With an average annual discharge of 19,830 m3/s, River Brahmaputra ranks fourth among the large rivers of the world (Goswami, 1998). The river is a trans-Himalayan river as it rises in south Tibet from the glaciers of Mount Kailash at an elevation of about 5150m above sea level (a.s.l) at 30°31´N and 82°10´E (Burred et al., 1934; Dhar and Nandargi, 2000). The Brahmaputra River basin with an approximate area of 651,334 sq.km., covers four countries:
China (50.5%), India (33.6%), Bangladesh (8.1%) and Bhutan (7.8%) (IUCN, 2014). Its basin in India is shared by Arunachal Pradesh (41.88%), Assam (36.33%), Nagaland (5.57%), Meghalaya (6.10%), Sikim (3.75%) and West Bengal (6.47%) (Singh et al., 2004). The basin lies between 23°N to 32°N latitude and 82°E to 97°5´E longitude. The Brahmaputra basin is highly influenced by monsoon rainfall (Mirza, 2002). Most part of the runoff of the Brahmaputra river is because of the heavy rainfall of 510 cm-640 cm in the Abor and Mishmi hills in Arunachal Pradesh and 250-510 cm in the Brahmaputra plains (Murthy, 1981).
The rivers originating from the Himalayas experience two high-water seasons, one in early summer caused by snowmelt in the mountains, and one in late summer caused by runoff from monsoon rains. Thus, apart from heavy rainfall, the streamflow of the Brahmaputra River is also strongly influenced by the snow and ice melting in the upstream parts of the basin. The Brahmaputra River receives a substantial amount of snowmelt runoff into its annual streamflow (Singh et al., 1994a, 1997a). Eastern Himalaya ranges are the main source of glaciers that contribute to the discharge of Brahmaputra River. The elevation of the Tibetan plateau that falls under Brahmaputra basin varies between 3000-5000m a.s.l. with numerous glaciers covering the plateau (Singh et al., 2004). According to Immerzeel et al., (2008), Tibetan Plateau alone covers 44.4% of the total basin area, thus contributing more to the snowmelt runoff of the river. The snow and glacier melt contribution compared to total runoff has been estimated to be about 27% for Brahmaputra River (Immerzeel et al., 2010).
Global climate change has substantially increased the atmospheric concentration of carbon dioxide and other trace gases over the last century. It is expected thatthe concentration of carbon dioxide will be double by the middle or latter part of the next century (NAs, 1979;
Pearman, 1980; IPCC, 1995; Houghton et al., 2001; Kamga, 2001). It is obvious that, global warming caused by increased atmospheric concentration of carbon dioxide and other trace gases will alter the radiation balance of the atmosphere. This in turn will cause increases in temperature and changes in precipitation pattern and other climatic variables (Changchun et al., 2008). Higher greenhouse gas concentrations have trapped more thermal radiation and consequently warmed the planet (IPCC, 2007). In its Fifth Assessment Report, the Intergovernmental Panel on Climate Change reported that averaged combined land and sea surface temperature warmed between 0.65°C and 1.06°C from 1880 to 2012 (IPCC, 2013).
As reported by the NASA Goddard Institute for Space Studies in January 2013, 2012 was the ninth-warmest year since 1880, and except for 1988, the nine warmest years in the 132-year record have all been since the year 2000. The year 2005 was the hottest year on record (GISS, 2013). Again, the precipitation is expected to increase under global warming at high latitudes and in the vicinity of the equator, but decreases in the subtropics (Watterson and Whetton, 2011). The GCMs can be used to simulate the present day and projected future climate conditions under different scenarios. However, the GCMs perform reasonably well at large spatial scales and perform poorly at the smaller space and time scales relevant to local and regional impact analyses. Various downscaling techniques have been developed to overcome this limitation of the GCMs. These are dynamic downscaling, statistical downscaling and downscaling using Artificial Neural Network. The statistical downscaling technique has been used extensively because of its less computational time.
Snow cover which is a crucial hydrological parameter in global water cycle and an important ingredient of the cryosphere, is considered as an active and multivariate part of the climate system. Snow is a sensitive indicator of climate (Foster et al., 1982; Namias, 1985; Gleick, 1987) and hence, seasonal snow is considered as an important part of Earth’s climate system.
In terms of area, snow cover is the single largest component of the cryosphere, covering an average of46 million sq.km. of the Earth’s surface each year and out of this, about 98 percent of the Earth’s snow cover is located in the Northern Hemisphere (Brodzic and Armstrong, 2013). Because snow is highly reflective in nature, a vast amount of sunlight that hits the snow is reflected back into space instead of warming the planet. About 80-90% of incoming solar energy is reflected by snow cover whereas only 10-20% of solar energy is reflected by snow-free surfaces such as soil or vegetation. Without snow, the ground absorbs more sun’s energy adding heat to the system, thereby causing even more snow to melt. Thus, surface temperature is highly dependent on the presence or absence of snow cover, and temperature
trends have been linked to changes in snow cover (Groisman et al., 1994). Studies on snow cover area indicated that in some regions, the snow cover area has been decreasing since 1980’s (Groisman et al., 1994; Frei and Robinson, 1999) with rising temperature, whereas in some other regions snow cover area has been increasing with increasing temperature (Ke and Li, 1998) which gives an idea of existence of regional disparity.
Climate change is likely to lead to an intensification of the global hydrological cycle and to have a major impact on regional water resources (Arnell, 1999). Among the major river systems, the impact of climate change on the hydrological behavior of the Ganges- Brahmaputra Basin is expected to be particularly strong. As the Brahmaputra Basin is highly influenced by the monsoon rainfall, the climate change that results in variation in intensity of the monsoon, will affect both high and low flows leading to increased flooding and variability of available water both in space and time (Postel et al., 1996) in the basin.
Himalayas are called the reservoirs of snow and glaciers. The ice mass over this region is the third-largest on the earth, after the Arctic/Greenland and Antarctic regions (Immerzeel, 2007). Bernett et al., (2005) indicated that Himalaya is perhaps the most critical area, where melting of snow and glaciers will negatively affect water supply in the next few decades, because of the region’s high population. In this region, the increase in temperature will accelerate snow and glacial melt which in turn will increase water availability. According to Frauenfelder and Kaab (2009), total glacier area in the upper Brahmaputra River basin has been decreasing by 7% to 10% per decade from 1970/80 to 2000. Another study carried out by Bolch et al., (2010) for five glaciers in the south-eastern centre of the Tibetan Plateau, which feeds into the Tsangpo-Brahmaputra River revealed that glaciers in this area had been retreating at a rate of around 10m/year for the period 1976 to 2009. The increased rates of snow and glacial melt will increase summer flows in the river for a few decades, followed by a reduction in flow as the glaciers disappear and snowfall diminishes (Immerzeel, 2008).
Brahmaputra River is most susceptible to reductions of flow, threatening the food security of an estimated 26 million people (Immerzeel et al., 2010).
During its course, the river Brahmaputra is joined by important tributaries from the Himalayan ranges of Arunachhal Pradesh and Bhutan in the north viz., Subansiri, Kameng, Dhansiri and Manas and from the south by Dihing, Disang, Dikhou and Kopili (Dhar and Nandargi, 2000). Out of these tributaries, the river Subansiri is considered as the major north bank tributary of the river Brahmaputra which contributes around 10% of the total annual discharge of Brahmaputra. Subansiri River flows through Tibet for 170 km, 250 km through
Arunachal Pradesh after which it enters Assam at Dulangmukh in Dhemaji district and flows for another 130 km through the plains of Assam before it joins the River Brahmaputra near Jamugurighat in Assam. During the long journey, Subansiri receives the discharges of numerous mountainous streams. The number of its tributaries is more in Siwalik foothills than in other zones. The total length of known and well-defined tributaries of Subansiri is 1,960 km (Sarkar, 2015). Subansiri basin is the largest river system of Arunachal Pradesh which covers more than 19,000 sq.km. in the central part of the State. The catchment area of Subansiri River is spread out further than the grand Himalayan mountain ranges, covering Tsona Dzong till the Great Loop of the Tsang Po River in Tibetan Territory. The entire Subansiri basin can be divided into four parts (Sarkar, 2015). These are, (i) the distant Tibetan mountains beyond the international border, (ii) the reach lying between the international boundary and Miri hills of Arunachal Pradesh, (iii) the Arunachal Pradesh portion between the outskirt of Miri hills and the inter-state boundary of Assam and Arunachal Pradesh and (iv) the plains of Assam. Out of these four parts, the first two belong to the great Himalayan range, the third belongs to the Sub-Himalayan and the fourth belongs to the fertile plains of Assam. It is evident from these categories that the upper part of Subansiri basin falls in the Himalayas and hence is covered with Himalayan snow and glaciers. Due to global warming this snow cover area of the basin changes which leads to the change in runoff of Subansiri River in its downstream. The heavy precipitation also strongly influences the discharge of the river. Subansiri basin extends from tropical to temperate zones and hence exhibits a great diversity in rainfall characteristics. In the Northern and Central Himalayan tracts, precipitation is scarce on account of high altitudes. On the other hand, Southeast part of the Subansiri basin comprising the sub-Himalayan and the plain tract in Arunachal Pradesh and Assam, lies in the tropics. Due to Northeast as well as Southwest monsoon, precipitation occurs in this region in abundant quantities. Particularly, Southwest monsoon causes very heavy precipitation in the entire Subansiri basin during May to October.
July and August are the high flood months (Sarkar, 2015).
Not only the effect of snowmelt and rainfall, the change in land use/land cover that is taking place in the Subansiri basin also influences the change in runoff of the river. Fohrer et al., (2001) mentioned that surface runoff is highly affected by land use/land cover changes over a watershed. Researchers have investigated the relationship between climate, land use and hydrological processes and their studies showed that stream flow generation capacity is also dependent on vegetation type (Dwarakish and Ganasri, 2015). The rapid growth in population
has led to change in land use in terms of deforestation for improving the agricultural production (Lorup et al., 1998). This leads to decrease in infiltration rate and increase in surface runoff. However, Ozturk et al., (2013) reported that influence of climate variability is more significant on surface hydrology than the land use change.
The Subansiri river system has its practical importance as it holds high water resources as well as hydropower potential for the country. The change in runoff of the river due to change in climate and land use/land cover will have impact on the performance of the Subansiri Lower Hydrelectric Project (SLHEP). It is one of the largest under construction hydropower projects in India with an installed capacity of 2000MW. As such, there is a need to study the impact of climate change on power potential of this hydropower project. There is a possibility that rapid accumulation of water in the glacial lakes in the upstream of the Brahmaputra basin can lead to a sudden breaching of the unstable dams behind which they have formed reservoirs. The resultant discharge of huge amounts of water would have catastrophic effects on people, both upstream and downstream. This will affect the discharge of the river Subansiri and thus the power potential of the Subansiri Lower Hydroelectric Project will be changed.
There are various hydrologic models ranging from lumped conceptual model to physically based distributed models which are in use worldwide, for flow forecasting of rivers based on meteorological data and catchment characteristics (Lorup et al., 1998). The physically based models take into consideration the physical characteristics of a watershed whereas the semi- distributed and fully-distributed models are capable of representing the spatial heterogeneity of the watershed (Dwarakish and Ganasri, 2015). Integration of land use models with the rainfall-runoff models provides quantitative information about the land use changes on hydrological output. Also, different land uses in different sub-watersheds yield different hydrological outputs (Lin et al., 2009).