A comprehensive framework for HSPF hydrological parameter sensitivity, optimization and uncertainty evaluation based on SVM surrogate model- A case study in Qinglong River watershed, China. (September 2021)

• Record Type: Journal Article
• Authors: Xingpo, Liu; Muzi, Lu; Yaozhi, Chai; Jue, Tang; Jinyan, Gao
• Is Part Of: Environmental modelling & software. Volume 143(2021)
• Publication Date: 2021-09
• Subjects: Hydrological simulation program-FORTRAN(HSPF) -- Surrogate model -- Support vector machine (SVM) -- Parameter sensitivity -- Parameter optimization -- Parameter uncertainty; 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.2021.105126
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: To improve HSPF hydrological and water quality simulation, a new SVM surrogate modeling method was investigated and a comprehensive framework for hydrological parameter sensitivity, optimization and uncertainty analysis was established for Qinglong River watershed, Hebei Province, China. SVM surrogate model was set up based on pairs of parameter sets and Nash-Sutcliffe efficiency coefficients. It was concluded that: (1) SVM surrogate model performs well in both reliability and efficiency. (2) Sensitivity of eleven parameters was evaluated: AGWRC is extremely sensitive parameters, AGWETP, DEEPFR, BASETP are sensitive parameters and UZSN, LZSN, LZETP, INTFW, CEPSC, IRC, INFILT are not influential parameters. (3) Recommended parameter intervals were: LZSN [2.0, 5.82], INFILT [0.21, 0.47], AGWRC [0.85, 0.87], DEEPFR [0.001, 0.17], BASETP [0.001, 0.09], AGWETP [0.0011, 0.13], CEPSC [0.01, 0.29], UZSN [0.05, 1.20], IRC [0.3, 0.62], LZETP [0.34, 0.85], INTFW [1.0, 5.77] and the optima were obtained respectively. (4) Posterior distributions of eleven parameters were obtained. Graphical abstract: Image 1 Highlights: The sensitivity analysis, optimization and uncertainty analysis were done by a SVM surrogate model. A parameter sensitivity method was proposed by SVM interpolation and regression. A case study in Qinglong River watershed was carried out and showed feasibility of SVM surrogate model.