Classifying southwest Pacific Tropical Cyclone Tracks by their Orientation and Shape

Abstract:

Forecasting of tropical cyclone (TC) movement post-formation presents an ongoing challenge since accurate track prediction facilitates preparedness of societies to the destructive impacts of TCs. Evaluation of variations in the morphology of TC tracks (i.e. shape, direction) is essential to refine the predicted trajectory and potential impact zones. Therefore, this study aimed to identify specific clusters of TC tracks and their morphology for the southwest Pacific (SWP) region and identify any trends over the last 70 years (1948 to 2017). Data was obtained from the South Pacific Enhanced Archive of Tropical Cyclones (SPEArTC) database, and a probabilistic clustering technique was applied to group TC tracks into statistically distinct clusters, and the primary cyclone trajectories were assessed for the region. In addition, TC tracks were classified into four sinuosity indexed categories based on their morphology. The cluster analysis identified five clusters initiating in the distinct portions of the SWP basin, all spatially exhibiting the typical northwest to southeast orientation, except for some TCs forming over/near the Gulf of Carpentaria that move towards the Indian Ocean. A significant increase in the number of TC tracks was observed for the eastern SWP cluster, while other clusters displayed decreasing or no trend. The sinuosity analysis revealed that the morphology of TC tracks has changed over the past seven decades, both spatially and temporally. While the western SWP is typically exposed to erratic tracks, the eastern SWP observes substantial changes in track morphology from straight/quasi straight trajectories to more sinuous (i.e. erratic) tracks. It was also observed that some TCs (of varying morphologies) occur within the same spatial extent yet follow different paths. In particular, the highly erratic TCs were common within the Coral Sea, hence amplifying the landfall risk for Northeastern Australia. These findings suggest that combined cluster analysis and TC track sinuosity analysis is an important tool in generalising the TC track regimes, refining predicted trajectories and understanding impacts on SWP island nations.