Accurate time and site-specific forecasts of streamflow are important in effective reservoir management and scheduling. The present study aimed to compare the efficiency of Least Square Support Vector Machine (LS-SVM) as a new data driven model and a conceptual hydrologic model (Hymod) to simulate the daily streamflow in a representative watershed in US, Leaf River Watershed (1950 km2). For this purpose, 5-years period (1958-1962) of daily data including rainfall, potential evapotranspiration and streamflow were used. First 3-years were used as calibration (training) period in Hymod and LS-SVM and two remaining years were selected for validation (testing) periods in two models respectively. Performances criteria (Kling Gupta Efficiency (KGE), correlation coefficient (R2 ) and the Nash-Sutcliffe (NS) coefficient) for both LS-SVM and Hymod models in verification period were calculated and demonstrated that LS-SVM is a very potential candidate for the prediction of long-term discharges and then can be used as a promising method for hydrological prediction in un-gauged area.
Pourreza Bilondi, M. , Khashei Siuki, A. and sadeghi tabas, S. (2015). Daily rainfall-runoff modeling with Least Square Support Vector Machine (LS-SVM). Journal of Water and Soil Conservation, 21(6), 293-304.
MLA
Pourreza Bilondi, M. , , Khashei Siuki, A. , and sadeghi tabas, S. . "Daily rainfall-runoff modeling with Least Square Support Vector Machine (LS-SVM)", Journal of Water and Soil Conservation, 21, 6, 2015, 293-304.
HARVARD
Pourreza Bilondi, M., Khashei Siuki, A., sadeghi tabas, S. (2015). 'Daily rainfall-runoff modeling with Least Square Support Vector Machine (LS-SVM)', Journal of Water and Soil Conservation, 21(6), pp. 293-304.
CHICAGO
M. Pourreza Bilondi , A. Khashei Siuki and S. sadeghi tabas, "Daily rainfall-runoff modeling with Least Square Support Vector Machine (LS-SVM)," Journal of Water and Soil Conservation, 21 6 (2015): 293-304,
VANCOUVER
Pourreza Bilondi, M., Khashei Siuki, A., sadeghi tabas, S. Daily rainfall-runoff modeling with Least Square Support Vector Machine (LS-SVM). Journal of Water and Soil Conservation, 2015; 21(6): 293-304.
)