Tropical Cyclone Track Forecasting Using Optimized Consensus in the Western North Pacific
Abstract
This study investigates an optimized consensus method using superensemble, which had the highest skill for 2012 tropical cyclone forecast in NHC, to provide good guidance to forecasters. The superensemble forecast is partitioned into a training phase and forecast phase. In the training phase, weights and bias estimation for the models are calculated on the basis of past forecasts via linear regression. Each forecasted track (latitude, longitude) per forecast period (24h, 48h, 72h, etc.) is obtained by the regression coefficients optimized in the past 2-year training period. We used the KMA (Korea Meteorological Administration) analysis data as the reference value and model (ECMWF, GFS, etc.) forecast data over the western North Pacific during 2011-2013 to verify the superensemble forecast results. The optimized consensus track forecast method, e.g., the monthly weighting, typhoon track classification, has been tested for tropical cyclones in 2013.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.A23E0286J
- Keywords:
-
- 4315 NATURAL HAZARDS Monitoring;
- forecasting;
- prediction