Seasonal Prediction of Tropical Cyclone Frequency in the Western North Pacific using GDAPS Ensemble Prediction System

GDAPS 앙상블 예보 시스템을 이용한 북서태평양에서의 태풍 발생 계절 예측

  • Kim, Ji-Sun (Department of Atmospheric Sciences / Typhoon Research Center, Kongju National University) ;
  • Kwon, H. Joe (Department of Atmospheric Sciences / Typhoon Research Center, Kongju National University)
  • 김지선 (공주대학교 대기과학과/태풍연구센터) ;
  • 권혁조 (공주대학교 대기과학과/태풍연구센터)
  • Received : 2007.06.12
  • Accepted : 2007.07.12
  • Published : 2007.09.30

Abstract

This study investigates the possibility of seasonal prediction for tropical cyclone activity in the western North Pacific by using a dynamical modeling approach. We use data from the SMIP/HFP (Seasonal Prediction Model Inter-comparison Project/Historical Forecast Project) experiment with the Korea Meteorological Administration's GDAPS (Global Data Assimilation and Prediction System) T106 model, focusing our analysis on model-generated tropical cyclones. It is found that the prediction depends primarily on the tropical cyclone (TC) detecting criteria. Additionally, a scaling factor and a different weighting to each ensemble member are found to be essential for the best predictions of summertime TC activity. This approach indeed shows a certain skill not only in the category forecast but in the standard verifications such as Brier score and relative operating characteristics (ROC).

Keywords