A space-time typhoon trajectories analysis in the vicinity of Taiwan
Journal
Stochastic Environmental Research and Risk Assessment
Journal Volume
29
Journal Issue
7
Pages
1857-1866
Date Issued
2015
Author(s)
Abstract
Tropical cyclones are one of the most serious natural disasters in northwestern Pacific Ocean. In general, an average of three to four typhoons invades the vicinity of Taiwan annually, which brings heavy rainfalls and strong winds resulting in disasters including flooding, mudflows, and landslides, leading to severe damage to economies and casualties. Studies show that different tracks of typhoon can cause distinct spatio-temporal patterns of rainfall events at different regions of Taiwan. As a result, understanding the trajectories of tropical cyclones and their relationship to climatic variables at global scale is crucial for hydrological modeling and disaster migration in Taiwan, especially under the conditions of climate change. This study applied a probabilistic curve clustering technique, which is based on a regression mixture model, to classify the best tracks of typhoons across the area within 6° around Taiwan during the period of 1951–2009. For the purposes of modeling and forecasting the typhoon trajectories, the track cluster is performed separately in different seasons due to their distinct driving forces to typhoon movements. A generalized linear model (GLM) is used to characterize the relationship between the identified typhoon tracks and the dominant climate features derived from NCEP reanalysis data. Results showed the six major typhoon tracks in the vicinity of Taiwan for different seasons respectively. The result of GLM cross validation showed that the frequency of typhoon tracks passing cross Taiwan in summer can significantly depend upon with two empirical orthogonal functions (EOFs) of sea level pressure, and the third EOF of sea surface temperature. © 2014, Springer-Verlag Berlin Heidelberg.
Other Subjects
Atmospheric temperature; Climate change; Climate models; Disasters; Landslides; Orthogonal functions; Rain; Sea level; Storms; Surface waters; Trajectories; Clustering techniques; Empirical Orthogonal Function; Generalized linear model; Hydrological modeling; Modeling and forecasting; Sea surface temperature (SST); Spatiotemporal patterns; Trajectory modeling; Hurricanes; climate change; climate variation; hydrological modeling; modeling; natural disaster; rainfall; storm track; trajectory; tropical cyclone; typhoon; Taiwan
Type
journal article