Optimization of spatial input parameter in distributed hydrological model

• Published in: ARPN Journal of Engineering and Applied Sciences
• Main Author: Khalid K.; Ali M.F.; Rahman N.F.A.; Mispan M.R.; Rasid M.Z.A.; Haron S.H.; Mohd M.S.F.
• Format: Article
• Language: English
• Published: Asian Research Publishing Network 2015
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-84940054615&partnerID=40&md5=8a3eb3308b208567f05e53f0938e4ebb

Hydrologic models are particularly useful tools in enabling the modeler to investigate many practical and significant issues that arise during planning, design, operation, and management of water resources systems. Distributed models should pass through a careful calibration procedure before they are utilized as the process of decision-making aids in the planning and administration of water resources. Although manual approaches are still repeatedly used for calibration, they are tedious, time-consuming and have the need of experienced personnel. This paper describes a semi-automatic approach for calibrating long term daily streamflow periods estimated by the Soil and Water Assessment Tool (SWAT) hydrological model. Optimization of three different sets of spatial input parameters were tested using SUFI-2 algorithms by firstly focusing on the sets of the groundwater inputs parameter. The second set is for the soil input parameters and the final set consists of 21 SWAT input parameters that reflect sensitivity on the streamflow simulation. For Langat river basin, the Rchrg-dp.gw, GW_Delay.gw and CN2.mgt were found to be most sensitive input parameters. SOL_AWC.sol was established to be the most sensitive to soil input parameter and followed by SOL_BD.sol and SOL_K.sol. On the final sets, it was shown that the three input parameters of OV_N.hru, SL_SUBBSN.hru, and HRU_SLP.hru were included as sensitive parameters in addition to the previous parameters. The next step should be conducting a long-term continuous hydrological modeling into SWAT 2012 model with all the selected sensitive SWAT input parameters in order to finalize the objective functions for the watershed. © 2006-2015 Asian Research Publishing Network (ARPN).