Parameterization in hydrological models through clustering of the simulation time period and multi-objective optimization based calibration. (April 2021)

• Record Type: Journal Article
• Authors: G., Lakshmi; Sudheer, K.P.
• Is Part Of: Environmental modelling & software. Volume 138(2021)
• Publication Date: 2021-04
• Subjects: Discrete parameterization -- Soil moisture -- Clustering -- Multi-objective optimization -- Streamflow -- Water balance; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2021.104981
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Distinct calibration for wet and dry periods of the simulation period has resulted in fixed discretization of the simulation periods throughout the year with a repetitive pattern for all the simulation years ignoring the variability of rainfall and soil moisture conditions along time. This may induce anomalies in the water balance of the basin. This study focused on developing a discretization methodology using clustering methods that help eliminate the existing limitations. Subsequently, the clusters are calibrated independently, as well as simultaneous with multiple objectives. The procedure is illustrated through two models viz. SWAT and a grid based model; set up for the US watersheds Cedar Creek, Indiana and Riesel Y2, Texas respectively. The proposed approach resulted in improved stream flow simulations with higher NSE (0.83 in multi-objective; 0.68 in traditional calibration) in Cedar Creek. The simulation of water balance components was found to be more effective in both the basins. Graphical abstract: Image 1 Highlights: Misrepresentation of temporal process dynamics in discrete calibration approaches, especially during the transition periods. Clustering of time period based on antecedent moisture conditions and calibration of the selected clusters is a better choice. Simultaneous multi-objective optimization of independent clusters resulted in improved performance and process representation.