Abstract
Tropical cyclones (TCs) are one of the most destructive natural hazards in the tropical South Pacific region, so understanding characteristics of their movement is important for hazard risk assessment and adaptation. This study examines the statistical characteristics of track shape for 291 TCs from 1970 to 2008 in the South Pacific (160°E–120°W, 0–25°S). The particular focus is on TC track sinuosity properties and how these may be characterised and grouped using a robust technique such that categories so constructed within which TCs are of similar sinuosity characteristics. In this paper, we propose a mathematical programming approach to determine the optimum boundary points of the categories that seeks minimisation of the sum of weighted variance of sinuosity index values between categories. The problem is formulated as a nonlinear programming problem which is then solved using a dynamic programming technique. Applying the technique we proposed five homogeneous categories of TC tracks found to be (1) straight, (2) near straight, (3) curving, (4) sinuous, and (5) convoluted tracks. The results are compared with the track-shape categories that are obtained by the K-mean cluster analysis method and a hierarchical cluster analysis with Ward’s method. The comparison shows that categories constructed by the proposed mathematical programming approach are more homogenous than the categories obtained by other methods.
Author: Arti Pratap Chand, M. G. M. Khan, Gennady Gienko and James P. Terry
Received on: May, 2017
Accepted on: January, 2019