GLACIO-HYDROLOGICAL STUDIES OF THE GANGOTRI GLACIER BASIN, HIMALAYAS

Abstract

Glaciers currently cover about 10.7% of the earth's surface. Total water resources of our planet are 1359x106km3, out of which only 38x106 km3 (2.78%) is available in the form of fresh water. Glaciers and large ice sheets are frozen reservoirs of fresh water. About 75% of the world's total freshwater is stored in the form of glacier ice and out of this about 90% is stored in Antarctica alone. Apart from the Polar Regions, the lofty Himalayas with snow clad peaks are abode to one of the most significant glacier systems in the world and is an important source of fresh water for the perennial rivers originating from the Himalayan mountains. Increased pressures are mounting on the availability of fresh water because of growing requirement due to population growth, urbanization, and ever increasing demand on the finite amount of water for different uses such as drinking, industries, agriculture and hydropower. Hydrological investigations for the glacierized basins are of direct practical utility in the water sector of the country. Recent inventory of glaciers has shown that there are about 5000 glaciers covering an area of about 38,000 km2 area in the Indian territory of Himalayas. Particularly during summer these glaciers contribute significant amounts of fresh water to the flows of the rivers originating from the Himalayas. For example, the Indus, Ganges and Brahmaputra constitute three of the twelve major river basins in the country and the freshwater resource availability in these basins is estimated to be 1757 m3 yr"1 per capita in Indus basin, 1473 m3 yr"1 per capita in Ganges basin and 18,417 m3 yr"1 per capita in Brahmaputra. These estimates are on the lower side with global averages ranging from 1000 m3 yr"1 per capita to over 50,000 m3 yr"1 per capita. Therefore, a collective effort is needed for improved management of water resources of the rivers originating from Himalayan region. Himalayan region is witnessing a large economic growth and demands for hydroelectric power, irrigation and drinking water supply is ever increasing. The provision of appropriate flood control measures in the lowland areas is also of major concern. Such aspects require proper understanding of the hydrological processes of the glacierized basins of the Himalayan region. An accurate assessment and forecasting of the total volume of discharge, melt rate and its distribution in time, weather pattern and sediment transportation is of vital importance for the planning and management of water resources including flood forecasting, reservoir operation and design of hydraulic structures. At the same time proper, precise and periodical information of the snow cover is also necessary for modelling of streamflow and monitoring the health of glaciers. Keeping in view all these aspects, a study has been carried out to understand the hydro-meteorological behaviour and sediment delivery response of Himalayan glacierized basin and to simulate its meltwater runoff for different melt seasons. In addition to this, a detailed study ofsnow cover mapping using satellite images and various geomorphological features developed in the area has also been studied. Further, to develop the necessary concept and understanding, an extensive literature survey has been made on each of these aspects. The review of such studies in Himalayan region shows that very limited work has been carried out for the Himalayan basins. It is also found that remote sensing and GIS techniques have good potential for application in snow and glacier studies. Therefore, it is planned to initiate such study in a Himalayan basin, which has practical importance in the country's water resources sector. For carrying out hydrological, meteorological and sediment transport studies the Gangotri Glacier basin (latitude 30°43'N-31o01'N and longitudes 79°00'E-79°17'E) in the Central Himalayas has been selected for the detailed hydrological investigations. The Gangotri Glacier is one of the largest glaciers of the Himalayas. The total catchment area of the study basin up to the discharge gauging site is about 556 km2, out of which about 286 km2 is covered by snow and ice. Bhagirathi River sprouts from the terminus of Gangotri Glacier. In India, this terminus is known as Gaumukh. After confluence of Alaknanda River with Bhagirathi River at Devprayag, the combined river is known as Ganges River. A meteorological observatory was established at about 3 km downstream to the terminus of the glacier. Rainfall, temperature, wind speed, wind direction, humidity, sunshine hours and evaporation data were collected. Also a discharge gauging site was established adjacent to the meteorological observatory and automatic water level recorder was installed for round the clock monitoring of discharge. Samples of suspended sediment were also collected at this gauging site and concentration of suspended sediment was determined The geomorphological features studied include various types of erosional, depositional and paraglacial landforms developed in the area using field survey as well as satellite imageries. These features include horn; u-shaped valley; roches mountonnees; supraglacial, englacial and proglacial lakes; supraglacial lateral and medial moraines, morainic ridges, tillite hillocks, paraglacial fans, glaciofluvial terraces and outwash plain. Digital Elevation Model (DEM) has been prepared from the topographic maps to represent the spatial variation of elevation. A detailed mapping of glacier body from satellite images has also been carried out and further used as an input to the computation of meltwater runoff model. Glacio-hydrological studies require emphasis on meteorological aspects as well because these parameters control the melting pattern of glacier. Meteorological data (viz. rainfall and temperature) observed near the terminus of the glacier has been correlated with the discharge and suspended sediment data. Further, these data provide the base for the simulation of meltwater runoff as they are used as direct input in the model. Meteorological analysis shows that the distribution and amount of rainfall varied significantly from year to year. On an average melt season rainfall (May-October) was about 260 mm, consisting of mostly drizzle type rainfall. Daily rainfall hardly exceeded 15 mm. Mean monthly temperatures for May, June, July, August, September and October were 8.9, 10.3, 11.7, 10.8, 7.7 and 5.3°C, respectively, suggesting that July was the warmest month. Average daily maximum and minimum temperatures over the melt season was observed to be 14.8°C and 4.1°C, respectively. Day-time wind speed was observed to be about 4 times stronger than the night-time wind speed. The relative humidity was observed to be about 69% in the beginning and about 85% towards the end. Mean daily sunshine hours were 5.5 hours. Monthly pan evaporation is 150.7, 113.4, 106.9, 85.5, 87.7 and 96.6 mm for the months of May, June, July, August, September and October, respectively. The total pan evaporation during the four successive melt periods varied between 594 and 680 mm. Over all meteorological records represented dry weather conditions in the study area. The main sources of runoff from any glacierized basin are melting of ice and snow, and rainfall. In the study area, the rainfall is very less and, therefore, streamflow is highly dominated by the melting of ice and snow. The proper distribution of such contribution in annual flow of the river is very important for planning and development of water resources. Hydrological observations show that mean daily discharge ranged between 5 to 194 m3 s"1. The mean monthly discharge observed for May, June, July, August, September and October was 27.3, 74.1, 121.8, 105.7, 60.7 and 22.2 nvV. The distribution of observed runoff indicates maximum discharge was observed in July followed by August. The months of July and August contributed about 56% to the total melt season discharge. Meltwater storage characteristics of the glacier have direct impact on the distribution of runoff. The storage characteristics ofthe glaciers are responsible for delayed response of meltwater generated over the glacier surface into runoff. Moreover, storage and drainage behaviour, diurnal variation in melt rate and melt runoff delaying characteristics of the glacier have also been studied. The comparable magnitude of runoff observed during day-time and night-time shows strong storage characteristics of the Gangotri Glacier. A detailed analysis of diurnal variations in discharge for selected clear weather days during 2001 illustrates that discharge starts rising in June, attains its maximum in July and then starts declining. Further, an attempt has been made to develop a relationship between discharge and temperature, which shows good correlation between them. This characteristic typically reflects that melting from the Gangotri Glacier is mostly temperature driven. Flow duration curve analysis for the study basin has also been carried out. Several hydropower projects in the Himalayan region are established over streams originating from the high altitude glacierized basins. These projects are facing severe problem due to high sediment transportation from these basins. The sediment generated from these basins is deposited in the reservoirs constructed downstream, thus reducing their storage capacity and restricting the life of the reservoirs for which they are designed. Thus, estimates of suspended sediment transport becomes essential for the management and operation ofsuch reservoirs. Suspended sediment analysis includes quantification and variations in suspended sediment concentration (SSC), suspended sediment load (SSL), sediment yield, sediment erosion rate and particle size distribution of the glacier. Mean monthly suspended sediment concentration for May, June, July, August, September and October during the study period is 1942, 2063, 3658, 2551, 734 and 136 mg I"1, respectively, while mean monthly total suspended sediment load for corresponding months has been found to be 149, 423, 1220, 746, 143 and 5x103 tonnes, respectively. SSL is found to be more variable (Cv =1.1) than SSC (Cv =0.8). The delivery patterns of discharge and SSC/SSL have been studied and found that delivery of SSC is always earlier than the discharge. The delivery response of SSL shows that it is slower in the beginning of the season than the later part in comparison to discharge. The sediment yield from the basin is found to be very high (4834 tonnes km"2). Attempts were made to investigate the hysteresis pattern between discharge and SSC, and correlation of suspended sediment with discharge, temperature and rain. Arating curve between SSC and discharge has also been established. Particle size distribution shows that suspended sediment contains predominately silt sized grains (0.002 - 0.060 mm). Keeping in view the availability of limited field information on the extent of snow/ice cover and its practical importance in hydrological studies, an attempt has been made to find out the variation in snow cover area over the melt season using IRS-1C/1D-LISS-III multispectral data. Due to some limitations in satellite sensor, an accurate assessment snow cover area requires proper methodology. Pre processing of the data includes conversion of DN values into reflectance values. Normalized difference snow index (NDSI) method has been used to distinguish snow from non-snow areas. This method is useful as it distinguishes snow from cloud and can easily map snow cover under mountain shadow. The study developed a new methodology to distinguish dry/wet snow from optical sensor data. In spite of the fact that wet snow and exposed glacial ice contributes maximum to the river runoff, no such methodology is available to distinguish them from other features using optical data. Moreover, four different water-bearing zones in the glacierized basin have been mapped, namely, dry snow zone, wet snow zone, exposed glacial ice, and moraine-covered glacial ice. All the four zones posses different hydrological behaviour. A comparison has been made between dry snow area and non-melting area obtained from the different values of lapse rate. Finally, it is found that non-melting snow area computed using a lapse rate of 0.60°C/100 m matches well with the dry snow cover area derived from remote sensing data. Therefore, this value of lapse rate has been used for the modelling of runoff from the Gangotri Glacier basin.In order to simulate the streamflow, the conceptual snowmelt model (SNOWMOD) has been used. Using the basic concepts of the model some changes have been made according to the requirement of the glacierized basin. The model is based on temperature index or degree-day approach. For simulation purpose the basin was divided into nine elevation zones. The primary input data included temperature, rainfall, glacierized area. In spite of very less rainfall in the study area, the model computes and routes both meltwater and rainfall-runoff separately using cascade reservoir approach. The model was calibrated using the data sets of 2000 and 2001 and simulated for 2002 and 2003 using those calibrated parameters. The model simulated daily streamflow satisfactorily for all the years (2000-2003) (R2 = 0.96). The average difference in computed and observed volume of discharge was about -2.5%. The model also computed the contribution of glaciermelt and rain in runoff. It is found that on the seasonal scale the contribution of glaciermelt to the river runoff is about 97%, whereas the contribution of rain to runoff is only 3%. In view of the technical and economical constraints in getting periodic information on snow cover area, the streamflow simulations were performed using constant value of glacierized area for each zone derived from the toposheets. However, an attempt has been made to simulate the streamflow using remote sensing derived snow cover area (SCA) and glacier cover area (GCA) for the year 2000. The values of all other parameters have been kept similar to the abovecalibrated model. Although in this case also simulated streamflow were well reproduced (R2 =0.94), but the efficiency was reduced as compared with previous simulations. Overall it is found that model overestimated the runoff in comparison to observed discharge with Dv = 13.11%. It is understood that for such analysis, periodical satellite mapping of the seasonal snow cover is needed. These hydrological investigations play an important role in the Himalayan water resources planning, development and management in general and for the Bhagirathi/Ganges River Basin in particular. Considering the hydrological importance of the region it can be said that similar studies for other basins located in this region could be carried out using the outline developed in this study.