Spiral band model optimization by chaos immune evolutionary algorithm for locating tropical cyclones

Abstract

A foundation of forecasting Tropical Cyclones (TCs) is to locate their centers. There are still many difficulties in TC center locating, especially when there is an unclear eye or when spiral bands are difficult to be recognized. It is usually done by hand operationally, and multiple forecasters' results may be inconsistent. In this paper, we treat the locating problem as an optimization one. An objective locating scheme is developed, using infrared (IR) satellite cloud images. A novel Spiral Band Model (SBM) is designed to extract and describe the spiral pattern of a spiral band which spirals out from a TC's center. SBM is based on a spiral band region, different from existing curvilinear models which are based on curves. A simple criterion is also defined to score the matching degree between a template derived by SBM and the spiral band of a TC. In this paper the TC center locating problem is transformed to an optimization problem, and the Chaos Immune Evolutionary Algorithm (CIEA) is employed to solve this optimization problem. The spiral center of the template found at the end of the optimization process is the estimated location of the true TC's center. The experimental result demonstrates that the proposed scheme obtains the best average error of 0.0930° in latitude/longitude in comparison with the best track data. It is well within the relative error of about 0.3° from the results of different TC warning centers. It provides objective information for forecasters to estimate TCs' centers.