Process-based eco-hydrological modeling in an Eastern Himalayan watershed using RHESSys

Abstract

Hydro-ecological models are important tools used to estimate streamflow from mountainous watersheds as driven by precipitation and snowmelt under the influence of different spatially variable hydro-ecological response variables. These models enable us to gain insight into hydrological and ecological processes occurring in the watershed. In the present study, a hydro-ecological model, Regional Hydro-Ecological Simulation System (RHESSys) was set up for a small representative watershed in Eastern Himalayan region of Arunachal Pradesh to investigate spatial variation in saturation deficit during snow depletion period. RHESSys model is capable of estimating variations in evaporation, transpiration, streamflow, overland flow, saturation deficit, etc. from the landscape. Studies related to hydrological and ecological sustainability of the Eastern Himalayan Rivers are limited, although a major part of the region is covered with evergreen forests, where ecological parameters may have a great impact on watershed hydrology. Nuranang watershed, which is located in Tawang district of Arunachal Pradesh, India, was selected as the study area. The hydro-meteorological data from 2004 to 2008 were collected from Central Water Commission (CWC). Digital Elevation Model (DEM) with 30\xa0m spatial resolution was downloaded from NASA website. Calibration of the RHESSys model was performed for the depletion period of 2004 (13th April–21st August) and 2005 (17th April–4th September) and the model was validated for the years 2008 (17th April–20th September) and 2009 (13th April–23rd September) using observed streamflow at the watershed outlet. Comparison plot of the actual and log values of observed and simulated daily discharges showed that the model captured the variations in total observed discharge and subsurface flows reasonably well for both the calibration and validation years. Daily time-series outputs of various hydrological variables, i.e., subsurface flow, overland flow, streamflow, saturation deficit, and percentage saturated area, were obtained for both calibration and validation simulation. For simulation years, overland flow dominated stream discharge in all the months, and contributed around 60% of the total streamflow, whereas subsurface flow contributed around 40% in all the months. Comparatively, saturation deficit was more in the month of April due to less precipitation and it was reduced in the month of May because of pre-monsoon rainfall. It increased again in the month of June as the precipitation in the month was less compared to May. Then, it decreased again in the subsequent months of July and August due to high rainfall during the monsoon season and increased in the month of September. Average saturation deficit was estimated to be 389.69\xa0mm and average percentage saturated area was estimated as 9.52%. The number of saturated pixels was found to be minimum in the month of April, while a maximum number of saturated pixels were observed in the month of August indicating the existence of inverse relationship between precipitation and saturation deficit for the study area.