Comparative implementation of melody search in auto-calibrating SWAT

Alireza Borhani Dariane,Reza Bagheri,Mahboobeh Ghasemi,Roza Asadi
DOI: https://doi.org/10.1007/s12517-024-11974-9
2024-05-08
Arabian Journal of Geosciences
Abstract:Rainfall-runoff simulation models require calibration to ensure accurate results. The present study assesses the efficacy of the Melody Search algorithm (MeS) in auto-calibration of the SWAT, by comparing it to other commonly used calibration methods in SWAT-CUP, namely, SUF12, GLUE, ParaSol, and PSO. In order to assess the MeS algorithm's performance, the continuous rainfall-runoff simulation was implemented using daily time-step data from the Taleghan Basin in northern Iran, spanning a 10-year period. The calibration of parameters was established through developing a FORTRAN program. The Nash–Sutcliffe efficiency (NSE) index indicator in SUFI2, GLUE, ParaSol, PSO, and MeS algorithms was found to be 0.647, 0.6, 0.628, 0.625, and 0.66, respectively. Thus, based on these results, the proposed MeS algorithm outperforms all the native calibrating methods in SWAT in all iterations. MeS specifically outperforms all native approaches in reaching a solution after 100 iterations. The results imply that MeS is an algorithm that quickly converges and functions as a reliable calibration technique for SWAT applications.
geosciences, multidisciplinary
What problem does this paper attempt to address?