SPATIAL TECHNIQUES IN SNOWCOVER AND SNOWMELT RUNOFF STUDIES IN WESTERN HIMALAYA

Abstract

Snow is of great importance as a key environmental parameter. It not only influences earth's radiation balance but also plays a significant role in river discharge. Amajor source of runoff and groundwater recharge in middle and higher latitudes are contributed through snowmelt from seasonal snow covered areas of the Earth's mountain region. The Himalayan mountain system is the source of one of the world's largest suppliers of freshwater. All the major south Asian rivers originate in the Himalayas and their upper catchments are covered with snow and glaciers. The Indus, Ganga and Brahmaputra river systems, originating from the Himalayan region, receive substantial amounts of snowmelt water and are considered as the lifeline of the Indian sub-continent. There is no detailed scientific evaluation available for Himalayan water resources, firstly, due to an insufficient network of observations for both precipitation and stream discharge measurements and secondly, Himalayan terrain being most rugged and inaccessible. Nevertheless, the available estimates show that the water yield from high Himalayan basins is roughly double that of an equivalent one located in Peninsular India. This is mainly due to inputs from snow and ice melt contributions. The perennial nature of Himalayan Rivers and the suitable topographic setting of the region provide a substantial exploitable hydropower potential in this region. Therefore, near real time estimation of snow cover is of utmost importance for effective management of water resources and can serve as a guideline for reservoir operations. Moreover, planning of new hydroelectric projects on the Himalayan Rivers emphasizes the need for reliable estimation of snow and glacier runoff. In the present study, Satluj River Basin (up to the Bhakra dam) located in the Himalayan mountain range was considered. The Satluj is the longest of the five rivers of Punjab state that flows through Northern India. River rises in the lakes of Mansarovar and Rakastal in the Tibetan Plateau at an elevation of about 4572 m and is one of the main tributaries of the Indus River. The total catchment area of Satluj River up to the Bhakra dam is about 56,500 km2 out of which about 22,305 km2 lies in India. The basinal area considered for the study is about Spatial Techniques in Snowcover and Snowmelt Runoff Studies in Western Himalaya 22,305 km2. The elevation of the basin varies widely from 500 mto 7000 m; although only a very small area exists above 6000 m. Owing to large differences in seasonal temperatures and great range of elevation in the catchment, the snowline is highly variable, descending to an elevation of about 2000 m during the summer months. The datasets used in this present study are satellite data (NOAA-AVHRR, IRS-WiFS and Terra/Aqua-MODIS images), digital elevation model (USGS-DEM and SRTM-DEM) and meteorological data. MODIS data, USGS-DEM and SRTMDEM were acquired from Internet, NOAA-AVHRR data (2002 to 2005) were procured from IITR-SES. The NOAA-AVHRR images prior to 2002 were acquired from Internet. IRS-WiFS data were procured from National Remote Sensing Agency (NRSA), Hyderabad. The meteorological data were collected from Bhakra Beas Management Board (BBMB) for ten ground stations located within the basin. There were few gaps in the meteorological data for the period November 2003-October-2004. Hence, the year 2003-2004 could not be considered for snowmelt runoff simulation. The present study was carried out to understand the hydro-meteorological behavior of the Satluj River and simulate the snowmelt runoff for different years using the snowmelt runoff model (SNOWMOD). The climate of the Himalaya is changing alarmingly, particularly meteorological variables like, temperature, rainfall and streamflow. Any simulation or forecast of streamflow is not complete without considering the trend in climate change. Therefore, the impact of climate change on the basin was analyzed using parametric and non-parametric models and hypothetical climate scenarios were created to understand the behavior of snowmelt runoff under the changed climatic conditions. The two important input parameters of the snowmelt runoff model are the snow cover area (SCA) and temperature. The efficiency of the simulation of the model lies in accurate estimation of SCA and interpolation of temperature. Hence, an accuracy estimation of mapping SCA using different remote sensing satellite data sets was carried out. Since, cloud poses a challenge to visible remote sensing, a technique was developed to estimate SCA using temperature data, which was used in the snowmelt runoff model. Another important parameter of snowmelt runoff model is temperature, which is interpolated for the whole basin using temperature lapse rate (TLR) technique. Presently, a fixed TLR value (0.65°C/ 100 m) is used for modeling in Himalayan basin. However, TLR varies with season and region. For seasonal estimation of TLR, a technique was developed in this study using satellite based land surface temperature (LST) maps. It was observed that while using the seasonally varying TLR in the snowmelt runoff model, the overall efficiency of simulation of melt water runoff has improved significantly. Due to the limited availability of field information on SCA in rugged terrain like Himalaya, satellite images are proving as an active tool in mapping it. Until now NOAA-AVHRR and IRS-WiFS, data were widely used for SCA estimation in several Himalayan basins. The suit of snow cover products produced from Terra/Aqua-MODIS data were not yet been used in SCA estimation and snowmelt runoff modeling in any Himalayan basin. To understand the snow mapping potential of these three data sets under different topographic and climatic conditions, accuracy assessment was done. It was observed that satellite-based estimation of SCA was in a good agreement with ground-based estimation with an error of 1.5-1.7 km2. Further MODIS data proved better in estimating SCA in higher elevation and different aspect classes. The dataset was effective in mapping snow under mountain shadow condition. The chances of getting cloud-free scenes in case of MODIS were higher due to higher temporal resolution. Besides that, MODIS has an automated snow-mapping algorithm, which reduces the time and errors incorporated during processing satellite data manually. Therefore, MODIS data were used for obtaining the SCA depletion curve for the Satluj basin. However, unlike other visible satellite images cloudcover was the main drawback with MODIS data especially during monsoon season. Hence, a methodology was developed to obtain information of SCA when satellite images were cloud-covered using daily mean air temperature data. The SCA was linearly correlated with the air temperature in a distributed manner, where R2 values above 0.9 were obtained for the years 2000, 2001 and 2002. The SCA obtained from this technique reduced the number of satellite images and provided the solution of cloud-covered images. Another essential parameter for conceptual temperature index based snowmelt runoff analysis is the TLR. Previously TLR was estimated using station air temperature data. However, in rugged and varied terrain like Himalaya, such stations represent only local air temperature values and are very sparsely located. Hence, a representative TLR could not be estimated. Hence, a methodology was developed to estimate TLR using MODIS LST data products and LST maps generated from AVHRR data, which is a continuous dataset and better representative of a terrain. The information on elevation was derived from the USGS DEM. It was observed that MODIS LST maps were better correlated with the elevation map. A comparison between MODIS LST values and air temperature shows that both the datasets are in good agreement and the standard error of temperature was between 1.4-1.7 °C. Hence, MODIS LST data product was used to determine TLR for the study area in different seasons for a period of six years (2000-2005). To simulate the snowmelt runoff in the Satluj basin, SNOWMOD, a snowmelt runoff model was used which is a temperature index approach of snowmelt computation. It simulates all components of runoff, i.e. snowmelt runoff, rainfall-induced runoff and base flow, using limited data. The model simulated daily streamflow satisfactorily for the years 2000-2001, 2001-2002, 2002-2003, and 2004-2005. Due to gap in the field data, the year 2003-2004 was not considered for simulation. The seasonal lapse rate estimated from the LST maps as discussed above was used in the model and considerable improvement in simulation was observed. The SCA estimated from the proxy method using air temperature was used in the model for the missing dates when satellite data were not available. It was observed that the overall efficiency increased while using varying TLR. Further, the trend in climatic variables like temperature, rainfall and streamflow of the basin was also studied using parametric and non-parametric models. It was observed that rainfall decreased significantly in the whole basin in last thirty years particularly in the middle elevation (1000-3600 m). The temperature has an increasing trend particularly in the middle elevation. The influence of the change in rainfall and temperature was reflected in the stream flow. Further, a climate change impact study was carried out for Satluj basin. For this purpose three hypothetical scenarios (T+1, T+2 and T+3 °C) were created to study the impact of climate change on stream flow. The change in computed stream flow due to change in climate scenarios provided an indication of the influence of climate change. In the lower and middle part of the basin which experiences higher snowmelt in summer, is found to have reduced snowmelt runoff under the changed climatic scenario. During winter season, when there is less snowmelt due to lower temperature in the present scenario shows more snowmelt under changed climatic condition because of more temperature. It has been observed that there is a fall in runoff during summer period, which is crucial period for irrigation and hydropower generation. The present approach of estimating SCA using mean daily air temperature data was a unique technique. The method of TLR estimation using satellite LST maps was done for the first time anywhere in the world and promising results were observed. The improvement in TLR and SCA estimation increased the efficiency of snowmelt simulation considerably. The study of climate change analysis at a basin scale was unique to any Himalayan basin. The hypothetical scenarios developed to study the impact of climate change provided a real picture on future streamflow under the present climatic trend. Such studies are very important for planning and utilizing the hydrological potential of Himalayan Rivers, which are sustaining the life and economy of the world's second highest populous country.