| Article ID | Journal | Published Year | Pages | File Type |
|---|---|---|---|---|
| 4458781 | Remote Sensing Applications: Society and Environment | 2015 | 16 Pages |
•WinSRM is applied in a small seasonally snow-covered data scarce catchment.•Cloud free satellite images are interpolated to derive daily snow cover area (SCA%).•A logarithmic relation is developed between SCA% and average mean air temperature.•SRM is shown to be satisfactory in such data scarce eastern Himalayan catchments.•Explained methods to handle limited data conditions and estimate model parameters.
Snowmelt runoff modelling is very important tool to estimate the ice and snow induced runoff volume in the mountain basins. However, unavailability of data is a key challenge for snowmelt runoff modelling especially in the eastern Himalayan region. In this paper, an attempt has been made to apply the Windows-based Snowmelt Runoff Model (WinSRM) to estimate the snowmelt runoff from a small representative seasonally snow-covered eastern Himalayan catchment with high data scarcity located in Arunachal Pradesh. The 52 km2 catchment, having very steep average slope with an elevation difference of more than 1800 m from lowest to highest point, has been divided into three elevation zones. The satellite images of the catchment are acquired from Indian Remote Sensing Satellite IRS-P6 (LISS-III/AWiFS). Snow cover mapping is done using normalised difference snow index (NDSI) method. Snow depletion curves of each zone are generated from daily snow covered area percentage (SCA%) obtained by interpolating periodic SCA% determined from cloud free images, which is also a rarity in this region during the snow depletion season. A logarithmic relationship is established between SCA% and average mean air temperature (AMAT) to estimate the SCA% for validation period of 2004 for which cloud free IRS-P6 Satellite images were not available. The WinSRM model is calibrated using three years data of 2006, 2007, and 2009 for the depletion period and is validated for depletion period of 2004. It is shown that the SRM model can be used satisfactorily in such data scarce catchments of eastern Himalaya. The procedures adopted in this study to handle different limited data conditions and estimate some of the required model parameters are also explained in this paper.
