Ice Thickness Lowest Position of Ice Margins around the Tibetan Plateau and the Depression of the Equilibrium Line during the Last Glacial Period LGP

2.3.1. Dhaulagiri- and Annapurna Himalaya

Determined from the position of 46 glaciers, the present ELA on the southern edge of Tibet, south and north of the Dhaulagiri- and Annapurna Himalaya (Fig.1, No.26), lies at about 5550 m asl. It was calculated following the established methods of von Hofer (1879), Louis (1955) and Lichtenecker (1938), outlined also in Kuhle 1982a, 1983, 1988a. In particular, the ELA north of the main crest of the Himalaya is at 5620 m and south of it at 5490 m. Moraines of the LGM can be found on the south slope down to 1100 m in the Mayangdi (Khola) Valley (28°23'N/83°23'E) and the Thak (Khola) Valley (1060 m asl, 28°24'N/83°36'E) (Fig.1, from No.19 to 26; Fig.3a: between Dhaulagiri and Annapurna, see Kuhle 1980, 1982a, 1983; and below section 2.3.4). On the basis of five valley glaciers the equilibrium line during the LGM was calculated as being at 4060 m asl.

The ELA-depression is calculated as follows:

where tp = recent terminus of the glacier tongue (m asl) ti = past terminus of the glacier tongue (m asl) Sp = recent equilibrium line (m asl) Si = past equilibrium line (m asl) S = equilibrium line = ELA (m asl)

ACF= average altitude of the crest fringe of the glacier (m asl) ELA, S i.e. Sp or Si = ACF - tp (m asl)/2 + tp (m asl).

By means of the glacial position of ice margins, evident in 31 terminal moraines, it was possible to establish the Paleo equilibrium line north of the main crest as being at about 3980 m asl. The large number of glacier margins is explained by the very deeply incised transverse valley of the Thak Khola, together with a catchment area of the South Tibetan Mountains that exceeds 6000 m only slightly. The transverse valley cuts through at almost 2000 m asl. North of the main crest it has only reached 2700 m altitude, that is very far below the recent ELA (Sp = recent equilibrium line). It follows that today in this area (Fig.1, No.19) hanging glaciers and those from longitudinal valleys that formerly have been outlet glaciers of the South Tibetan inland-ice and ice-stream network (Fig.3a, I3) do not coalesce into one single tongue (Kuhle 1982a Fig. 184). An integral equilibrium line for this section of the South

Tibetan Plateau edge during the LGM was at 4020 m. This is based on the depression of the equilibrium line by 1530 m (see Table). This calculation includes the altitude of the ice margin of the Jhong Khola Glacier (Muktinath Basin) during the Late Glacial Stadium at 3250 m (near Kingar: 28°49N/ 83°50'E; see map in Kuhle 1982a Figs.184, 38). It is situated in a very dry valley north of the main Himalayan crest. It is evidence of an equilibrium line at about 4450 m, the result of a local equilibrium line depression of 1210 m (Kuhle 1980, 1982a). This moraine has also been observed by Iwata et al. (1982, p.87) and Yamanaka (1982/83). The C14 dated minimum age of the corresponding glacier basin is being 8760 +/-210 YBP. This Late Glacial age and the equilibrium line depression brings it in line with other moraines, such as the terminal moraine at Ghasa ("Ghasa Stadium I" after Kuhle 1980, 1982a Figs.91, 92) in the Thak Khola transverse valley (Fig.1, No.19; Table).

Ice marginal positions and the extent of outlet glaciers flowing down from South Tibet are based on riverside moraine terraces that rise up to 570 m above the valley bottoms. They reach down to 1060 m asl on well-marked terminal moraines with erratics.

In addition there are trough profiles with glacier striae descending as far as 1700 m asl to the evergreen forest (with quercus semecarpifolia) and grooves as evidence of glacier thicknesses of 1600 m in the central Mayangdi Khola (Valley) area (28°08' N/83°23' E, Kuhle 1980, 1982a, 1983). It must be emphasized, that the depression of the ELA to 3980 m asl north of the main crest is more than 1000 m below the average level of the valley floors of the South Tibetan Mountains (Tibetan part of the Himalaya, Kuhle 1982a Fig.184). Even with a Last Glacial line altitude of 4450 m asl (see Table) there was still a difference of 600 m to the valley floors. This implies, necessarily, glaciation of the entire Tibetan Himalaya which overwhelmed the relief. The detailed hints to glacial striae, through valleys, moraines and references to comprehensive expedition-reports with photos underline this. The ice extended further north (Fig.3a, I3 north of Dhaulagiri and Annapurna). There it contributes to an inland ice.

In 2005 sheets of lodgement till containing erratics have been investigated in the upper Thak Khola transverse valley in Mustang (29°25'-28°53'N, 83°45'-84°15'E) (Fig.1, above No.19). They indicate flow of a 700-1000 m-thick outlet glacier over the 4661 m-high Kore La across the water divide from Tibet towards south. This outlet glacier received local influx from the two flanking mountain groups of the Sangda- and Damodar Himal (Tibetan Himalaya). It reached the glaciations of Dhaulagiri and Annapurna, so that a joint lowest ice margin position at less than 1100 (1060) m asl has been developed between the settlements Ranipauwa and Beni (cf. above). During the LGP this Mustang-Thak Khola outlet glacier coming down from the inland ice margin in S Tibet flowed on a very thick pedestal-groundmorane. It filled the bottom of the Thak Khola between Dhaulagiri and Annapurna up to 3000 m asl, i.e. it was 400-600 m thick.

2.3.2. The Outlet Glaciers between Shisha Pangma-, Cho Oyu-, Lhotse- and Mt. Everest Massifs and the Former Bo Chu (Valley)- and Dhudh Khosi (Valley) Glacier Network and some Remarks on Interglacial Weathering (Figs.1 and 3)

Approximately 340 km further west on the northern side of Mt. Everest conditions are similar. In this area (28°-29°50N/85°24'-91°13'E; Fig.1, No.14) investigations established a recent macro-climatic equilibrium line at almost 5900 m asl. This is the average of 15 values which are based on field work (Kuhle 1988c). The Chinese Map for snow lines and glacier equilibrium lines (Xie Zichu, 1984, 1:2000 000) indicates values between 5500 and 5900 m asl for this area. Evidence indicates that the maximum ELA depression in this area amounts to 1180 m (Kuhle 1988c). This is based on eight ice margin positions (see 2.3.3 below). The climatic equilibrium line, accordingly, ran during the LGP at 4700 m asl (Kuhle 1984/85, 1985, 1987a, 1988c).

This ELA is based on ice-margin positions of the Late Glacial, e.g. after reatreat from the LGM positions. The reason is that documents of the LGM positions have not been observed. The LGM ELA was in this area most likely lower. It is likely, that ice margin positions from the LGM can only be found near the terminal points of outlet glaciers which formerly flowed through the High Himalaya. The Bo Chu (or Sun Kosi) Glacier of the LGP (28°50'N/ 86°09'E) may serve as an example (Fig.1, No. 18) for such a terminal point of an outlet glacier. Located on the Tibetan Plateau, this glacier was supplied from the north, east and south-west flanks of the Shisha Pangma massive, as well as from the Menlungtse massif. The lowest LGP ice-margin was below ca. 900 m asl. (Kuhle 1999). Concerning research -carried out from the lowest glacial ice margin position valley-upwards - resulted in the following observations: The Sun Kosi valley chamber (e.g. valley chamber of Lamosango, 27°40'-48'N/85°45' -55'E) into which the Bo Chu outlet glacier extended, is situated at a height of the thalweg of ca. 700 and 900 m asl. There are 3 to 4 m long, rounded erratic gneiss boulders (augen-gneiss). These occur as bedrock on the Himalaya main ridge and also on the Shisha Pangma massif (Kuhle 1988c, p. 483 Fig.43). They lie on bedrock schists and metamorphic siltstones. Some of them are in the riverbed. Potholes in this augengneiss-boulders can be observed (Fig.1, No.24). For their development the boulders might have been held by the hanging glacier ice and then flushed out through the subglacial meltwater, which flowed under high pressure what resulted in such kind of cavitation corrasion. On the valley floor the in this low altitudes usually observed red weathering is lacking.

The mainly 0,5 - 2 m deep weathering (soil type: laterit, ferrosol) was developed under monsoontropic warm humid climatic conditions with summerprecipitation. Thus, during various warm interglacials with temperatures 2.5-3°C above pre-industrial values as observed in the Vostok ice-core (Petit et al. 1999) the study area was characterized by evergreen vegetation and 2500-6000 mm precipitation per year.

It is evidence of an ELA at 3900 m asl during the LGP. This is nearly the same level as in the Dhaulagiri- and Annapurna areas. Sometimes these large boulders have been dislocated by glacier water (high energy flows) or mudflows. Here they are observed in the moraine-like matrix formation. Downslope of this valley chamber the lateritic red weathering from the monsoon tropic periods auf the Pleistocene, probably the interglacial periods, is observed on the valley bottom and on the debris slopes. (Kuhle 1999).

In the further E, in the Khumbu Himalaya, an ice stream network and valley glacier system has been reconstructed (Fig.1, S of No.14) for the LGP (Wurmian, Last Ice Age, MIS 4-2, 60-18 KaBP, Table Stage 0) with glaciogeomorphological and sedimentological methods (field- and laboratory data; Kuhle 1987a; 2005 Figs.3, 11, 19). It was a part of the glacier system of the Himalaya and has communicated across transfluence passes with the neighbouring ice stream networks toward the W and E (Kuhle 2005 Fig.2 No.1, Fig.4) The ice stream network has also received inflow from the N, from a Tibetan ice stream network, by the Kyetrak-Nangpa-Bote Koshi Drangka in the W (Kuhle 1999 Figs.2 and 3), by the W-Rongbuk Glacier valley into the Ngozumpa Drangka, by the Central Rongbuk Glacier valley into the Khumbu Drangka (see below; Kuhle 1988c Fig.2) and by the Barun-Arun Glacier from the antecedent Arun Nadi transverse-valley in the E of this investigation area (see 2.3.4.;

Kuhle 1991a Fig.43). The ice thickness of the valley glacier sections, the surface of which was situated above the ELA, amounted to 1000-1450 m (Kuhle 2005 Figs.3 and 4). The most extended parent valley glaciers have measured approx. 70 km in length (Dudh Koshi Glacier). The Dudh Koshi parent glacier tongue was from the confluence of the Imja- and Nangpo Tsangpo Valley (Bote Koshi-) Glacier up to its terminal 38 km-long.

Heuberger (1986, p. 30ff) postulated the lowest past ice margin of the Dudh Koshi glacier at 1580 m asl below the Khari-Khola inflow.

The author's results show that the glacier terminal was situated near the Inkhu Khola (Valley) - junction in ca. 900 m asl (27°28'30"N/86°43'20"E; Kuhle 2005 Fig.4 and 11). This lowest ice margin position is an extrapolation from glaciogemorphological indicators, verifying in detail a decrease in ice thickness of the Dudh Koshi Glacier from ca. 1300 m, via 1200 m and 1000 m to ca. 850 m (Kuhle 2005 Fig.52-55) in the junction area of the Deku Khola (Valley) (Kuhle 1986, 1987a, 1988d, p. 587, 199; 2001, p. 389-391; 2005 Fig.4 Pro.31, Fig.11). At the Deku Khola (Valley) the altitude of the valley ground amounts to only ca. 1500 m asl. The ice thickness of 850 m has existed 12-15 km up-valley of the extrapolated lowest ice margin positions. Down to the Khari Khola (Valley) inflow, the Dudh Koshi Nadi (Valley) shows sections of a trough-like form (Kuhle 2005 Fig.52, 53, 54), i.e. of a trough-shaped glacial gorge. This is typical of steep Himalayan cross-valleys, created by glaciers with a strong, decakilometre-long subglacial meltwater erosion, which have flowed down far below the ELA (Kuhle 2005 Fig.11 on the left below No.73 up to Deku K.).

In addition near the Jubing settlement, situated in the confluence of Deku Khola (V alley) and Dudh Koshi Nad (Valley), the author has observed smooth denudation forms on the rock flanks, neither corresponding to linear erosion nor to crumblings. The interpretation is to regard them as glacigenic abrasion forms. Alternatively, as above-mentioned lateritic weathering, they formed during warm interglacials. Additional indicators to support this interpretation are the bedrock properties (phyllite).

The Dudh Koshi parent glacier tongue has received an inflow by two orographic right and three left, mostly short and steep tributary glaciers. A longer tributary glacier was the 18 km-extended Lumding Khola (Valley) Glacier, which as the lowest tributary stream has joined the parent glacier from the still glaciated right tributary valley of the same name (Kuhle 2005, Fig.11 on the right of No.16; Photo 131 below No.16). During the Late Glacial Dhampu Stage (III, Table), the tongue-end of the parent glacier has passed the Kyashar Khola(Valley) -junction by ca. 1.5 km and reached ca. 2750 m asl (Kuhle 2005 Fig.4 above Pro.28, cf. Fig.11: III on the right below No.17). During the Taglung Stage (II, Table) the end of the parent glacier was situated ca. 2 km down-valley from the Thado Koshi Khola (Valley)-junction about 2500 m asl (Kuhle 2005 Fig.4 in Pro.29; cf. Fig.11: II above Deku K.) and during the oldest Late Glacial Stage, the Ghasa Stage (I, Table), the tongue end has approx. reached the Lumding Khola (Valley) -junction at 1800 m asl (Kuhle 2005 Fig.4 in Pro.30; cf. Fig.11: I above Deku K.)

The tongue end of the Dudh Koshi Glacier has flowed down to ca. 900 m asl. At heights of the catchment areas of 8205 m (Cho Oyu massif), 8501 m (Lhotse massif), i.e. 8848 (or 8872) m (Mt. Everest, Sagarmatha, Chogolungma) this is a vertical distance of the Ice Age glaciation of 7300-8000 m. The steep faces towering up to 2000 m above the névé areas of the 6000-7000 m-high surfaces of the ice stream network were located 2000-5000 m above the ELA. Accordingly, their temperatures were so low, that their rock surfaces were free of flank ice and ice balconies. From the maximum past glacier extension up to the current glacier margins, 13 (altogether 14) glacier stages have been differentiated and in part 14C-dated (Kuhle 2005 Tabs. 2-4). They were four glacier stages of the late glacial period, three of the neoglacial period and six of the historical period. By means of 130 medium-sized valley glaciers the corresponding ELA-depressions have been calculated in comparison with the current courses of the orographic snow-line. The number of the glacier stages since the maximum glaciation approx. agrees with that e.g. in the Alps and the Rocky Mountains since the last glacial period. Accordingly, it is interpreted as an indication of the LGP (Wurmian age) of the lowest ice margin positions. The current climatic, i.e. average glacier snow-line in the research area runs about 5500 m asl. The ELA-depression of the LGP calculated by four methods has run about 3870 m asl, so that an ELA-depression of ca. 1630 m has been determined. This corresponds to a lowering of the annual temperature by ca. 8, i.e. 10°C according to the specific humid conditions at that time.

Just like the Mayangdi, Thak Khola and Bo Chu Glaciers, the Dudh Koshi glacier was an outlet of a large inland ice glacier or ice stream network which descended from the Tibetan Plateau (Kuhle 1999). The Dudh Koshi Glacier was formed by the confluence of the Nangpa La (pass) (28°06N/86°35'E; 5716 m asl) west of Cho Oyu massif, and the rerouted Central Rongbuk glacier, which had come down from the Lho La (pass) (28°00N/86°53'E; 6010 m asl) west of Mt. Everest (see 2.2.11.). Glacial abrasion and undercutting below the west shoulder of Mt. Everest (Kuhle 1988c Figs.38, 66, 69) at an altitude of 6500 m provides evidence of overflows from the north side to the Himalaya south side. In the northern forefield of the present northward draining Rongbuk Glacier the gradient of the lateral moraines tilts, so that at an ice level of 900 m above the valley floor the gradient begins to face south. This 180° reorientation of the Rongbuk Glacier explains the absence of older endmoraines more than 8 km down-valley from the present glacier tongue. There, 8 km away from the recent terminal of the Rongbuk Glacier, the lowest endmoraine of the valley is situated near the Rongbuk Monastery. It can be classified as belonging to Neoglacial-Stages V-'VII (Table). During the LGM and the LGP the glacier drained to the south (Fig.3a, north of I3). The overflows onto the steep southern ramp of the Himalaya may also explain the slight thickness of the LGM and Late Ice Age Rongbuk Glacier. Its lateral moraines run about 600 m above the recent glacier surface (Kuhle 1988c Fig.58). Further thickening of the ice became impossible due to overflow onto the steep southern side, only 5 km away. In the southern part of the highland, with a mean altitude of 4800-5000 m and valley floors of 4200 m asl at the lowest, the reconstructed equilibrium line depression of at least 4720-4300 m asl led to a relief-filling glaciation.

Ice levels were determined by topographic proximity to the steep southern edge of Tibet. The result was an approximately 900-1200 m thick ice stream network.

Glacier filling of the Tibetan mountain landscape can be compared to the Rhone Valley during the LGP. The Rhone Valley, having been filled with 1800 m of ice, finds its valley floor to be 1500 m below the LGP equilibrium line (Fig.3a, north of I3). The ELA position was here 300 m below the ice-surface. - Here on the southern edge of Tibet the ELA ran even ca. 1300-1800 m below the reconstructed ice-surface.

2.3.3. The Lowest Ice Margin Positions of Tibetan Outlet Glaciers in the Kangchendzônga Massif (South/East Himalaya; Fig.1, No.20) and the ELA Depression of the LGP

The 8585m high Kangchendzônga Massif (Fig.3a) lies ca. 160 km E of Mt Everest, Lhotse and Makalu Massifs on the south-edge of Tibet. Tied to the feeding areas above 6000 m asl the present glaciers in the Kangchendzônga area no longer exeed 10-20 km (Kangchendzônga and Yalung Glaciers, Kuhle 1990 Fig.9). During the LGP the Ghunsa and the Simbua Valleys discharged the decakilometers long ice streams of the massif. They were connected with Tibetan Ice by way of the 6114 m high Jongsang La (pass). Evidence of their dimensions is provided by trough valley forms with scouring extending up to more than 500 m (Kuhle 1990 Figs.1-4). Both valleys, the Ghunsa- and the Simbua Valley, join the Tamur Valley leading down from South Tibet. Here, at a confluence of 1500 m asl near the Hellok settlement, the oldest Late Glacial terminal moraine (Kuhle 1990 Fig.5) of the Ghasa-Stadium (I, Table) is located. In the lower 4 km of the Simbua Khola, the Late Glacial lateral moraines of the Ghasa-Stadium I have been found, extending at the Hellok settlement into the Tamur main valley (Kuhle 1990, p. 418-421). During the field work in 1999, the author (Kuhle 2001, p. 391) has encountered a ground moraine pedestal in an increasing thickness in the upper to middle Simbua Khola, which has been multifariously modified and fluvially dissected since the deglaciation. Down-valley this pedestal peters out into the level mapped as lateral moraines (Kuhle 1990 Fig.9 (I) NE of Hellok). Accordingly, it is also a dissected ground moraine pedestal. In the orographic left flank of the Simbua Khola at 3400-3430 m asl, decameter-thick ground moraine deposits with rounded components and polymict boulders have been observed as far as the Lamite Bhanjyang pass (27°30'59"N/87°53'45"E; Kuhle 2001, Photo 196). Thus the Simbua Khola glacier has reached the Tamur parent glacier and, until the Late Glacial, was one of its tributary streams. It was at least 600 m-thick in its middle course. It thus overflowed the Lamite Bhanjyang (pass), situated just 400 m above the valley bottom, down into the SE-adjacent Kabeli Khola at a thickness of at least 200 m. Above observations show that the Simbua Khola glacier functioned as an overspill from a the Himalaya ice stream network into a Himalaya fore-chain-valley without a self-glaciation worth mentioning. It must have remained rather constant from the LGP (Stadium 0) as far as into the Late Glacial (Ghasa stadium I or even II).

The lowest glacial ice terminus (endmoraine, LGP, Stadium 0) at 890 m asl lies 610 m below that of the Ghasa-stadium (I). The horizontal distance is 19 km. No endmoraines occur in this section but traces of accumulation near the settlement Marijam (near Hellok, Kuhle 1990 Figs.5 and 9), only 350 m above the lower limit, suggest an intermediate terminal position between the settlements of Tapethok and Chirwa at 1240 m altitude. These are assumed to be part of the Pre-Ghasa ice stagnation (Kuhle 1982a, p.153 Fig.184 1/2).

The lowest past glacial terminal (glacier tongue position) is witnessed by much higher ice marginal forms lying above the settlements Tapethok and Kkejinim (Kuhle 1990 Fig.9). Clear evidence for this higher marginal forms are the slope abrasions and polishings on the orographic left slope and the "glacial mills" (potholes) 500 m up-valley from the settlement Chirwa and above the settlement Mitlung (Kuhle 1990 Fig.6). 1.5 km up-valley from the endmoraine at 890 m asl on the orographic left are striae and roches moutonnées up to 250 m above the valley bottom (Kuhle 1990 Fig.7).

The lowest past glacial terminal (LGP) occurs at the hanging bridge at Thuma (890 m asl; Fig.3a Kanchendzônga) and, in a well developed glacial tongue basin, includes moraines 40-

120 m high and classical earth pyramids (Kuhle 1990 Fig.8). There are abraded and polished features 1.2 km down-valley but there is no clear evidence of moraines. In view of the considerable altitude of the confluence region (6900 m) the ELA calculated from the ice terminal at Thuma of 3900 m asl represents an ELA-depression of 1660 m. The Tamur network of ice streams functioned as the outlet system for South-Tibetan ice, north of the main range of the Himalayas (see 2.3.1. and 2.3.2.) with glaciers, as in the Tamur Valley, reaching a thickness of more than 1000 m (cf. Kuhle 1982a, pp.57-63).

The present-day climatic ELA in the Kangchendzonga Group lies at 5560 (to 5540) m (N: 5720 m, S: 5650 m, W and S down to 5250 m asl), about 10-70 m higher than in the Central Himalayas (Dhaulagiri-Himal; Kuhle 1980, p. 246; 1982, p. 168). The difference in exposure between the southern and northern sides (i.e. between windward and leeward sides of the Himalayas) and southern Tibet results in the ELA difference of 70 m (i.e. 5650 m versus 5720 m asl). This windward/leeward effect can be recognized in the whole arc of the Central Himalayas. The depression of the equilibrium line of 1660 m to 3900 m asl, is the largest calculated depression found for the whole mountain system of the Tibetan Plateau and corresponds with glacial tongues at 890 m asl. It is interesting to compare this with the lowest LGP ice terminus below 870m asl, 1750 km to the west in the Indus Valley (Fig.1, No.21; Kuhle 1988d, p.588; 1988f, p. 606; 1996a, p. 153; 1997, p. 123 and 239). It is five to ten times drier in the Karakorum and Nanga Parbat Massif at the altitude of the Ice Age-ELA (3500-4000 m asl) than in the eastern Himalayas. In the Karakorum and Nanga Parbat Massif the monsoon is absent. Therefore the similar extent of glaciers is an indicator of the lack of summer-monsoon in the Himalayas, too, during the LGP and possibly also during older glacials.

If a lapse rate of 0.6°C/100 m is used, the observed ELA depression allows the reduction of summer temperature to be calculated as at least 9.6°C. The very probable reduction in summer precipitation clearly suggests an even greater cooling than this.

2.3.4. Extension of the Barun-Arun Glacier System (Khumbakarna Himal) and of Further Outlet Glaciers From South Tibet During the LGP

This section discusses the LGP (Stadium 0, Table) extent of further outlet glaciers (Fig.1, Nos.24, 22) which flowed down from South Tibet through the Central Himalayas (Fig.3a between Nanda Devi and Kangchendzonga), following the transverse valleys, as well as lowest ice margin positions of exemplary valleys of the Himalaya S-slope in the Kumbakarna Himal, Langtang-, Ghanes-, Rolwaling-, Manaslu-, Annapurna-, Dhaulagiri- and Kanjiroba Himalaya (Dolpo). These are the lowest LGP terminal positions of the Barun-Arun Nadi (Valley), Bhote Kosi (Valley), Tamba Kosi (Valley), Buri Gandaki (Valley), Marsyandi Khola (Valley), Madi Khola (Valley), Seti Khola (Valley), Modi Khola (Valley), Thak Khola (Valley), Mayandi Khola (Valley) and Barbung-Bheri Khola (Valley) (details in Kuhle 1980, 1982a, 1983, 1998, 1998a, 2001, 2004).

During the LGP a dendritic valley glacier system has joined in the Arun parent glacier (Fig.1, No.22). It has been fed by the S-Tibetan ice stream network (Kuhle 1991a Fig.43 Nos.56, 59) as well as by the Karma-, Barun- and Irkhuwa glaciers and, accordingly, has also been nourished by the High Himalaya (Fig.3a I3 between Everest and Kangchendzonga; details Kuhle 2005 Fig.4). This composition of the parent glacier resulted from the arrangement of the Arun Nadi as an antecedent Himalayan transverse valley leading down from the Tibetan Plateau. The Arun oulet glacier descending from there, i.e. from the valley chamber of the settlement Kada, was ca. 110 km long and flowed down to ca. 450 m asl up to the inflow of the Sankhuwa Khola (27°27N/87°08'45"E; Kuhle 2005 Fig.11). The tributary streams and glaciers of the Arun parent glacier were the Barun- and Irkhuwa glacier (Kuhle 2005 Fig.4). They flowed down from the Khumbarkana Himalaya SE- to SSE-slope, from the Makalu- and Chamlang massif, up to the junction with the Arun parent glacier over a length of 61, i.e. 40 km, as far as 1100 and 700 m asl (Kuhle 1998a; 2005 Figs.4, 11). The whole length of the Barun glacier, including the fully 33 km-long tongue of the Arun parent glacier from the confluence of the two ice streams up to the lowest joint ice margin position at 450 m asl (see above), amounted to 94 km.

During the High Glacial (LGP maximum) the ice thicknesses of the Barun- and Irkhuwa glacier amounted to at least 1300 (Kuhle 2005 Figs.10, 12), i.e. 1400-1600 m according to the underlying thickness of the ground moraine and 1100 m (Kuhle 2005 Fig.14, cf. Fig.4). At low-lying valley floors of merely ca. 1100 and 720 m asl, the lower Arun parent glacier has even just reached 1100 (Kuhle 2005 Fig.15, cf. Fig.4) and 700-830 m (Kuhle 2005 Fig.16, Photo 47). In the feeding areas of the source branches of the Barun glacier, the Barun- i.e. Upper- and Lower Barun substream, the altitudes of the valley glacier levels lay at 6200-6450 m (Kuhle 2005 Figs.7-10), so that ice transfluences have taken place across 6070-6275 m-high passes into the N-adjacent Kangchung Nadi (Valley) or Karma Chu, into the W-adjacent Imja Khola (Valley) and into the E-adjacent Arun Nadi (Valley) (see 2.3.2.; Kuhle 2005 Figs. 3, 11). Due to additional transfluences from or into the Kharta valley via the Karma Chu (Valley) - and also via the upper Arun Nadi (Valley)- a further connection to the S-Tibetan ice stream network (Kuhle 1991a, p. 216) has existed in the N. There was also a most important, a fully 500 m-transfluence to the SW-adjacent Irkhuwa glacier across the Iswa La (Pass) (Kuhle 2005 Figs.3, 4, 11 No.47). Here, the joint Barun-Irkhuwa glacier level situated above this pass has dropped from the Barun glacier surface at ca. 5900 m asl down to the Irkhuwa glacier level at ca. 5300 m.

The current height of the snow-line in the catchment area of the Barun- and Irkhuwa glacier, being the source areas of the LGP Arun glacier in the Himalaya SSE-slope, amounts to 5450 m asl. This is a 200 m lower value than has been calculated for the S-side of the Kangchendzonga Himal situated 100 km further to the E (Kuhle 1990, p. 420). The calculation is based on the current orographic ELA of the S-exposed hanging glacier of the Chamlan group in the left Barun valley flank at 5300 m asl and of the Irkhuwa glacier in a SE-exposition at 5600 m asl.

The glacier tongue of the S-glacier of the Chamlang-group discussed here, ends at 4590 m asl. From this follows a difference in height of 4140 m to the lowest ice margin position of the Ice Age Barun-Arun parent glacier at 450 m asl (see above). Accordingly, a snow-line depression of 2070 m has existed (calculation of the ELA-depr.: 4590-450=4140; 4140:2=2070). As to the Irkhuwa glacier terminus at 4100 m asl the difference in height to the end of the parent glacier at 450 m asl is 3650 m. Thus, the ELA-depression was 1825 m (calculation of the ELA- depr.: 4100-450=3650; 3650:2=1825). For the S- to SE-exposition an ELA-depression of 1950 m can be calculated (2070+1825=3895; 3895:2=1947.5).

This is to say that the LGP snow-line of the Arun glacier system, i.e. in the relevant area of the Himalaya S- to SE-exposition, has run at ca. 3500 m asl (5450-1947.5=3502.5) - an ELA-value confirmed by the reconstructed LGP cirque glaciers and the corresponding altitude of the cirque level between 3300 and 3600 m asl (Kuhle 2005 Photo 26, Fig.11 on the right and half-right above No.46). The Kasuwa glacier, which in an extreme windward position is exposed to precipitation, testifies to a local Ice Age ELA at only just 3025 m asl.

The Arun parent glacier is at the same time an outlet glacier from the margin of the S-Tibetan ice stream network (Fig.3a, I3). Thus, the Himalaya lee-side with its snow-line about 4200-4300 m asl, has also been considered with regard to the joint lowest ice margin position at 450 m asl. An averaging of the orographic snow-lines in windward- and leeward positions yields an ELA at ca. 3640 m asl (4250-3025=1225; 1225:2=612.5; 3025+612.5=3637.5). A favourable factor, which despite this snow-line - which against the 3500 m-ELA (see above) calculated for the Himalaya SSE-slope was 140 m-higher - enabled the Arun outlet- and parent glacier to extend down to 450 m asl, is the ca. 1300-2000 m-thickness of the LGP ice stream network on the S-margin of Tibet. This feed-back self-heightening of the cold-based ice stream network due to the flat gradient of discharge, must have been the cause for a secondary heightening and extension of the feeding areas.

As for the lowest ice margin position at 450 m asl (see above) the LGP snow-line (ELA) of the Arun glacier established about 3500 m asl (see above) corresponds to a medium height of the glacier feeding area of 6550 m (3500-450=3050; 3500+3050=6550). This applies approximately to the catchment area built up from S-Tibet and the High Himalaya which here rises up to 8481 m asl (Makalu Massif). Comparable heights of catchment areas made up of N- and S-slope have also been calculated for the Dhaulagiri- and Annapurna Himalaya (see 2.3.1.; Kuhle 1982a, p.150-152) and the Mt. Everest- and Shisha Pangma Massifs (see 2.3.2.; Kuhle 1986, p.443-452, Tab.3; 1988c, p.468-470; 2005).

In the Kangchendzonga massif 100 km further to the E, the reconstructed High Glacial ELA ran at about 3900 m asl, i.e. ca. 400 m-higher than in this investigation area (see 2.3.3.) According to the current state of knowledge the reconstructed snow-line in the Kangchendzonga Massif amounted to ca. 1660 m (Kuhle 1990, p. 420), i.e. 290 m less than the ELA-depression of 1950 m calculated for this research area.

The LGP ice-margin positions of the Langtang-Himal SW- and Ganesh Himal SE-slope and in the S-slope of the Menlungtse-group, Rolwaling-Himal and of the Manaslu Himalaya (Fig.1, No.25) are described as follows: Langtang- Ganesh-Himal: On the W-flank of the Bhote Kosi (Valley) as far down as its talweg (Trisuli river), decametres-thick ground moraines have been preserved (Kuhle 2001 Photo 192). The lowest occurrences of ground moraine have been met at ca. 900-1000 m asl near the Donga settlement. It proves that the terminus of the LGP Langtang glacier tongue has reached the junction of the Mailung Khola (Valley) (28°05'N/85°13'20"E). This valley glacier tongue was a joint outlet glacier tongue of the connected Langtang- and Ganesh Himal ice stream network (Kuhle 2001).

Rolwaling-Himal: The LGP-glacier had in the Tamba Kosi, in the cross-profile near the location of the Jagat settlement, a minimum thickness of 800 m (Kuhle 2001, Photo 193). Down-valley at 900 m asl, a kame near the Malepu settlement provides evidence of an ice thickness of still 200 m (Kuhle 2001, Photo 194). Thus the glacier tongue end might have been situated 4 km down-valley at 860 m asl (27°38N/86°06'E; at the start of the valley narrow SW below the Marbu settlement; Kuhle 2001).

Manaslu-Himal: The Buri Gandaki outlet glacier also discharged the S-Tibetan ice stream network (Fig.3a, I3 near Manaslu). It passes the Himalaya E of Manaslu (Fig.3a). Its glacio- geomorphological clear traces are completely mapped from the Chinese border in S-Tibet (N of the Himalaya) up to the southern Himalaya foothills. "Glacier mills" or potholes occur on the glacially polished valley flanks at 200-300 m above the thalweg, i.e. without being bound to a run-off rill of the slope. The same applies to flank abrasion and roches moutonnées. The glacier flowed down to at least 680 m asl at 28°08'N/84°51'E (Fig.1, No.24; Kuhle 1998a).

The westward adjacent Marsyandi outlet glacier flowed down from S-Tibet (Fig.1 between Nos.19 and 23) through the Himalaya Range between the massifs of Manaslu- and Annapurna (Kuhle 1980, 1982a, 1983). Cross-sections of trough valleys, glacially shaped horns, flank abrasion and polishing provide evidence of the glacier's discharge during the LGP. Overall this is proven by a very extended landscape of ground-, lateral- and endmoraines with large to very large erratic augen-gneiss and granite blocks, reaching up to the settlement of Dumre at the Himalaya foothills (28°07'20"N/84°26'E, Kuhle 1998a). The lower course of this young outlet glacier is very well set off against the deep tropical red weathering (ferraltic or laterite weathering) of the adjacent low lying areas. The glacier tongue terminal reached down to 460 m asl. (Fig.1, No.25). That is the lowest LGP glacier margin position the author was able to observe on the edge of High Asia.

Now the lowest LGM ice margin positions of the Annapurna Himalaya (Fig.3a) are discussed (details: Kuhle 1982a).

In the Madi Khola (Valley), located to the W of Marsyandi Khola (Valley), another lowest glacier end was reconstructed. More than 2 km long lateral moraine on the E flank and large erratic gneiss blocks on phyllite bedrocks on the W flank are found down to ca. 630 m asl (28°12'20"N/84°05'20"E, Fig.1 No.26; Kuhle 1998, p.87; 2001, Photo 191) in the Madi Khola (Valley). The glaciers that formed these flowed down from the Annapurna IV, II and Lamjung Himal,

The LGP Seti Khola glacier has been mapped as being uncertain (details in Kuhle 1982a). Based on two up to 3 m-long erratic gneiss boulders on a rock head of schist bedrock W of the Ghachok settlement at 1500-1540 m asl (Kuhle 2001, pp. 383/384, Photo 190), 350390 m above the current talweg, its LGP terminal position can be extrapolated at ca. 1000 m asl. At an ice thickness of ca. 300-400 m and a glacier width of 2.5-3 km at Ghachok, the LGP Seti Khola glacier has most probably reached the junction of the Seti- and Yamdi Khola situated 6 km further down-valley (28°16'20"N/83°57'30"E). The erratic boulders derive from the Himalaya main crest (Annapurna III-IV). According to the arrangement of their positions they can neither be explained by mudflows (or related phenomena) nor by a Late Glacial glacial extent (Kuhle 2001, 2004).

The tongue of the Modi Khola glacier received an influx from the Chomrong Khola- and Kyumnu Khola glacier during the LGP (Stage 0) and flowed down as far as ca. 800 m asl up to the Dobila locality at the junction of the Jare Khola (Valley) (28°13'50"N/83°43'E; Kuhle 2001, Photo 188). The trough valley cross-profile of the lower Modi Khola (Valley) reaches up to there, as does an orographic right kame-terrace. Down-valley the glacier mouth gravel floor terraces of Chuwa (Kusma) set in (Kuhle 1982a Bd. I, p. 71/72, Bd. II, Abb. 7 and p. 105-108; Kuhle 198, .p 193-196). Among others, the following observations testify to this past glacier extension: ground moraines up to at least 2400 m asl at the spur between the Chomrong- and Modi Khola (Valley) and on the orographic right side at the exit of the junction area of the Kyumnu Khola up to 2000 m asl (near settlement Udi); prehistoric subglacial potholes on the orographic left between settlement Landrung and Tolka at a height of 300-500 m above the current talweg and an orographic right ground moraine cover at the inflow of the Bhurungdi Khola at ca. 1450 m asl. Accordingly, the Modi Khola glacier near

Birethanti (see Kuhle 1982a, Bd. II Abb.8,d) still had a thickness of approximately 400 m (Kuhle 2001, Photo 189; 2004).

The South-Tibetan Ice Age glacier areas of Dolpo and Kanjiroba (Fig.1, No.53) were located WNW of the Thak Khola and Mayangdi Khola outlet glaciers (mentioned above), which flowed down to at least 1100 m asl (Fig.1, from No.19 up to 26; Kuhle 1982a Fig.8h and e). During the LGP the Barbung-Bheri Khola outlet glacier discharged these areas. Preserved in exposed positions, fresh ground moraines and large erratic granite blocks at the settlement of Tripurakot (29°02'N/82°46'E) suggest a minimum thickness of this glacier of still 400 m at 1900 m asl (Fig.1, No.27; Kuhle 1998a). The lowest glacier end was located down-valley in the Bheri Gorge. It could not be visited.

2.3.5. The Ice Age Glaciation of the Garhwal (Chamoli) Himalaya

In this marginal area of southern Tibet investigations focussed on the valley system of the Nanda Devi-Kamet Group (Fig.1, No.28), which converges in the Alaknanda Nala (Valley). The lowest ice margin position in this main valley is at Pipalkoti (30°34'N/79°25'E; Kuhle

1997, p. 127) at 1000-1100m asl. The lowest glacier tongue of the Nanda Devi-Kamet ice stream net, which received supplies from inland ice outlet glaciers from southern Tibet via the ca. 5000 m high passes of the Tun Jun La (Marphi La) and the Shashal La, ended near this village in the Alaknanda Valley (Fig.3a: I3 between Kamet and Nanda Devi). Branching out into hundreds of kilometres of valley glacier, this ice stream net had preserved all these lowest end- and lateral moraine exposures. The mean altitude of its catchment area is at about 6000 m asl. The Ice Age ELA on this precipitation-rich SSW edge of Tibet is calculated to be at ca. 3500-3600 m asl (implying an ELA depression of ca. 1400 m). In the Gohna Nala (Valley), leading down from the 6300 m high Nanda Ghunti (western Trisul Massif), LGP (Table, Stadium 0 or I) moraines of an ice marginal position were found at 1800m asl. These moraines confirm a climatic snow line at about 3600 m asl

The calculation of the ELA is explained above. Using above principle: Mean altitude of the catchment area: 5400 m minus altitude of the observed ice-margin (1800 m) is 3600, divided by 2 is 1800 plus 1800 (altitude of the observed ice-margin): The result is 3600 m asl.

This implies that tectonic uplift that might have taken place in the last 18000 years or 2.75 Ma is not accounted for. If the tectonic uplift of 10-25 mm/a that is observed today, applied also to the LGP, the value would be ca. 200-500 m lower. The same applies in a modified form to the end of the preceding glacial.

In order to confirm this value, the Nandakini Nala (Valley) on the Trisul south-western slope was investigated. The present Trisul South-Glacier extends down to 3500 m asl. The mean altitude of its catchment area is 6500 m asl, indicating an ELA at about 5000 m asl:(6500-3500)/2 = 1500+3500 = 5000. The lowest former ice margin position in the valley, evidenced by polished rock surfaces, lateral and ground moraines with erratic blocks, was at 1200-1400 m asl (near the settlement Khunana: 30°16'45''N, 79°24'20''E). It represents a snow line depression in the Nandakini Valley of ca. 1050 m - 1150 m to 3850 m asl (Kuhle

During the LGP, in the western parallel valley of the Alaknanda Nala (Valley), the Mandakini Nala (Valley), in the S-slope of the 6940 m-high Kedarnath massif, the dendritic Mandakini valley glacier network reached down to 1100 m asl up to the Okhimath settlement (30°31N/79°06'E). This could be verified by geomorphologic field investigations and sedimentological analyses in 2004 (Kuhle 2004, Map 1:1 Mio; 2005 Fig.2).

In the main valley, the Bhagirathi Nala, situated still further west, the LGP Bhagirathi main glacier tongue of the ice stream network west of the 7138 m-high Chaukhamba massif, as the continuation of the currently 31 km-long Gangotri glacier, has reached down to ca. 1050 m asl. As indicated by moraine material and abruptly starting large gravel terraces, this glacier tongue end has reached the Slalam Gad (tributary valley) (30°44'N/78°24'E; Fig.1 No.28) 3 km down-valley of the city of Uttarkashi (Kuhle 2004, Map 1:1 Mio; 2005 Fig.2).

2.3.6. Ladakh Range and Zanskar Himalaya (Upper Indus Valley), Lahaul valley and SE Pir Panjal Range (Fig. No.54)

There are two parallel valleys on the south slope of the Ladakh Range (Fig.1, No.29), the Phyang Valley and the Leh (Puchu Chu) Valley. In both of them well preserved end moraines extend down to 3400 m asl. They are regarded by relative dating as being remnants of the Late Glacial (Table, Stadium I). These ice margin positions are evidence of a substantial snow line depression of ca. 1200 m. Assuming the mean altitude of the catchment area to be 5300 m asl, the altitude of the former snow line is calculated to be 4350 m asl. Ground moraine and roches moutonnées, outside and below these tongue basins, are evidence of LGP glaciation in the upper Indus Valley (altitude at more than 3000 m asl). Scouring bands prove, that in the junction area of the Taglang La North-Valley, which runs down from the 5100 m high Taglang La (Pass) on the 6401 m high Ruberung Massif, the valley floor of the Indus Valley had been covered by glacier ice approximately 1000 m thick (see below 2.3.9; 2.3.10).

On the flanks of the Taglang La North-Valley lateral moraines with erratic blocks on outcropping crystalline slates (33°40'N/77°43'E) extend to above 4000 m asl, thus confirming corresponding ice thicknesses for the Indus tributary valleys.

Considering the glacially abraded flanks and truncated spurs above the moraines, this ice thickness might have been still 400-600 m more, i.e. the polished slopes were lying above the pertinent snow line. Ground moraines, up to decameter thicknesses, occur in the upper Indus Valley between the Nimu (mouth of the Tara Phu (Valley)) and Khalsi settlements (34°08'-20N/77°25'- 76°50'E). There are up to 900 m high abraded and polished mountain forms (roches moutonnées). Up and down valley from the Khalsi settlement large erratic blocks of granite on outcropping metamorphites are widespread. In many places they appear to have been washed out from the ground moraine matrix, possibly, by the former (Holocene) Indus River. These observations are evidence that the upper Indus Valley between Upshi and Khalsi had been filled by a valley glacier that was the main branch of the Ladakh Zansgar ice stream net (see below in addition 2.3.9 and 2.3.10).

Research in the western Zanskar Himalaya, in the Nun Kun Group (Fig.3a; 33°59'N/76°01'E), yielded provided data about a very large ice stream during the LGP. The entire Nun Kun Group, the highest peak of which reaches 7315 m, was filled with glaciers up to a level of at least 4200 m asl, as shown by granite erratics on slates. The floor of the Karcha or Suru Valley has roche moutonnée forms at Parkachik that lie between 3200-3400 m asl (Fig.1, No.30; details Kuhle 1998, p. 89, Fig.10). In the Kargil Basin, 70 km away from the Nun Kun Massif, ground and lateral moraines, as high as 3150-3200 m asl, provide evidence of LGP glacier thicknesses of at least 600 m. A NNW facing corrie, with end moraines at 3700 m asl, is evidence of a snow line at 3900-4100 m asl.

Between the Ruberung Massif and the Takh settlement (North-Lahoul), 100 km further south (Fig.1 between No.29 and 31), there is an elevated area with peaks of 6000-6632 m and valley floors at about 4400-4700 m asl. According to the character of its landscape, this area is part of Western Tibet (at 33°N/78°E). This Western Tibetan area has a network of trough valleys, separated from one another by flat transfluence passes. These valleys tend to be without valley heads (i.e. forming a widely-branching network of merging trough valleys) similar to those found in Southern Scandinavia (Jotunheimen). This is characteristic of a high valley landscape scoured by inland ice. There the ice-filled valley-network is comparable. The mountains themselves pierced well above the inland-ice.

Further south, the mountain groups known as Lahaul (Lahoul) or Pir Panjal, up to the Manali settlement in the Kullu Valley or Solang Nala had also been filled during the LGP (Fig.1, No.31). The 3980 m-high Rothang (Jot) pass (32°21'45"N/77°14'50"E) was a transfluence pass over which the Lahaul glacier (Chandra glacier) has flowed into the Kullu valley adjacent to its S (Kuhle 2001, pp.381/382, Photo 185). This is evidenced by glacigenic abrasion forms like trough valley profiles with steep flanks, polished bands, transfluence passes and roches moutonnées and their ground moraine covers. Due to this transfluence the level of the Lahaul glacier has run at a height just about 4300-4400 m. Its ice thickness has amounted here to ca. 1100-1200 m (Kuhle 1998, 2001).

The two parallel valleys which from the 4971 m-high SE Pir Panjal massif NE of Dharamsala peter out into the Himalaya foreland to the SSW, have been glaciated down to the foreland. Their lowest ice margin positions, recognizable by well-preserved lateral- and end moraines, can be considered as belonging to the LGM. They are situated below the exits of the Tori valley at 1250 m (32°12'30"N/76°22'30"E) and the Triund valley (32°12'50"N/76°21'10"E) at 1200 m asl (Kuhle 2001, pp. 383/384, Photos 186, 187).

2.3.7. Extent of Glaciers and Snow Line Depression in the North-Western Karakorum (Fig.1, No.55; Photo 1 and 2)

The Shimshal Pass (4600 m asl; 36°26'N/75°41'E) is a broad saddle in the Ghujerab Mountains north of the Karakorum main crest. The pass leads from the north slope (Shaksgam Valley; see also 2.3.10) to the south slope via the Shimshal and Hunza Valleys, and down to the Indus Valley (also 2.3.9 and 2.3.10; details Kuhle 1988d). It was a transfluence pass, with a minimum ice thickness flow of 250-400 m (Fig.1, between No.55 and 32). Evidence of this is found as elongated moraines SSE of the pass, containing blocks of granite, limestone and metamorphites which extend up to 4850 m asl. In some places the moraines interfinger with glacio-fluvial drift, implying that at the time of deposition the ELA must have been above 4850 m asl. The ELA is now running at about 5200-5300 m asl, indicating that transfluence over the Shimshal Pass is part of a snow line depression of, at most, 500 m (350-450 m). Only based on the ELA depression, it is tentatively dated as Late Glacial (Table, Stadium IV). In the LGP, however, the snow line was 1100-1300 m lower (see 2.3.8, 2.3.9, 2.3.10) than at present. Thus during the LGM (Table, Stadium 0) transfluence ice thickness of the Karakorum north- and south slopes must have been several hundred metres more than 250-400 m (Fig.3b Karakoram). Ground moraine is observed in the Shimshal Pass. Twenty-seven kilometres down-valley towards the west, on the N side near the junction with the Shimshal Valley, there are erratic blocks of granite on a limestone mountain spur at 4300-4450 m asl (Chatmerk Pass, 36°28'N/75°26'E), about 1300-1400 m above the floor of the Shimshal Valley. The parent rock of the erratics outcrops approximately 20 km up-valley in the area of the Shimshal Pass. Evidence of ice thickness in the Shimshal Valley is observed as abraded valley flanks and truncated spurs on the W side of the Shimshal Valley. However, this relatively large ice thickness may be regarded as Late

Glacial, since the erratics observed must have been deposited below the snow line. The Late Glacial ELA occurred at more than 4450 m asl and, most likely 4450m asl, not more than 750-850 m below the present ELA. Preserved 1400-1600 above the valley bottom, a ledge of lateral moraine remains on the E side in the Shimshal Valley, down-valley from the camp Ziarat, at 4100-4300 m asl (36°32'N/75°04'E). Thus the maximum ELA depression is 700900 m (present ELA level at this location: 5000 m asl) to allow for moraine deposition below the snow line. This places the ca. 1400-1600 m thick Shimshal Glacier, as deduced from above-mentioned moraines, tentatively into Late Glacial times (Table, Stadium II-III). Quartzite roches moutonnées on the dividing ridge between the Lupghar and Momhil Valleys (4300m asl; 36°28'N/75°02'30"E) infer an equivalent or even more substantial LGP thickness in the valleys. The longitudinal axis of the roches moutonnées, transverse to the mountain ridge, is evidence of the transfluence of the glacier over the ridge from east to west with an ice thickness of several hundred meters. The occurrence of glacially abraded valley flanks, truncated spurs and ice scour limits, indicates an LGP (LGM) surface at 4500-4800 m asl. The considerable ice thickness in the lower Shimshal Valley area matches interpretations made in the Hunza Valley. Schneider (1959, p.209) observed erratics on the Shanoz Ridge at a height of 4000-4150m asl (36°35N/74°41'E). The erratics occur south of the Batura Glacier (1000 m above the present-day surface), and 1500 m above the floor of the Hunza Valley.

West of the junction of the Shimshal Valley with the Hunza Valley (36°28'30"N/74°00'50"E) (Fig.1, between No.55 and 33; Kuhle 1988d) large polymict erratic blocks - such as gneiss, quartzite, porphyries and granite - were found on a round-polished limestone (calcite and marl) mountain near its peak at 3530 m asl (details Kuhle 1998, p. 92, Fig.12). Adjacent to the erratics are relicts of fine-grained moraines that provide evidence of a young, probably Late Glacial (Table) age ice stream. These moraines and erratics are evidence of a Hunza Glacier of more than 1000m thickness and a glacier surface at 3500m asl. As erratics indicate minimum glacier thickness in the area below the snow line, it is assumed that in the area of the former Hunza Glacier net with a glacier surface at around 4250-4500 m asl, the LGP glacier thickness was at least 1750-2000 m (see below The Hunza ice stream joined the Gilgit Glacier and contributed to the supply of the Indus Glacier enabling it to reach its lowest ice margin position down valley of the Sazin settlement at below 870 m asl (Fig.1, No.21; Kuhle 1988d, p.588; 1988f, p. 606; 1996a, p.153; 1997, p.123 and 239). In discordance with the here presented results Haserodt (1989) and Derbyshire et al. (1984) have recorded the Late Glacial ice margins of the Hunza Glacier (see below 2.3.9). Thus according to Haserodt and Derbyshire et al. the Hunza glacier was much shorter than the here summarized evidences prove. As no doubts about the observations of both the author and Haserodt (1989) and Derbyshire et al. (1984) exists, the following interpretation intergares both observations: their ice margin positions belong to the Late Glacial.

Indications of a thick ice stream in the upper Hunza Valley were found in the intramontane basin of Sost, in the Chapursan Valley (Photo 1 and 2) and further north/east as far as the Kunjerab Pass (see below 2.3.10 and 2.3.13 and Kuhle 1988d).

2.3.8. Reconstruction of the LGM Glaciation in the Nanga Parbat Massif (35°05'-40'N/74°20'-75°E, Fig.1, No.34; Fig.3a)

Details about this massif can be found in Kuhle 1988d; 1996a; 1997, Fig.28; 2001. Ground moraines and polished flanks are preserved on both sides of the upper Rupal Gah

(Valley) above the Toshain Glacier at 4000-4670 m asl (35°10'N/74°27'-30'E). A mountain spur, leading down to the south from the Mazeno crest, has been eroded back to 4680 (or 4700) m by the Ice Age Rupal Glacier. It got the shape of a truncated spur (35°10'40"N/74°32'20"E). Late Glacial lateral moraines of Stadia III-IV (Table), which are preserved in this valley, provide evidence of an ELA depression of 700-800 m, i.e. to ca. 4400-4500 m asl. On the flanks of the Rupal Valley, in the section between Shaigiri- and Bazhin Glacier (Kuhle 1996a Fig.1; 1997 Fig.28), former glacial flank polish margins and moraine remnants with large blocks (35°11'N/74°34'35"E; 35°12N/74°35'30"-74°37'E; 35°12'40"N/74°37'10"E) have been observed. They prove Rupal Glacier levels between 4460 m and 4680 m (Kuhle 1997 Fig.28, No.5-8). Ground moraines and traces of glacier scouring, reaching still further up, are proof of Rupal Glacier levels between 4550 and 4650 m asl between the junction of Bazhin- and Chhungphar Glacier (ibid. Fig. 28, No.9-11). Evidences also have been mapped on the SE side of the Rupal valley flank (ibid. Fig.28, No.19-21). The down-valley joining Chhungphar Gah shows very well preserved glacially polished flanks up to 4600 m asl, so for instance at 35°16'39"N/74°44'40"E, below the Sharsingi Peak (ibid. Fig.28, No.26). This provides evidence of an ice thickness of more than 1400 m (Kuhle 1996a Fig.2; 1997). Further down to 3700 m asl, there are preserved meter- to decameter thick ground moraines. This is also true of the confluence with the Rupal Gah (Valley) (ibid. Fig.28, Nos.27, 28). An E side lateral moraine remnant (ibid. Fig.28, No.29) is located far below of the earlier polish margins of the LGM (see above). Due to its altitude relative to the ELA it probably can be classified as belonging to the Late Glacial Taglung Stadium II (Table).

The described observations in the Rupal Gah (V alley), provide evidence of the Ice Age glaciation of the Nanga Parbat SE face (details see ibid. Fig.28). The observations show, that the Rupal Glacier joined the Astor Valley with increasing thickness. The ice flow directions of the tributary glaciers of the Astor ice stream net, which fringed the Nanga Parbat to the E, are also recorded (Kuhle 1988d, 1997, 2001). We now omit a large number of observations and turn to the junction of the Astor Valley with the largest main valley, the Indus Valley (ibid. Fig.28 Nos.57-64). In the Astor Valley at the junction with the Indus Valley polished flanks are preserved on the NNE side at 2900 m asl (35°33'30"N/74°43'E). Below that altitude there are in places concrete-like remnants of ground moraine (ibid Fig.28 No.58). From here, down to the Indus Valley, these ground moraines cover the NNE lower slopes extensively (35°34'40"N/74°41'30"E). Glacial rock roundings are preserved up to 2800 m asl (Kuhle 1988d; 1991b; 2001). Following the SW side of the Astor Valley from the Indus Valley upward, a decameter thick ground moraine (35°34N/74°40'E) is deposited above the Ramghat Pul (bridge) at ca. 1250-1950 m asl. Further above, at the spur peak Hattu Pir, moraine sediments occur at 3000 m asl. One km upward the Astor Valley glacially polished outcrops of the stratum occur up to almost the top of Hattu Pir. Above the thalweg, the LGP Astor Glacier has attached thick ground moraines to the abraded rock surface at 1000-1200 m asl (35°32'50"N/ 74°40'30"E). Further up this SW side valley flank, continue decameter to more than 100 m thick ground moraines, which are interrupted by glacially polished rock (35°32'30"-32'N/74°40'30"-42'30"E). The ground moraines can be easily diagnosed by funnel erosion in the area of the settlements of Doian and Mang Doian. The highest, i.e. LGM Astor glacier level can be reconstructed with the help of the upper edges of these ground moraines: In in upward direction at ca. 3100-3300 m asl they become lateral moraine remnants at Hattu Pir in the area, where the Astor Glacier joins the Indus Glacier. The classification as LGM is based on the moraines (Kuhle 1996a Fig.3), which, though exposed to erosion, because they stick to the very steep valley flanks, are to a great extent well preserved. The author received absolute 14C-datings only from the historic lateral moraines of the Chhungphar Gah Glacier (Table, StadiaVII-XI). Here, the thickness of the Astor Glacier was 1500-1600 m. This allows to draw the conclusion, that the Indus main glacier was of similar thickness (see Kuhle 1996a). W of the Astor Glacier, the Lichar-, Buldar- and Rakhiot (Tato Gah) Glaciers joined the Indus Glacier. As a representative example, we concentrate on the reconstruction of the Rakhiot Glacier. There are two lateral moraine ledges on the left-hand side of the upper Rakhiot (Tato) Gah, the higher one lies at 4370 m asl, i.e. 370 m above the recent Ganalo Glacier surface (35°19'43"N/74°34'17"E; ibid. Fig.28 No.69). On the orographic right-hand side valley flank (ibid. Fig. 28 No.70) is a corresponding lateral moraine or kame-terrace complex at 4380 m asl (35°19'30"N/74°38'E) ca. 500 m above the present glacier surface. These moraines have been formed at an ELA depression of ca. 400-500 m (actual ELA 4900 m to ca. 4500-4400 m asl) and are to classify as being of the latest stadium of the Late Glacial (IV) (Table) (Kuhle 1996a; 1997 1988d;). Down-valley glacial polishes and abrasions on both valley flanks reach higher than the other moraine remnants, i.e. up to max. 4350 m asl (orographic left-hand side, cf. ibid. Fig.28 No.69, 78, 81; orographic right-hand side, No.72, 73, 76, 77). They provide evidence of a Rakhiot Glacier thickness of ca. 1000-1400 m during the Main Ice Age (Table, Stadium 0). In this context the observation of a ground moraine cover on the 750 x 450 m valley shoulder at the 3822 m high point E of Bezar Gali (ibid. Fig.28 No.81; 35°25'23"N/74°33'48"E) is of importance. The fine ground mass contains isolated large granite blocks, which are round at the edges or round. This ground moraine overlies forms of roches moutonnées and is cut off from any supply with slope debris (Kuhle 1996a; 1997). It is evidence of a Rakhiot Glacier thickness (LGM) in this valley cross profile of at least 1400 m. The ground moraine position is only 8,4 km away from the merely 1150 m asl high Indus Valley thalweg (ibid. Fig.28 No.82). From the valley head below the Nanga Parbat-N-face up to near the junction with the Indus Glacier, the level of the 26 km long Rakhiot (Tato Gah) Glacier has decreased from 4350 m to 3822 m, i.e. merely ca. 500 m. Therefore, an Indus glacier level at ca. 3000 m can be supposed. This is confirmed by the Astor glacier level in its junction area with the Indus Glacier at Hattu Pir (ibid. Fig.28 No.61, 87; see 2.3.7; 2.3.9; 2.3.10) at about 3000-3100 m asl (Kuhle 1996a). Numerous direct observations provide evidence of the thickness of the Ice Age (LGP, LGM) Indus main glacier of ca. 1800 m (cf. ibid. Fig.28 No.82-105). Here are given some examples: on the orographic left-hand side the trough valley flank of the Indus Valley has been polished and abraded by the Indus Glacier up to 3000 m asl (ibid. Fig.28 Nos.84-90, 92). Overlying ground moraines are preserved in numerous, partly decametre thick remnants (ibid. Fig.28 No.83, between Nos. 85 and 86, Nos.82, 101, 102; Kuhle 1996a Fig.4). Ground moraines of metre to more than 100 m thickness, partly up to an extent of 4 km, have been mapped at many localities on the orographic right-hand side flank (ibid. Fig.28, Nos.94-98, 100); there are also glacial flank abrasions (e.g. ibid. Fig.28 between Nos.93 and 94) and roches moutonnées, jutting out of the valley slope (ibid. Fig.28 above Nos.96 and 98) (see Kuhle 1996a). One of these ground moraines is deposited at point 1758 m asl, 620 m above the Indus thalweg (35°34'19"N/74°36'55"E; ibid. Fig.28, No.94); another one up to point 2401 m asl (35°31'10"N/74°35'58"E; ibid. Fig.28 No.95) at the T.P. Gor Gali Peak NE-slope. On the S-slope of this mountain, the ground moraine reaches up to ca. 2650 m asl (ibid. Fig.28 No.97). The most significant Main Ice Age (LGM) Indus Glacier thickness and -width is indicated by the ca. 1 km long lateral moraine ridge, which culminates at 2850 m asl above the Dirkil settlement (35°32'30"N/74°33'11"E; ibid. Fig.28 No. 105). The moraine fringes the "Dead Valley", that is a classic glacier lateral valley. It has developed between the Indus Glacier and the orographic right-hand side Indus Valley flank at 1700-1750 m above the present thalweg. Since the deglaciation it is fossilized, i.e. without water-flow, and therefore bears its name rightly. Haserodt (1989, p. 208) also identified these moraines, but, in contrast to the author (Kuhle 1996a; 1997), understood them as local moraines of a small S-exposed hanging glacier from the Luthi Gal Chamuri flank, down from Terimal. In this case, however, there would be necessary an ELA depression of 1800-1900 m, i.e. 600 m more than the author suggests for the LGM (Kuhle 1988f; 1988d). - Further bank deposits of the Indus Glacier occur at the settlements Gor and Dirkil; so for instance moraines, kames and bank outwash (sandar) (ibid. Fig.28 Nos.100, 103, 104). The highest round-polished mountain ridges reach 3037 m asl at Gor Gali (ibid. Fig.28, No.106). - These observations point to a complete ice stream net glaciation of the Nanga Parbat Massif (cf. ibid. Fig.28) with an ice surface between 3000 and 4850 m, ice thicknesses of 800-1400 m and an ELA depression of at least 1200 m according to a snow line altitude of ca. 3600 m asl during the Last High Glacial Maximum (Table, Stadium 0). On Nanga Parbat the Indus main glacier reached ice thicknesses of even 1800-1900 m (Kuhle 1996a; 1997) (see. 2.3.9 and 2.3.10).

2.3.9. The Southern Slope of the Karakoram (Fig.1, between No.33 and 34)

The present, almost 60 km long glaciers terminate in the catchment area of the Indus Valley at altitudes below 2500-2700 m (Hunza-Karakorum). These are the glaciers that descended furthest in High Asia. According to Ward (1926) the Namche Bawar north glacier (Fig.1, No.13, humid eastern Himalaya) makes a similar steep descent to the Tsangpo Gorge. On Nanga Parbat the glaciers reached altitudes between 2900 and 3600 m asl (Kuhle 1997, Fig.28). During the LGP, however, all the tributary glaciers from the Karakorum south slope and the Nanga Parbat Group united in a 120000 to 180000 km2 (ibid. Fig.28; 3 above) ice stream network, the Indus ice stream network (Kuhle 2001 Figs.2; 2/1; 2/2; see 2.3.7; 2.3.8; 2.3.10). At an altitude of 980 m or lower the largest outlet glacier reached the lowest ice marginal position at Sazin at the mouth of the rivers Daret and Tangir, a little E ward the bend of the Indus (35°34'N/73°28'E, Fig.1, No.35 and Kuhle 1988d). At (35°30'N/73°25'E) the author observed ground-moraines down to 850 m asl (Kuhle 1997).

There are two normally graded lateral moraines, more than 100 m thick (Kuhle 1991b, Photo 3), with a final bend exhibiting the terminus of the moraine (Kuhle 1988d; 1988f). These moraines occur at a distance of about 10 km from one another, each flanking the Indus Valley over several kilometres. The moraines consist of typical glacier diamictites of polymict composition (Kuhle 1991b, Photo 3). Down-valley, the valley cross section becomes a V-shaped valley gorge as a result of former subglacial meltwater erosion. Up-valley from the position of the ice margin a concave, worn-down U-profile is observed that eventually expands to a trough profile with well-preserved abrasions and polishes on the flanks (Kuhle 1991b, Photo 3). Sixty km up-valley, at about 1100 m asl, there is a basin-shaped opening - the "Chilas Chamber". Here, remnants of lateral moraines on the southern bank provide evidence of the LGP ice thickness. NW of the Chilas settlement, glacial abrasion and polishings on the NW side provide evidence of a minimum glacier thickness of 450-550 m. Between Chilas and the lowest moraines, the valley glacier edge during the Late Glacial (Table) can be traced by stable lines of light coloured glacio-limnic bank sediments.

In some places the limnites were dammed bac

Was this article helpful?

0 0

Post a comment