Assessment of Streamflow in Ungauged Basin by Using Physical similarity approach

Abstract Estimating runoff in an ungauged basin is important for explaining hydrological phenomena and managing water resources within the basin. Regionalization methods are considered most common approaches for estimating the runoff in an ungauged basin. Among them physical similarity method is one, but very few studies have considered the climatic attribute for catchment classification and for determining the donor catchments under physical similarity method. In the present study, based on physical similarity of catchments 3 groups were categorized. A total of 10 catchments were used in the present study. K-mean Clustering algorithm was used to classify the catchments into 3 groups within each categorized group. Donor catchments were evaluated for each group within each categorized group based on the performance of GR4J hydrological model. KGE, R & Pbias were used as the statistical measure to evaluate the model performance as well as to select the donor catchment. Model parameters evaluated from donor catchment were transferred to target catchment for streamflow assessment & model performance evaluation. Results showed that the GR4J model performed reasonably well for the catchments having lower mean elevation and large area. Among 3 categorized groups, PA group performed reasonably well. This may be due to inclusion of both the physical and climate attributes for catchment classification.