| || || Chand, Arti Pratap|
| || || Climatic variability : a study of tropical cyclone track sinuosity in the Southwest Pacific|
Author:Chand, Arti Pratap
Institution: University of the South Pacific.
Subject: Climatic changes -- Oceania , Cyclone tracks -- Tropics -- Oceania , Cyclones -- Oceania
Call No.: Pac QC 903 .2 .O3 C532 2012
Copyright:Under 10% of this thesis may be copied without the authors written permission
Abstract: Tropical cyclones (TCs) are one of the most destructive natural hazards in the tropical Pacific, with large impacts on socio-economic and environmental sectors of island nations. Improved understanding of the characteristics of these intense storms is critical. A continuing problem lies in forecasting TC movement after formation. One way to add to existing knowledge in this area is to analyse available data on cyclone track shape, in order to identify any special patterns. In this context, this study examines statistical characteristics of several TC track parameters, using archived data from 1970 to 2008 for the South Pacific region. The dataset includes information on 292 TCs, which includes all storms with wind intensity of 35 knots and above that have their genesis in tropical waters. TC paths are analysed within the geographical grid covered by 0 – 25°S and 160° E – 120° W. The particular focus of this study is on track sinuosity values and how these may be characterised and grouped. River sinuosity has contributed a lot in understanding fluvial geomorphology (Terry and Feng, 2010) and therefore extending the technique to study TC track maybe useful. A sinuous track having loops and curves will affect many more islands than a TC moving along a straight path. Some Islands may be affected more than once or may be exposed to a TC for a longer time period if the TC makes a loop during its journey. Sinuosity values for all TC tracks were calculated by measuring the total distance travelled by each TC and then dividing this by the vector displacement between cyclogenesis and decay positions. In this study, the problem of categorising the TCs based on sinuosity index (SI) values obtained by transformation of sinuosity values allows the grouping of similar TCs. The SI categories are so constructed that the variance of groups is as small as possible. Thus in this thesis a technique is developed to construct the SI categories of the TCs that seek minimization of the sum of weighted deviations of SI from the mean of group. Then the problem is solved for determining the optimum boundary points of the groups by using a dynamic programming technique.