Rainstorm monitoring based on symbolic dynamics and entropy

Rainstorm is one of the major natural disasters in the world. Because of the complexity and non-linearity, using the current methods to correctly monitor and predict rainfalls is difficult. in recent years, with the rapid developments of nonlinear science, nonlinear time series analysis has been widely used hi many scientific and technological fields. In this study two kinds of nonlinear methods, i.e., the robust symbolic dynamics and information entropy, are used for nonlinear time series analysis of rainstorms. The theoretical bases on symbolic dynamics, information entropy and nonlinear time series analysis are introduced, first. Then, a new algorithm for rainstorm monitoring, including data preprocessing, time series symbolizing, symbolic time series segmentation and information entropy calculations, is described. Finally, 45 cases of heavy rainstorms around the world are analyzed. The preliminary results show that the method developed in this paper is promising for rainstorm monitoring. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Conference ESIAT2011 Organization Committee.