HYDROLOGICAL RESPONSE OF DUNAGIRI GLACIER STREAMFLOW (ALAKNANDA RIVER BASIN) TOWARDS CHANGING CLIMATE, CENTRAL HIMALAYA (UTTARAKHAND), INDIA

Abstract

Glaciers are considered as the fresh water resource and the lifeline of majority of the rivers on the Earth. Existence of glaciers plays a major role in controlling regional climate and consequently considered as the natural coolers of Earth by maintaining the temperature and energy balance (Raina, 2008). The global warming and climate change are the major challenges in today’s era. Effects of increasing temperatures are causing number of dreadful disasters such as storms, heat waves, floods, droughts, etc. Due to global warming and climate change, glaciers are melting rapidly all over the world and also contributing in disaster events in form of sea level rising, glacier lakes outburst floods, avalanches, etc. More melting of glaciers means more exposure of Earth surface to direct sunlight. This will result in lower albedo over the Earth surface and causes more absorption of input energy. This will have a direct impact on the water resource management and planning. Health of the glaciers has direct relationship with fluctuations in climate. Therefore, to understand the influence of climate change on water resources, it is necessary to identify the response of glaciers towards climate forcing. The Himalayan glaciers are the source of the ten important river systems of Asia viz. the Amu Darya, Tarim, Indus, Ganga, Brahmaputra, Irrawaddy, Salween, Mekong, Yangtze, and Yellow (Bajracharya and Shrestha, 2011). These Himalayan rivers are the water lifeline for almost 1.3 billion people of South Asia or about 10 percent total regional human population (IPCC, 2007, Bolch, 2012). The Indian part of Himalaya contains 9,575 glaciers (Raina and Srivastava, 2008). These glaciers provide great support to the local population by maintaining flow of rivers throughout the year. The three major Indian rivers viz. Indus, Ganga and Brahmaputra depend on the snow/glacier melt to maintain their flow during the summer. The contribution of glacier melt is high during the non-monsoon season. But changing climate has increased the rate of glacier melting which will be threaten to long term existence of these glaciers in future. The present study aims to understand the mechanism of hydrology generation processes of a glacierized basin and response of climate change to the streamflow using hydrological model. The streamflow generated from a glacierized basin have three major components – i) contribution of rain, ii) contribution of Snow and iii) contribution of glacier melt. To study the variation in streamflow of the catchment, better understanding of the climatic (temperature, rainfall, evaporation, radiative fluxes), non-climatic ( vii ) parameters (topography aspect, glacierized and non-glacierized area) and hydrological parameters (snow/ice melting, streamflow fluctuation) are necessary. Keeping this in view, study on the glacier evolution, snow to ice transformation (glacier formation), classification of glaciers and glacial geomorphology is carried out to the understand the Glaciology. A comprehensive review on the Glacier-Climate Interaction, Indian Himalayan Glacier System, Major River System and Climatic Conditions of Uttarakhand, Hydro-Meteorological Measurements in glacierized catchment (Methods and Instruments), Hydrological Modelling of Streamflow and available Hydro-Meteorological studies, was carried out with special reference to Indian Himalaya. A detailed description of standard semi-distributed conceptual hydrological model (HBV) is presented to understand the working principle and model routines (Snow routine, glacier routine, soil moisture routine, response routine and routing routine) for streamflow generation processes in the glacierized catchment. To study the entire glacier in Indian Himalaya by ground survey or instrumental methods is difficult due to rugged terrain and harsh weather conditions. In the present study, the Dunagiri Glacier has been selected for studying the response of glacier towards climate forcing. This glacier is a north facing valley type glacier having an area of 2.56 km2 and lies in the Dhauliganga River basin (Central Himalaya, Uttarakhand). Meltwater generated from Dunagiri Glacier contribute to Dhauliganga River which is a major tributary of Alaknanda River. The total area of Dunagiri Glacier catchment up to the meteorological station is 17.9 km2 having 14.3 % glacierized area. The total extension of catchment is from ~3800 m a.s.l to ~6400 m a.s.l. The Dunagiri Glacier was firstly studied in 1984 by the Geological Survey of India (Srivastava, 1992). The glacier was studied in term of hydro-meteorology, snout fluctuation, mass-balance during the ablation seasons of 1984 to 1989. A hydro-meteorological observatory was established in 1984 near the glacier snout (3800 m a.s.l.) to collect the hydrological and meteorological records during the summer season. Available records of hydro-meteorology from 1984 to 1989 are used for hydrological modelling purpose. The average temperature in Dunagiri Glacier catchment ranges between 5.9°C to 13.7oC (Table 4.2). The average temperature was observed 17.7°C, 11.3°C, 11.9°C, 10.7°C and 9.7°C for the years of 1984, 1985, 1987, 1988 and 1989, respectively. August month records maximum precipitation of 40.9, 81.7, 134.5 and 142.1 mm for all the year except 1985 in which September month record maximum precipitation of 87.6 ( viii ) mm (Table 4.2 and Fig. 4.4). 1988 is wettest year with maximum amount of rainfall 269.3 mm. Average value of Rh is 64.2, 87.5, 82.6, 86.9 and 82.9 % & wind speed value is 7.9, 4.9, 5.2, 3.5 and 4.8 kmph for the year 1984, 1985, 1987, 1987, 1988 and 1989 respectively (Table 4.2 and Fig. 4.4). Average discharge during the ablation season of 1985, 1987, 1988 and 1989 is1.7, 2.5, 2.3 and 1.9 m3 s-1, respectively (Table 4.2 and Fig. 4.4). A semi-distributed conceptual hydrological model (HBV) is used to study the streamflow characteristic of the Dunagiri Glacier. The model consists of five different routines like Snow, Glacier, Soil Moisture, Response Runoff and Routing routines. The selection of routines is based on the characteristic of the catchment. Since, the studied catchment consists of snow and glacier, therefore, all five routines have been selected for streamflow generation of Dunagiri Glacier. The model runs on the daily meteorological parameters (precipitation, temperature and evaporation) of the catchment. The meteorological data collected near the snout of the Dunagiri Glacier during the ablation seasons of 1985, 1987, 1988 and 1989 are used to run the model. The glacier catchment (3800 m a.s.l to 6400 m a.s.l.) is divided into 13 elevation bands - E1 (3800-4000 m a.s.l.) to E13 (6200-6400 m a.s.l.), with an interval of 200 m for the distribute application of the hydrological model. The model interpolates precipitation and temperature for each elevation band separately. Similarly, streamflow simulations for the glacierized and nonglacierized zones of the catchment are performed for each elevation band separately. Result of interpolated temperature and precipitation records suggest that temperature in the glacier catchment varies between 7.49°C to -6.91°C, 7.18 to -7.22°C, 6.90 to -7.50°C and 6.25 to -8.15°C for the years of 1985, 1987, 1988 and 1989, respectively. The month of July is the warmest month for all the years followed by August and September. Variation of precipitation is 3.01 to 9.02 mm/d for 1985, 2.34 to 7.02 mm/d for 1987, 4.20 to 12.60 mm/d for 1988 and 3.74 to 11.23 mm/d for 1989. Mixed form of precipitation (rain and snow) is calculated in the catchment for the study period. Generally, precipitation above 5000 m a.s.l. (ELA of glacier during the study period) is calculated in the form of snow due to continuous temperature records of ≤ 0°C and in the form of rain below the 5000 m a.s.l. up to positive temperature records throughout the study period. The variation of soil moisture (SM) is only in the nonglacierized area due to continuous movement of water through infiltration and percolation. This variation shows a decreasing pattern with elevation due to presence of ( ix ) snow at higher elevations. The evapotranspiration only varies with temperature and its value becomes zero when there is negative temperature in a particular band. The model parameters are calibrated for the available dataset of 1985 and 1987 and further validated using the calibrated parameters for the period of 1988 and 1989. The Dunagiri Gad streamflow is calculated for the periods of 1985, 1987, 1988 and 1989 and results are compared with observed data set. The simulated streamflow of Dunagiri Glacier has three major components – Rain, Snow and Glacier melt. During the months of July and August, the rainfall contribution is high. The contribution of snowmelt is very less (4-6%) to the streamflow as compared to rain and glacier melt (Fig. 5.1, 5.2, 5.3, 5.4 and 5.6), because of nonavailability of winter snow cover data. The glacier melt contribution to the total streamflow majorly depends on the contrasting meteorological conditions (temperature, precipitation and radiative fluxes) and physical characteristics of the catchment. Contribution of glacier melt to the streamflow varies between 30 % and 50 % (Table 5.2 and Fig. 5.6). The glacier melt contribution was highest for the year of 1985 (43.7 %) followed by 1987 (43.3 %), 1988 (32.2%) and 1989 (30.1%). The melting pattern increases up to elevation 4600–4800 m a.s.l. band and then starts decreasing (Fig. 5.7). It is observed that melting of the glacier in upper elevation band (4400–4600 m a.s.l.) more as compared to lower bands (4200–4400 and 4400–4600 m a.s.l.). The possible reason may be the presence of large numbers of supraglacial lakes in this elevation bands which influence the glacier surface melting in a great extent. Results of the study showed that the model efficiency (Reff) is 0.76 and 0.70 for the period of 1985 and 1987 with coefficient of determination (R2) as 0.68 and 0.81, respectively during the calibration period. Similarly, the model efficiency is calculated to be 0.67 and 0.74 for the period of 1988 and 1989 with coefficient of determination of 0.60 and 0.72, respectively for the validation period. The simulated streamflow showed a good agreement with the observed streamflow which suggested that all important processes are moderately included in the HBV model.