Introduction

Changme-Khangpu glacier is located in the Tista river basin, in the upper catchment of one of its major tributaries, Lachung Chu, in the Sebu valley of north‒east Sikkim. It is a transverse valley glacier trending north‒south, having a length of 5.87 km and width varying between 600 m to 1 km. The glacier originates from the southern slope of Gurudongmar peak (lat. 27°58’N., long. 88°42’E.) which is an integral part of the Khangchengyao‒Pauhnri massif. The melt water of the glacier feeds into the Sebu Chu, a tributary of Lachung Chu.

The areal coverage of the glacierized part of the Changme-Khangpu basin is about 12.85 km², of which 10.35 km² comes under perennial ice cover and the rest is covered by rock surfaces only. The ratio between the accumulation and ablation zone areas is 5:7. Almost the entire ablation zone of the glacier is under a thick mantle of supraglacial morainic debris (Reference BhattacharyaBhattacharya, unpublished). The terminal part of the glacier is completely buried under the moraine which has generated enormous amounts of dust which is present everywhere. Presence of englacial dust is also well documented in the darker ice bands exposed in the vertical to sub-vertical faces. The lowest part of the snout is located at an altitude of 4850 m a.s.l. This part, when exposed from the supraglacial debris, is about 3 m in height and 7 m in width.

The Geological Survey of India has been studying the annual mass balance on this glacier since 1977. In order to obtain the flow rates of ice we have dated the snout using the isotope ³²Si produced by cosmic rays (Reference LalLai and Peters, 1962). In addition, we have measured the longitudinal and vertical profiles of ²¹⁰Pb. Nuclear debris, mainly from the Chinese nuclear tests, was detected in fresh snow (Reference Bhandari, Bhandari, Nijampurkar, Shukla and PuriBhandari and others, 1982) and its vertical profile in the accumulation zone enabled us to determine the net accumulation rate of Ice in this glacier to be 70 cm/year (Reference Shukla, Shukla, Bhandari, Nijampurkar, Kameswara Rao, Puri and SharmaShukla and others, 1983). The results are discussed here in terms of glacier dynamics and its past history as determined from the geomorphological studies

Experimental Techniques

Three types of samples were collected for the present study. Samples of snout ice (2.38 tonnes) and of snout water (1.64 tonnes) were collected and processed for ³²Si and ²¹⁰Pb (Table I). Small samples of surface ice were also collected from some locations for ²¹⁰Pb, and δ¹⁸0 analysis (Fig. 1). A core was taken at 5040 m altitude and one metre sections of ice were melted in plastic bottles for analysis of ^2l0 Pb, total dissolved solids, total beta activity, some chemical constituents, and isotopic ratio of oxygen. On the night of 28 August 1978, snow precipitation occurred all over the glacier. Samples (nominally 1ℓ) (Table. I and fig.1) of this fresh snow were collected for ²¹⁰Pb and δ¹⁸0 studies, Fig. 2. Here we only discuss ²¹⁰Pb and ³²Si data. The other data relating to nuclear debris and δ¹⁸0 have been reported elsewhere (Reference Shukla, Shukla, Bhandari, Nijampurkar, Kameswara Rao, Puri and SharmaShukla and others, 1983; Reference Bhandari, Bhandari, Nijampurkar and VohraBhandari and others, 1983; Reference Nijampurkar and BhandariNijampurkar and Bhandari, in press).

Table I. Experimental Data on ³²Si Measurements in Snout Samples From Changme-Khangpu Glacier, Sikkim

Fig.1. Distribution of ³²Si and ²¹⁰Pb activity in surface samples of Changme-Khangpu glacier. The sample distance from the accumulation zone has been determined from point A which is the centre of the accumulation zone.

Fig.2. Distribution of ²¹⁰ Pb activity (d.p.m./ℓ) in fresh snow samples on Changme-Khangpu glacier. The sample distance from the accumulation zone has been determined from point A which is the centre of the accumulation zone.

The chemical procedures and techniques of measurement are similar to those described by Reference Nijampurkar, V. N., Bhandari, Vohra and KrishnanNijampurkar and others (1982). ³²Si was estimated by analysis of its daughter ³² P and ²¹⁰ Pb by its daughter ²¹⁰Bi. The daughters were radiochemically purified, counted on low-background Geiger-Müller counters, and their decay was followed for several half-lives. The counting efficiencies of ³² P and ²¹⁰ Bi were around 35%. The chemical efficiency of extraction of phosphorus was about 50% and of bismuth about 80% except for a few cases of low yield of up to 20%. Replicate measurements made in some samples gave consistent activity.

The results of ³²Si measurements are given in Table I. Measurements of ³² P in a second extraction in both snout ice and snout water samples agree within the errors of measurements. In both the cases the silica recovered was more than expected from the content of dissolved silica as estimated from spectro-photometry. It was therefore not possible to determine the actual chemical extraction efficiency of silica. Based on our experience in analysis of similar samples, the efficiency of extraction of silica is assumed to be 95%.

Results and Discussions

The ages of various ice samples can be calculated from the observed activity, given in Tables I and II, using the radioactive decay equation, the values of their half-lives, and their fall-out values, if it is assumed that the change of activity is entirely due to decay. The average fall-out values in the past are difficult to estimate, although measurements over the past two decades are available at several geographic locations (Reference ElmoreGeorgieva and Dimchev, 1977; Reference Lal, Lal, Nijampurkar, Rajagopalan and Somaya-juluLal and others, 1979; and Reference Nijampurkar, V. N., Bhandari, Vohra and KrishnanNijampurkar and others, 1982). The ²¹⁰ Pb values in the precipitation of 28 August 1978 suggest an increase in activity with altitude (Fig. 3) except for one value of 2.9 d.p.m./ℓ at 5400 m altitude which does not fit this pattern. An increase of ²¹⁰ Pb with altitude was also observed by Reference ElmoreGeorgieva and Dimchev (1977). The measurements in other snow samples collected during 1981 at 5200 m yield a value of 8.7 d.p.m./ℓ for ²¹⁰Pb and 0.8 d.p.m./tonne for ³²Si. Taking these observations into account and based on arguments given earlier (Reference Nijampurkar, V. N., Bhandari, Vohra and KrishnanNijampurkar and others, 1982; Reference Bhandari, Bhandari, Bhatt, Nijampurkar, Singh, Srivatsava and VohraBhandari and others,1981) (fig.2 and fig.3.), we assume the fall-out values to be 8 d.p.m./ℓ for ²¹⁰Pb arid 0.7 d.p.m,/tonne for³²Si for the purpose of calculating ice ages. Probable errors in these values assumed for fall-out do not significantly alter the conclusions regarding ages of ice samples or flow rates discussed later on. An uncertainty of a factor of two in the fall-out value of ²¹⁰Pb changes the age of the snout by only 22 years whereas a 20% uncertainty in fall-out value of ³²Si changes the age by about 50 years. Thus in spite of the fact that the fall-out value may be changing in each precipitation and that there is possibly some uncertainty in the annual fallout, the chronology derived here should be qualitatively correct. The half-life values for ³²Si measured recently by Reference ElmoreElmore and others (1980) and Reference Kutschera and KutscheraKutschera and others (1980) are 108 ± 18 years, and 101 ± 18 years whereas so far a value of 300 years has been used, based on geophysical arguments given by Reference ClausenClausen (1973) and Reference DemasterDemaster (1980). We here adopt a value of 105 years for ³²Si and the value for ²¹⁰Pb is 22.3 years.

Fig.3. Altitude dependence of ²¹⁰Pb activity in precipitation (fresh snow) of 28 August, 1978.

The apparent ages of about 100 years, calculated from ³² Si and ²¹⁰ Pb activity in snout ice are consistent. Snout water, on the other hand, shows younger age of 21 ± 13 years based on ³² Si and of 59 ± 4 years based on ²¹⁰ Pb which may agree with each other within two standard deviations. Younger ages for snout water as compared to ice are of course expected since the melt water contains contributions from all over the glacier. Significantly young ages, as observed here, indicate a major contribution to the melt water from the upper reaches of the glacier, i.e. from recent precipitations. There are, however, many complications as far as the ”closed box” assumption is concerned. The core samples (Table III) show ²¹⁰ Pb activity of 0.5 d. p.m./l in all horizons except between 3 and 4 m and again between 11 and 12 m. The value of 7.5 d.p.m./ℓ at 3 to 4 m is much higher than expected as it yields an apparent age of 10 years and is overlain by very old ice (>100 years). Such inversions have also been encountered in Neh-nar glacier in two core samples collected during 1977 and 1978 (Reference Nijampurkar, V. N., Bhandari, Vohra and KrishnanNijampurkar and others, 1982; Reference Bhandari, Bhandari, Bhatt, Nijampurkar, Singh, Srivatsava and VohraBhandari and others, 1981). The enrichment of ²¹⁰ Pb can occur in many ways, from nuclear weapons testing, due to absorption on dust or due to chemical exchange or contribution from the in situ radiogenic ²¹⁰ Pb present in the dust. Percolation of melt water can also play an important role in modifying ²¹⁰ Pb concentration, particularly in temperate glaciers (Reference Glen, Glen, Homer and ParenGlen and others, 1977). None of these sources can, however, completely explain all the observations (Reference Shukla, Shukla, Bhandari, Nijampurkar, Kameswara Rao, Puri and SharmaShukla and others, 1983). In spite of these possible uncertainties, the observed longitudinal profile of ²¹⁰ Pb in surface ice shows a systematic decrease with decreasing altitude (Fig. 4), and can be attributed to decay of ²¹⁰ Pb with time. For ³² Si it is known that such processes are much less important (Reference NijampurkarNijampurkar, unpublished) and can probably be ignored. The ³² Si ages should therefore be more reliable.

Fig.4. Longitudinal profile of ²¹⁰ Pk in surface ice samples. ²¹⁰ Pb activity (d.p.m./ℓ) is plotted against (km) from the top of the accumulation zone. Point A is the mean value for the fresh snow (fallout) based on several samples and is expected to be uniform through out the accumulation zone (dotted line).

Table II. Experimental Data on ²¹⁰ Pb Measurements in Surface and Fresh Snow Samples From Changme-Khangpu Glacier, 1978

Flow Rates

The ²¹⁰ Pb ages in ice, subject to the corrections for other processes, can be used to determine the flow rates of the glacier in various regions. The three measurements at 5150 m, 4950 m, and 4800 m yield apparent average flow rates of 80 m/year and 20 m/year between the two altitude intervals respectively. The ³² Si and ²¹⁰ Pb data in surface ice from the snout yield an average flow rate from the accumulation zone to the terminus of 40 m/year for the Changme-Khangpu glacier.

These values can be compared with contemporary flow rates which have been determined by the Geological Survey of India and the Survey of India (Reference BhattacharyaGhosh and Sengupta, unpublished; personal communication from R. Nahak in 1980) for several years during summer. These velocities vary between 1 m/year around altitude of 4850 m to 58 m/year at the equilibrium line, at an altitude of 5250 m, and decrease again towards the accumulation zone, being 40 m/year at an altitude of 5300 m. No measurements have been made above this altitude but flow rates are expected to be low in the accumulation zone. These contours give an average area-weighted flow rate of 13 m/year, which can be treated as an upper limit considering that the accumulation zone is not included in this analysis. This value is much less than the 40 m/year obtained by radiometric methods, which is averaged over the mean life of ³² Si, i.e. about a century.

Table III. Experimental Data on ²¹⁰ Pb Measurements in Surface and Fresh Snow Samples From Changme-Khangpu Glacier, 1978

The discrepancy is probably significant enough to suggest that contemporary and past flow rates are different indicating that the glacier is not in a steady state at present. We therefore examine the geomorphic observations to see if this can give some clues to the past history of the glacier. This is particularly important since the glaciers of this valley are, in general, in the process of recession at present.

Geomorphic Evidence for Glacial Activity of Changme-Khangpu Glacier

The available glacier geomorphic evidence indicates that this part of the higher Himalayan morphogene had undergone multiple glacigenic episodes. Four, if not more, glacial phases and inter-phases are well documented from the multiple landscape around Changme-Khangpu glacier (Reference BhattacharyaBhattacharya, unpublished; Reference GhoshGhosh and Sengupta, unpublished) though the glacigenic landforms of the valley are generally obliterated or covered by the consequent geomorphic processes.

During the earliest advancing phase, Changme-Khangpu glacier, the largest of the ten glaciers in the Sebu valley, possibly had occupied the entire valley width and had advanced up to the Dongkyachu valley, where it coalesced with the trunk glacier. This advance probably had resulted in erosion of the roches moutonées which already existed at Yume Samdong along the course of the trunk glacier. The old tillites lying over these roches moutonées were likely to be the medial moraines separating the trunk glacier (Dongkya glacier) from the tributary glacier (Changme-Khangpu). Glacially eroded valley walls and the unpaired benches that one sees here nearly 200 m to 300 m higher up from the present valley floor are evidently due to this glacigenic episode. In the following advance, Changme-Khangpu glacier did not advance far enough to cover the entire valley width and length. However, a glacier advance at least up to the hot springs about 1 km down-stream of the present snout position could well be postulated. Lower level terraces noted between 100 m and 150 m higher than the present valley base and a set of dissected tillites are apparently remnants of old end moraines and represent this glacial phase. During the third advance, available field evidence indicates that the glacier occupied the western part of the upper reaches of the valley. As a result, a glacial trench developed between the eastern margin of the glacier and the eastern valley wall. During this advance of the glacier, however, the glacier abutted against its southern valley wall and was diverted to the south-east. The glacially eroded western valley wall of the Sebu valley as well as the swerving latero-terminal moraine ridges of Changme-Khangpu glacier corroborate this observation. During these phases this glacier evidently still acted as a barrier to the advance of other glaciers like Sebu or Changme-Khang. As a result, these glaciers during this advance either coalesced or abutted against Changme-Khangpu glacier.

In the recent past, the glacier had another advance which could be deciphered from the evidence still preserved nearly 300 m down-stream of the present terminus. The exact extent of the advance could not be pinpointed as much of the glacigenic deposits are eroded away by the consequent fluvio-glacial activities. Similar phenomena are also noted in some other glaciers in the valley. The maximum extension of Sebu glacier in the recent past is well indicated by its end moraines. Within the depression between the present position of this glacier and the end moraines the lake Sebu Chho is located. A small hanging glacier located to the south of Changme-Khangpu glacier also demonstrates clearly the advance that this glacier had in the recent past. In the case of Changme-Khangpu glacier the advance in recent time is noted from the half-truncated end-moraine deposits located nearly 750 m from the present ice limit of the glacier. Due to the rather rapid recession, this glacier has already transformed into a cirque glacier with no outflow (tongue) at all.

The smaller present-day surface velocity is therefore consistent with the recession of the glacier and the high velocity over the past century is probably related to the previous advancing phase. The radioisotopic tracers thus provide a method of determining the past behaviour (advance or recession) of glaciers. Using isotopes with different half-lives one can construct the past history of glaciers.