plateaus of the environment of the Haschi Scha settlement (Fig. 1, No.4; see Kuhle 2003 Fig.15, No.13).

The plains, at about 4150 m asl, consist of calcitic limestones. On these plains classical ground moraines (35°01'-12'N/98°49'-59'E), with striated quartzite blocks and erratic granite blocks as indicator-boulders are observed. They prove, that the western bedrock granites of the Animachin Massif (Figs.1, 3) had connection to the ice sheet. During 1981 the glaciation-history of this massif was investigated (details: Kuhle 1982c, 1987b). The granite blocks have been transported from the S, from the area near the Ischikai Station (Kuhle 2003 Fig.15, Nos. 13-14), over a distance of ca. 80 km. Thus here at Haschi Scha settlement was a confluence of two tributary ice-streams: One which initially came from the Mt. Payen Khola (with Bayan Har Massif) from the S, has been diverted by Mt. Burhan Budai (E-Kuen Lun; Fig.1, No.3, Fig.3a) to the E. The other came immediately from the S, i.e. from W of Mt. Animachin (Fig.1, No.44; Fig.3a). The N to NE adjacent outlet glaciers and their lowest ice margin positions have been reconstructed, too (Kuhle 1997; 2003 Fig.15) (see 2.3.15).

2.2.5. Central Tibet, Eastern Foreland of the Geladaindong Massif

In Central Tibet, in the eastern foreland of the Geladaindong Massif, north-western Tanggula Shan (33°30N/91°17'E; Fig.1, No.5; Fig.3a) there are large erratic blocks of granite overlying polished and abraded round mountain ridges of outcropping metamorphosed sedimentary rocks. However, they have not been transported far, since they originated from the central peaks of the Tanggula Shan Massif (see Kuhle 1991a Fig.43, Nos.1 and 2; 2003 Fig.1, No.3). The Tanggula Shan Massif rises up to 6640 m, and is only a few kilometers to decakilometers away. These mountain ridges, though, with erratic blocks at 5800 m asl, reach 500-600 m above the Tibetan Plateau. They are evidence of a plateau ice thickness of more than 500-600 m. Since the ELA now lies at 5600-5800 m asl, the minimum glacier level must have been above 5800 m asl, at or somewhat above the present level of the snow line. This also implies that the ice level was markedly above the then much depressed ELA. Thus Central Tibet was part of the feeding area of this inland ice.

This implies that the snow feeding surplus must have caused the thickness of the central Tibetan ice-dome surface to exceed 1000 m by far. The 5800 m high mountain ridges and even mountains which reached more than 6100 m asl (Kuhle 1991a Photos 1-4 and 6; 2003

Fig.1, No.3) have been abraded and polished round by a glaciation that covered them completely. It is thus for the LGP and pleistocene/neogene glacial intervals inferred that there was an inland ice thickness of 1000 m or more (Fig.3b), considering the basal altitude of the Tibet Plateau as 5000-5300 m asl.

2.2.6. Tanggula Shankou

Immediately south above the Tanggula Shankou (Tanggula Pass 32°50N/91°50'E, Fig.1, No.6; see Kuhle 1991a Fig. 43, No.4; 2003 Fig.1, Nos.2 and 4) there are moraine ledges with erratic granite blocks on bedrock metamorphites which run in N-S direction at 5500-5600 m asl. They are evidence that a continuous ice-cover stretched across the watershed of this central Tibetan pass. The ice surface was about 400-500 m above the origin of the valley on both sides of the pass. It belongs to the LGP (Table, I-IV). During the LGM (0 in Table) the ELA was too low to permit the formation of such lateral moraine ledges, i.e. the ice was much thicker during the LGP (see section 2.2.4). On both sides of the Tanggula Pass ground moraines containing coarse blocks are extending over tens of kilometers (Fig.1, from No.2 up to 7; see Kuhle 1991a Photos 9-13; 2003).

2.2.7. Pangnag and the Northern Nyainqentanglha

Between the settlement Pangnag and the northern Nyainqentanglha Mountains, (30°30'-32°N/91°-92 E; Fig.1, No.7, see also Kuhle 1991a Fig.43, Nos.6-16; 2003 Fig.1, Nos.4-8) there are extensive polished and abraded areas, evidence of which is found within topographical lows and on roches moutonnées. The roches moutonnées sit on the high plateau level between 4400 and 4600 m asl, are from meters to several hundred metrers high, with scattered patches of ground moraine (Kuhle 1991a, Photos 14-18, 21; 2003 Fig.1, Nos.5-8). The alignment of the roches moutonnées tends to be north to south, then veering from north/west to south/east with the flatter scour-sides facing north to north/west. Thus, they provide evidence that the central inland ice followed the general gradient of the high plateau. In many places abrasion and polishing marks as well as exaration rills on the surface of these roches moutonnées are proof of ice, scouring the outcrops of the sedimentary rocks, so that characteristic outcrop strip-polishings are wide-spread (Kuhle 1991a, Photos 16, 22, 23; 2003 Fig.1 Nos.6-8). In the same area of Central Tibet, ground moraine cover, with large facetted and striated blocks of granite and other material, is common (Kuhle 1991a, Photos 14, 15, 20; 2003 Fig. 1 Nos.5-8). Thus during the LGP a complete glacier cover results also for this area.


glacier stadium

gravel field (Sander)

approximated age (YBP)

ELA-depression (m)


= Riß (pre-last High Glacial maximum)

No. 6

150000 - 120000

c. 1400


= Würm (last High Glacial maximum)

No. 5

60000 - 18000

c. 1300


= Late Glacial

No. 4 - No. 1

17000 - 13000 or 10000

c. 1100 - 700


= Ghasa-stadium

No. 4

17000 - 15000

c. 1100


= Taglung-stadium

No. 3

15000 - 14250

c. 1000


= Dhampu-stadium

No. 2

14250 - 13500

c. 800 - 900


= Sirkung stadium

No. 1

13500 - 13000 (older than 12870)

c. 700


= Neo-Glacial

No. -0 - No. -2

5500 - 1700 (older than 1610)

c. 300 - 80


= Nauri-stadium

No. -0

5500 - 4000 (4165)

c. 150 - 300


= older Dhaulagiri-stadium

No. -1

4000 - 2000 (2050)

c. 100 - 200


= middle Dhaulagiri-stadium

No. -2

2000 - 1700 (older than 1610)

c. 80 - 150


= historical glacier stages

No. -3 - No. -6

1700 - 0(= 1950)

c. 80 - 20


= younger Dhaulagiri-stadium

No. -3

1700 - 400 (440 resp. older than 355)

c. 60 - 80


= stadium VIII

No. -4

400 - 300(320)

c. 50


= stadium IX

No. -5

300 - 180 (older than 155)

c. 40


= stadium X

No. -6

180 - 30 (before 1950)

c. 30 - 40


= stadium XI

No. -7

30 - 0 (=1950)

c. 20


= stadium XII = recent resp. present glacier stages

No. -8

+ 0 - +30 (1950-1980)

c. 10 - 20

Draft: M. Kuhle

Draft: M. Kuhle

Glacier stadia of the mountains in High Asia, i.e. in and surrounding Tibet (Himalaya, Karakoram, E-Zagros and Hindukush, E-Pamir, Tien Shan with Kirgisen Shan and Bogdo Uul, Quilian Shan, Kuenlun with Animachin, Nganclong Kangri, Tanggula Shan, Bayan Har, Gandise Shan, Nyainquen Tanglha, Namche Bawar, Minya Gonka) from the pre-Last High Glacial (pre-LGM) to the present-day glacier margins and the pertinent sanders (glaciofluvial gravel fields and gravel field terraces) with their approximate age (after Kuhle 1974-2005). The author infers comparable data for older pleistocene/neogene glaciations as well. Older times with a lower elevation of the Tibet Plateau are inferred to have also reduced glaciations.

2.2.8. East and West of Nyainqentanglha

A very wide valley, which acted as a channel for the south/west discharge of the inland ice, is situated east of Nyainqentanglha Mountains (Fig.1, No.8, see Kuhle 1991a Fig.43, Nos.15-20; 2003 Fig.1, Nos.5-8).

This 15 km wide space is accordingly marked by glacigenically abraded valley flanks (Kuhle 1991a, Photos 22, 23, 24, 27, 29; 2003 Fig.1, Nos.6-8). The flanks (preceding photos 24 and 29) of the orographic cuspate areas have been scoured out up to 6000 m asl. The flank abrasion and polishing is interrupted by tributary valleys and cirques that are presently glaciated. This presents a similar picture to mountain and valley forms in Scandinavia which were also formed by inland glaciation. Since the valley floor lies at altitudes varying between 4200 and 4600 m asl, the truncated spurs are evidence of an inland ice stream thickness of 1400-1800 m. Erratic blocks of granite (30°18'N/90°36'E; Fig.1, No.8) have been deposited up to 5270 m asl (Kuhle 1991a, Photo 30, Fig.43, No.20; 2003 Fig.1, No.8), whilst the valley floor has ground moraine deposits with erratic granite blocks (30°25N/90°56'E and 30°02'N/90°26'E) extending to several meters (Kuhle 1991a, Photos 27, 28 and 39; 2003 Fig.1, No.8). The valley-filling LGP glaciation received tributary ice flows from the Nam Co (lake) (Tengri Nor) in the north within 5200-5500 m high transfluence passes. Evidence of the tributary ice occurrs as classic U-shaped valley profiles (Kuhle 1991a, Photo 25, Fig.43 Nos. 17; 2003 Fig.1, No.8). One example is a valley leading from the Nyainqentanglha Mountains south/eastern longitudinal valley in S/E direction down to the Tsangpo River (i.e. S/E is the direction of Lhasa) which has been preserved as a glacigenic trough-valley. Up to the junction with the main valley, its tributaries are also trough-shaped valleys, thereby providing evidence of former glacier thicknesses of more than 800 m. On the valley floor of the central confluence area there are classic roches moutonnées with flat scour-sides and steeper lee-sides at 4120 m asl, which are preserved in the granite (30°02'N/90°37'E) (Kuhle 1991a, Photos 36 and 37; 2003 Fig.1 Nos.8-9).

In the area of the Lhasa Valley (Fig.1, No.9) and extending as far west to the east-trending furrow of the Tsangpo Valley, flank abrasions and polishings, polished ledges and polishing lines are preserved in the adjoining tributary valleys. They are evidence of ice thicknesses exceeding 700-1000 m (Fig.3b; Kuhle 1991a, Photos 72, 73; 2003 Fig.1, No.9).

2.2.9. Southern Tibet, North of the Tsangpo Valley

Another example is provided from an observation in southern Tibet. At 29°41'N/90°12'E, north of the Tsangpo Valley, there is a 5300 m high pass known as the Tschü Tschü La (Chalamba La) (Kuhle 1991a Fig.43, No.27, Photos 42, 43; 2003 Fig.1, below No.8). There the Trans-Himalaya valley network opens out onto the central plateau of Tibet to the E (Fig. 1 No.10).

Up to at least 200 m above the depression that forms the pass, lying on dark rhyolite bedrock with chlorite but without potassium feldspar, there are light-coloured tectonically marked (i.e. with broken and on shear faces faulted crystals) granite erratics with potassium feldspar components, but lacking chlorite (Kuhle 1988c Figs.4 and 5). Thus even at the pass the ice-thickness amounted to 200 m. The U-shaped valleys on both sides of the Tschü Tschü La pass reach a depth of about 1000 m below the pass. Ground moraine observed in the valleys indicate that they may have been filled with glacier ice well above the pass, up to at least 1200 m above the valley bottom. This suggests the existence of a network of ice streams. The nearest bedrock granite is known to occur 20-50 km further north/east. The direction of transport and the glacier direction, given by the valley system, point to numerous quasi-parallel outlet glaciers of substantial thickness. This is also suggested by roches moutonnées, which form hill- and mountain chains of several hundred metres height. Their rounded ridges rise up to 5600 m asl. Rough crest lines and peaks which begin above them, allow to estimate the thickness of the ice to 1500 m. The outlet glaciers left the Central Tibetan inland by way of an interposed section of the ice stream network interrupted by mountains of the Trans-Himalaya (Fig.3a, I2). At this longitude (about 89°E) they just missed the Tsangpo isobath at 3800 m (Kuhle 1985, 1987d). The final extremity of this particular outlet glacier ended at 4020 m asl within a side valley. It ends 6-10 km from the Tsangpo River. It was similar in the parallel Orio Machu valley (29°27'N/89°38'E; Fig.1, No.10; see Kuhle 1988c Fig.2, No.8, Fig.6).

Figure 3: (a) The reconstructed 2.4 million km2 ice sheet and ice stream network at the edges covering the Tibetan plateau (data from Kuhle, 1980, 1982a, 1982c, 1985, 1987a,b, 1988c,d, 1990d, 1991a,b, 1993, 1994, 1995b, 996a,b, 1997, 1998, 1998a, 1999, 2001, 2003, 2004, 2005, 2005a). The centers are marked I 1, I 2, I 3. Only peaks higher than 6000 m rise above the ice surface. See also Fig.1 (b) Cross section through the central ice sheet from Hindu Kush in the west to Minya Gonka in the east. Note that the ice-sheet is considerable north of the High Himalayas and the Mt. Everest. It is not a mountain glaciation but a large inland-ice in high altitudes and low latitudes. See also Pliocene temperatures and wind circulation shown in this volume.

Figure 3: (a) The reconstructed 2.4 million km2 ice sheet and ice stream network at the edges covering the Tibetan plateau (data from Kuhle, 1980, 1982a, 1982c, 1985, 1987a,b, 1988c,d, 1990d, 1991a,b, 1993, 1994, 1995b, 996a,b, 1997, 1998, 1998a, 1999, 2001, 2003, 2004, 2005, 2005a). The centers are marked I 1, I 2, I 3. Only peaks higher than 6000 m rise above the ice surface. See also Fig.1 (b) Cross section through the central ice sheet from Hindu Kush in the west to Minya Gonka in the east. Note that the ice-sheet is considerable north of the High Himalayas and the Mt. Everest. It is not a mountain glaciation but a large inland-ice in high altitudes and low latitudes. See also Pliocene temperatures and wind circulation shown in this volume.

Evidence of the ice stream extension may be found in the over 120 m high lateral moraines which coalesce into terminal moraine walls. From these a sandar or outwash cone emerges (Fig.1, No. 10). The moraine deposits with outwash cone are located in that side valley that joins the main valley near Lhasa (29°43N/91°04'E; Kuhle1988c Fig. 3). They are situated at 4250 and 3950 m asl. This implies an ELA depression of about 1075-1200 m in southern Tibet (Fig.1, No.9).

2.2.10. Lulu Valley

More to the south of Tibet, at 28°50'N/87°20'E is the Lulu Valley. It cuts deeply into hydrothermally decomposed basalt (50% pyroxenes, pseudomorphically replaced by dolomite and chlorite).

Between 4400 and 4950 m the valley floor is filled with diamictites, which extend 170 m up the valley flanks in some places. In the very fine groundmass there are isolated clasts of very coarse micaceous granite (Kuhle 1988c Figs. 18-23). Transported over long distances from the north, dozens of meters thick, spreading over tens of kilometres, these diamictites are regarded as ground moraine (Kuhle 1988c Fig.2 Nos.11-16; 1991a, Photos 76, 77).

A convergence with a debrisflow is ruled out for geomorphological, sedimentological, petrographic and topographic reasons (details in Kuhle 1988c, 1991a). In addition there are no topographic conditions on the plateau for former ice-dammed lakes, and sudden outburst of them which may have resulted in debrisflows. In case of debrisflows the erratic granite blocks for instance, deposited on the orographic right valley flank 170 m above the thalweg would require (a) a 170 m high filling of the valley with debrisflow and then (b) its complete erosion with the exception of the erratic blocks. Finally the diamictite cover passes continuously into the extensive ground moraines of the plateau, which lie above 4950 m asl and are rich in blocks.

2.2.11. The Himalaya North Slopes (Cho Oyu-Gurlamandata, Fig.1, Nos.11, 17 up to 61) to Central- (Transhimalaya-Gangdise Shan-Nganclong Kangri, Fig.1, No.60) and West Tibet (Lingzi Tang-Aksai Chin) (Fig.1, No.52): More Evidence of a Tibetan Inland Ice During the LGP

In 1996 detailed geomorphological work, including several cross-sections through representative reliefs of Tibet from the Central Himalaya to the Kuenlun have been carried out (Kuhle 1999). These included the mapping of ice-margin positions. ELA depressions between decameters and ca. 100-250 m have been inferred. The inter- and extrapolated lowest ice margin positions allowed the reconstruction of pertinent depressions of the snow line which, owing to the altitude of the Tibetan plateau reached a maximum of 400-700 m. Accordingly the early Late Glacial (Stadia I to II; Table) and High Glacial glacier traces (Riss or pre-LGP and LGM (Stadium I and/or 0) occurred over a horizontal distance of 1620 km across the plateau with an average height of 4700 m asl without the characteristics of ice margin positions.

From this profile, running from the Cho Oyu (Central Himalaya) in the SE via Gertse (Kaitse; Central Tibet; Fig.1, No.60) up to the Lingzi Thang- and Aksai Chin and from there into the Kuenlun and also from a parallel section of the Gurla Mandhata (Fig.1, No.61) in the SW up to the at present very arid Nako Tso (Fig.1, No.59), located centrally in the W, 20 sediment samples have been analysed, which provide evidence for a ground moraine genesis and therefore confirm the macroscopic field observations (Kuhle 1999 Fig.2). Only the relatively very small basin of Ali might have been - like the Indus valley chamber of Leh -free of ice even during the (LGM). Forms of glacial horns as well as roches moutonnées, flank polishings, backward abrasion of mountain spurs at intermediate valley ridges, glacial streamlined hills, but also high-lying erratics prove the wide-spread ice cover (Kuhle 1999 Fig.2). Important thicknesses of ice have been recognized by means of transfluences. Especially by and in the Nako Tso (lake) the limnic undercutting of roches moutonnées shows the Post Glacial infilling of the lake.

Summing up, the glacio-geological and geomorphological observations concerning the Tibetan investigation area 1996 (Fig.1, from No. 10 up to 52; Kuhle 1999 Figs.2 and 3), provide evidence of a glacier cover forming an inland ice, which marginally passed into outlet glaciers (Fig.3a, I3 between Shisha Pangma and Mt. Everest). The presentation of the cross section of the Tibetan margin north of Aksai Chin (Fig.1, No.52; Fig.3b left of Toze Kangri) explains the exponential increase of the glacier feeding areas as a result of the uplift of the plateau and the mountains on top of it above the ELA, i.e. as a result of the climatic drop of the ELA below the level of Tibet, or rather the relative movement of snow line and height of the plateau to each other. At a drop of the ELA from 5600 to 5000 m asl (i.e. a depression of 600 m) the feeding area already nearly triples and at a drop to 4400 m (i.e. a depression of l200 m), the glacier feeding area even reaches the margin of the plateau. This figure is regardless of the self-increase, which happens during the glacier development, and, thus, further enlargement of the feeding areas. In addition to this, we have to consider the expansion of the glacier ablation area, lying below the snow line. This enlarges the glacier surfaces by a further third. This means that already at an ELA depression of 600 m nearly the total test area of 57.581 km2 - in spite of its marginal plateau position, reaching far down - is covered with ice. With this in mind, one could perhaps get an idea how the partial uplift of Tibet above the snow line has led to a gradual expansion of the ice. Thus large-scale glaciogeomorphological findings become understandable (Kuhle 1982c; 1999).

Observations that lead to above conclusion are now described region by region. This enables future researchers to add new details, for example thicknesses of inland ice and outlet glaciers without redoing work. Readers who need only the results of this contribution can refer to the maps and tables and continue with section three, interpretation.

2.2.12. Tsangpo valley

In Central Tibet the most south-eastern area under investigation is the Tsangpo Valley. The area from the valley to the 7751 m high Namche Bawar Massif (Namcha Bawa) in the Tsangpo knee was studied for traces of past glaciations (Fig.1, No. 12, 13) (Kuhle 1991a Fig.43, Nos.36-45). In the area of the junction with the Nyang Qu (valley), near the Pula settlement, there is a ground moraine and a lateral moraine ledge (Kuhle 1991a, Photo 56; 29°27'N/94°02'E) on the orographic left flank of the Tsangpo Valley. The ledge reaches up to 250 m above the valley floor at 2950 m asl (Kuhle 1991a Fig.43, Nos.39, 40). The ledge of ground and lateral moraine can be followed down-valley for another 10 km, where it joins the lateral moraine of a tributary valley (Kuhle 1991a, Photo 57, 29°29'N/94°46'E). Reaching the Tsangpo Valley at 2950 m asl, this tributary valley moraine is evidence of an ELA depression of 900 m, using a mean height of the valley catchment area of 4750 m asl. During the LGP the Tsangpo Valley may had a far more substantial ice-filling than a mere 250 m from the valley floor (see 2.2.8-10 and Kuhle 1999). It also might have been glaciated during parts of the Late Glacial.

The following field observations support this: On the S side, that is the orographic right-hand side of the valley, 15 km up the Tsangpo Valley, there is an 80 m high exposure (Kuhle 1991a Fig.43, No.38), containing glaciolimnic sands, covered by 8 m thick varved clay (Kuhle 1991a, Photos 53-55, near the Ganga Bridge; 29°181N/94°21'E). These are the sediments of an ice-dammed lake, which existed during the LGP. This lake existed up to the Latest Ice Age (Late Glacial) when the Nyang Qu glacier advanced again into the main valley, damming-up the Tsangpo Valley. Its dating is based on the large coniferous trunks found at the base of the sediments. The lowest sample at ca. 3060 m asl has a C14-age of 9820+/-350 Yrs.BP. (sample No.26.9.89/1; laboratory No.17654; cf. Kuhle 1998 Tab.2). This minimum age indicates that during this period the Nyang Qu Glacier flowed towards a boreal coniferous forest. Such a forest continues to grow there. This proves that a minor climatic change led to the glaciation of the Central Tibet Plateau (see below 2.3.18).

The historic glacier fluctuations in the Namche Bawar area (Fig.1, No. 13) provide evidence of, possibly, extensive glaciation. In 1989 the western Namche Bawar Glacier terminated at 3900 m asl (Kuhle 1991a, Photo 68). In the 1950s its tongue advanced to the main valley at 2900 m asl, damming the Tsangpo River as a result of it. End moraines (Kuhle 1991a, Photos 70, 71) are found beyond the river. This advance extended 1000 m down the valley. Evidence of the event exists in the catastrophic discharge of the impounded lake, which destroyed a settlement further down the valley, killing more than 100 people. Further evidence of this event is the 30 years old birch found growing on the end moraine as well as the still existing dead ice complex in the moraine valley (Kuhle 1991a Fig.43, Nos.42-45). An earlier advance, down to less than 3000 m asl, created 200-450 m high lateral moraine terraces on the floor of the main valley. These terraces are (near to the surfaces) C14 dated at 4490+/-95 YBP (Kuhle 1998 Tab.2). In the lowest section of the Tsangpo bend, on the northern edge of the Namche Bawar Massif (Kuhle 1991a Fig.43, No.44) remnants of ground moraine (Kuhle 1991a, Photos 61, 62) have been preserved on the valley floor (2800 m asl, Kuhle 1991a, Photos 58, 59) beside roches moutonnées formed of bedrock gneiss. Possible LGP polished and abraded bands in trough valleys (Kuhle 1991a, Photos 64-66,69) extend up to 4800 m asl, and are evidence of glacier thicknesses of about 2000 m (Fig.1, No.13).

2.2.13 Mt. Everest, Lhotse andMakalu Area

The southern edge of Tibet, the area north of Mt. Everest, Lhotse and Makalu (Fig.1, Nos.14, 24, 22) shows many traces of glaciation by an ice stream net. This net (Fig.3a, I3) with large, continuous ice capped areas, represents the southern continuation of the inland ice area of Central Tibet (I2). It runs south of the Tsangpo Valley (see above section The relevant observations concern the area south of the Panga La (5200 m, 28°30'N/87°06'E) up to the crest of the Himalayas (Kuhle 1988c Fig.30). There are (Fig.1, No. 17) trough valleys (Kuhle 1988c Fig.2, No.20, Fig.29), which join the Dzakar Chu Valley (Kuhle 1988c Figs.53-59), which is a continuation of the Rongbuk Valley. On the valley flanks glacial polished bands are preserved on outcropping strata up to very great heights, i.e. 600-800 m above the thalweg (Kuhle 1991a, Photos 78, 79). In many places morainic materials with erratic quartzite blocks have been preserved (Kuhle 1988c Fig.28, 4400 m asl, 28°27'N/87°09'E). On the bottom of the Dzakar Chu Valley, decameter thick ground moraine deposits have been preserved (3950 m asl, 28°22'N/87°10'E, Kuhle 1991a Fig.43, No.55). Further down-valley, in the Kada (or Kharta) Valley chamber, lateral moraines, containing polymict erratic blocks, are preserved on outcropping phyllites on the orographic left- hand, that is the E valley slope at 4760 m asl, up to 1130 m above the valley floor, which reaches 3630 m asl (28°10'30"N/87°22'E, 4760 m, Kuhle 1991a Fig.43, No.56; Photos 84-86). The author interprets them as having formed during the Late Glacial times because during the Main Ice Age of LGP the glacier surface may have run above the ELA (above 4760 m asl). This could be the reason for the absence of LGP lateral moraines. This speculative interpretation is based on the following observations: in this valley cross profile, the glacigenically rounded mountain-forms, abrasions and polished bands extend up to 5680 m asl (Kuhle 1991a, Photos 82, 83 ,84 ,86 ,87). Evidence of a 2000 m thick main valley glacier is also suggested by the high scour marks of adjoining tributary valleys such as the Kada (or Kharta) Valley (Fig.1, No.14), for example (Kuhle 1991a Fig.43, Nos.57, 58 Photos 88-91). The Kharta Valley ice stream joined the Karma Valley ice stream of Mt Everest, Lhotse, Chomolonzo and Makalu via the 4890 m high Tsao La (Pass). Evidence of ice transfluence from southern Tibet in the north to the Himalayan south slope is provided by abraded and polished bands between 4900 and 5300 m asl and by an extended landscape of roches moutonnées in the vicinity of the pass (Fig.3a north of Mt. Everest; 28°00'30"N/87°16'E, Kuhle 1991a Fig.43, Nos.57, 58 Photos 92-94). With a characteristically steep surface gradient (dropping from 5200 to 3500 m asl) the Karma Valley and its tributary valleys had maximum glacier thicknesses of up to 1500 m (Kuhle 1991a Fig. 43, Nos.59-61, Photos 95101, 104). The common source area of the Kharta and Karma Valley glaciers was also shared by the north and east of Mt Everest, Changtse and Gyachung Kang (Kuhle 1991a, Photos 102-105; Kuhle 1988c Figs.58, 59 and 64-74). Here the ice reached levels of 6300-6800 m asl, and transfluences took place from the upper Rongbuk Valley (upper Dzakar Chu) via the 6010 m high Lho La (Pass) (Kuhle 1988c Figs.66, 69, 74) into the Khumbu Valley of the Himalayan south side (see below section 2.3.2), as well as via the 6548 m high Rapui La (Pass) (Fig.3a near Mt. Everest; Kuhle 1988c Fig.2, Nos.40, 41, Figs.67, 68) and the 6084 m high Karpo La (Pass) (Kuhle 1991a, Photo 105). Like all the valleys leading out of South Tibet and down through the Himalayas the LGP glacier discharged into the 2000 m thick Arun outlet glacier (see above 2.2.12 and below 2.3.4). Evidence of ice confluence (Fig.1, No.14 and 20; 28°18'N/87°22'E) is found in kilometer-wide valley junctions (Kuhle 1991a Fig.43, No.55, Photos 81, 82 very left, 83, 86 left).

2.2.14. Shaksgam Valley, Latzu Massif, Menlungtse Group, Northern Shisha Pangma Foreland

Erratics found in the Shaksgam Valley on the western edge of Tibet (36°06N/76°28'E, Fig.1, No.15) provide evidence of glaciation. Today this is the most arid part of High Asia with less than 40 mm of precipitation annually at 4000 m asl.

The area studied is a cross-section of the Muztagh Valley in the region of its confluence with the Karakorum north slope, to the north of the 8616 m high K2. The valley has been shaped into a glacial trough up to an altitude of 1200 m above the valley floor. At altitudes between 4400 and 4700 m gneiss and granite, as well as dolomite erratics (90% Do, 5% Ca, micritic and sparitic, Kuhle 1994 Figs.117, 118) have been found on roches moutonnées that consist of 90% pure calcite (Kuhle 1994 Figs.37, 38, 51, 73). The roches moutonnées are located 600 m above the valley floors on a pass that leads from the Shaksgam Valley to the Muztagh Valley. More than 1.5 m long, they occur both as single specimens and in the context of bands of lateral moraine material. Gneiss and granite erratics appear on the inner side of the Shaksgam Valley. Thus the blocks required transport along the valley at a high elevation in order to be deposited here. On the outward side of the valley to the west the pass is bounded by a 4730 m high glacial horn polished at its top (Kuhle 1994 Figs.52, 73). Up-valley and to the east of the saddle, the polished micritic calcite rises to more than 500 m above the floor of the pass. Its uppermost section is roughened. Substantial thicknesses of ice are deduced from softly polished roches moutonnées on the pass (Kuhle 1994 Fig.38). Ground polishing occurs at temperatures near the melting point as a result of increased ice pressure. Now, as well as during the glacial period, the mean annual temperature at the glacier surface was about -10 to -12°C at the equilibrium line. At the time of the LGP (LGM) the glacier surface in this area was in the accumulation zone and was located approximately 1000 m above the equilibrium line. This is an indication that the roches moutonnées within the pass are more likely to be part of the Late Glacial (e.g. younger than the LGM) since the ELA at this time was substantially higher.

Corresponding grooves can be found on the N side, that is the orographic right-hand side of the Shaksgam Trough (Kuhle 1994 Figs.69, 71) and on the 5466 m high "Shaksgam Horn" (Kuhle 1994 Fig.84) some 25 km up-valley and up to 1400 m above the gravel floor of the large West Tibetan longitudinal valley. This is the area in which a distributary stream of the Shaksgam Glacier system buried the 4863 m high Aghil Pass (36°11' N/ 76°36' E) under a 500 m thick cover (Kuhle 1994 Figs.86, 87, 116). It communicated with the Yarkand ice stream network (see below section 2.3.12.). A minimum elevation of the ice-surface is estimated for this point at ca. 5300 m asl. Glacial polishings in the massive limestones on the W, that is the orographic left-hand side of the Aghil Valley, and in the granite on the E, that is on the right, provide evidence of Aghil Valley glaciation (Kuhle 1994 Fig.90). Below 3700 m, well preserved striae in the quartzite bedrock of the Aghil-Surukwat Valley are encrusted with ferro-manganese (Kuhle 1994 Figs.40, 41, 93).

Striae have also been observed on sandstone outcrops at about 3600 m, and on roches moutonnées. Polished bands on metamorphosed schist outcrops were found at 3400-3600 m near Illik (36°23N/76°42'E, Kuhle 1994 Fig.128). This is the trough-shaped confluence area of the upper Yarkand Valley. Many other forms of glacial polishing (Kuhle 1988e; 1988d; 1994a) are also observed in this area (Fig.1, No.16). This data provide evidence of a West Tibetan Karakorum - Aghil-Kuenlun ice stream network at the time of the LGP (LGM). The glaciers of the central longitudinal valleys (Shaksgam and Yarkand Valleys, Fig.3 near K2) were large outlet glaciers which flowed down from the western margin of the Tibetan Plateau ice dome.

In addition to the observations of erratics and polished bands mentioned above, roches moutonnée fields in Central Tibet are evidence of an extensive ice cover (Kuhle 2003). In North- Central Tibet (Kuhle 1982c; 1987b), south of the Kuenlun Pass (35°33'N/93°57'E, Fig.1, from No.1 up to 2), there are roches moutonnées consisting of metamorphic sandstone and crystalline bedrock schist at 4800-5350 m asl. In the South-Tibetan Latzu Massif (28°55N/87°20'E; Fig.1, No.11; see also Kuhle 1988c Fig.2 No.16, Fig.24) roches moutonnées, consisting of basalt, occur at 5000-5500 m asl. 100-120 km further to the west, i.e. north of the Menlungtse Group (28°32N/86°09'-25'E; Fig.1, No.17), roches moutonnées, consisting of metamorphic sediments, occur between 4400 and 5100 m asl (Kuhle 1988c Fig.2, No.30). Another 50 km to the west there are roches moutonnées (metamorphites with quartzite) in the northern Shisha Pangma foreland (28°37'N/85°49'E; Fig.1, No.18; Fig.3a: Shisha Pangma). The predominantly trough-shaped valleys of northern and southern Tibet, although pebble fillings frequently lend them the appearance of box profiles, are of an appropriate character to suggest the existence of an in places greater than 2000 m thick inland ice cover (Fig.3b).

Was this article helpful?

0 0

Post a comment