http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Application of SVM and SWAT Models for Monthly Streamflow Prediction, a Case Study in South of Iran
Milad Jajarmizadeh,Elham Kakaei Lafdani,Sobri Harun,Azadeh Ahmadi 대한토목학회 2015 KSCE Journal of Civil Engineering Vol.19 No.1
The present study compares the results of the Soil and Water Assessment Tool (SWAT) with a Support Vector Machine (SVM) topredict the monthly streamflow of arid regions located in the southern part of Iran, namely the Roodan watershed. Data collected overa period of 19 years (1990-2008) was used to predict the monthly streamflow. Calibration (training) and validation (testing) wereperformed within the same period for both the models after the preparation of the required data. A semi auto-calibration was performedfor the SWAT model. Also, the best input combination of the SVM model was identified using the Gamma Test (GT). Finally, thereliability of the SWAT and SVM models were evaluated based on performance criteria such as the Nash-Sutcliffe (NS) modelefficiency coefficient and the Root Mean Square Error (RMSE). The obtained results from the development of the SWAT model andSVM model indicated satisfying performance in predicting the monthly streamflow in the large arid region. The SWAT obtained NSand RMSE values of 0.83 and 6.1 respectively, and the SVM obtained NS and RMSE values of 0.84 and 6.75 respectively for thevalidation (testing) period. Results indicate that for high flows of more than 19 (m3/s), both models predict flow with over and underestimation in the validation (testing) period. Moreover, the SVM has a closer value for the average flow in comparison to the SWATmodel; whereas the SWAT model outperformed for total runoff volume with a lower error in the validation period.