Environment

[E2-03] An advanced tropical cyclone center estimation approach using geostationary satellite-based infrared channels

*Juhyun Lee¹, Jungho Im¹ (1. Ulsan National Institute of Science and Technology)

Keywords:Tropical cyclone, center detection, rain band fitting

Since the tropical cyclone (TC) is one of the most destructive natural disasters, accurate monitoring is crucial. Operational monitoring of TC still relies on the forecasters’ subjectivities, this study proposed advanced fully automatic TC center estimation using geostationary satellite. The approach consists of two approaches: automatic setting of regions of interest (ROI) and TC center estimation using an advanced logarithmic spiral band (LSB) and score matrix (SCM). The setting of ROI enables to mask out the region where has less potential of TC center, and then, the accurate TC center would be determined considering TC size, length of rain band in relating on the intensity. Two schemes (i.e., schemes A and B), depending on the initial guess setting method, and two models (i.e., models A and B), whether SCM was considered with LSB or not, were examined. While both proposed models improved comparing to previous studies, the best performed model, considering both SCM and LSB, performed skill score of 20.8% in the scheme B. Especially, for the strong TC cases, the proposed model achieved the skill score of 72.3% for the scheme.