Geometric Clustering Analysis of Typhoon Track and Its Impact on Northwest Pacific Countries

EasyChair Preprint 7487

Abstract

Tropical cyclones (TCs) are among the most dangerous meteorological phenomena with the power to cause catastrophic damages to human lives, societies, and properties. Their activities and occurrences have been considerably altered as a consequence of climate change. This study investigates the impact of TC tracks on the Northwest Pacific (NWP) nations by using Unsupervised Machine Learning (UML) K-mean clustering. The results indicated that the optimal number for clustering K-mean analysis of TCs is three. In addition, the risk of each NWP nation to the clustered TCs was investigated. It is found that most countries are vulnerable to cluster no. 2 TCs, whereas China and Vietnam are highly prone to cluster no. 3 events. Also, the geometric clustering analysis is a potentially useful technique to redefine the forecast trajectories and interpret their influence on the NWP countries.

Keyphrases: Clustering, NWP, Tropical Cyclone, machine learning

@booklet{EasyChair:7487,
  author    = {Yuei-An Liou and Truong-Vinh Le},
  title     = {Geometric Clustering Analysis of Typhoon Track and Its Impact on Northwest Pacific Countries},
  howpublished = {EasyChair Preprint 7487},
  year      = {EasyChair, 2022}}