Abstract
Abstract: In this study, the equations for estimating the number of tropical cyclones (TCs) at a 6-month lead-time in the Vietnam East Sea (VES) have been developed and tested. Three multivariate linear regression models in which regression coefficients were determined by different methods, including 1) method of least squares (MLR), 2) minimum absolute deviation method (LAD), 3) minimax method (LMV). The artificial neural network model (ANN) and some combinations of the above regression models were also used. The VES was divided into the northern region above 15ºN (VES_N15) and the southern one below that latitude (VES_S15). The number of TCs was calculated from the data of the Japan Regional Specialized Meteorological Center (RMSC) for the period 1981-2017. Principal components of the 14 climate indicators were selected as predictors. Results for the training period showed that the ANN model performed best in all 12 times of forecasts, following by the ANN-MLR combination. The poorest result was obtained with the LMV model. Results for the independent dataset showed that the number of adequate forecasts based on the MSSS scores decreased sharply compared to the training period and the models generated generally similar errors. The MLR model tended to give out the best results. Better-forecast results were obtained in the VES_N15 region followed by the VES and then the VES_S15 regions.
 Keywords: Tropical cyclone, Seasonal prediction, Vietnam East Sea (VES).
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: VNU Journal of Science: Earth and Environmental Sciences
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
