Author(s): Jisha Joseph; Amey Pathak; Subimal Ghosh
Linked Author(s):
Keywords: Uncertainty; Variable infiltration capacity model; Monte Carlo simulation
Abstract: Various hydrological models are coupled with climate models in the recent past to study the hydrological responses to the climate change at river basin scale as it is very important for planning and management of water resources. However, these models have uncertainties due to the model parametrisation, choice of GCM and downscaling techniques, inaccurate data for calibration and validation. The calibration of the parameters is the major challenge in hydrological modelling as it demands very accurate observed data. The runoff data which is generally used for model calibration are controlled data due to the presence of dams, reservoirs and irrigation practices and are hence not reliable. Therefore it is very important to identify the major source of uncertainty so that the level of accuracy required in the model calibration can be studied. The objective of this study is to identify and compare the different sources of uncertainty in climate change simulations of Ganges basin. The Variable Infiltration Capacity (VIC) model used in this study is a semi distributed macroscale model. The Monte Carlo simulation (MCS) method is carried out for the four calibration parameters and vegetation parameters in VIC and the model is run with observed and reanalysis data for years 2000-2009. The results obtained are evaluated using observed soil moisture and evapotranspiration. The hydrological simulations are carried out for years 1979-2005 and 2011-2037 for different GCMs and downscaling methods. The spatial distribution of changes in soil moisture, evapotranspiration and water yield are plotted to obtain uncertainty band for each of the datasets. From the study it is demonstrated that the uncertainty due to choice in GCM is higher compared to the hydrological parameter uncertainty.
Year: 2016