Time series modeling and forecasting: Tropical cyclone prediction using ARIMA model

An Auto Regressive Integrated Moving Average (ARIMA) model is best suited for time series analysis to enhance the analysis of the data and to forecast the unknown data in the series also called forecasting, in Statistical Studies, Analysis, Business Econometrics, Big data and in Predictive analytics, especially for time series analysis,. Tropical Cyclone (TC) is among the most notorious hazard that cost human lives and causes tremendous economic loss, and damages especially to the peoples who reside in the coastal areas. Therefore, it is of considerable importance to correctly predict TC, its tracks and intensities for effective disaster analysis, responses, mitigation and management. This paper presents a Statistical Time Series Modeler (TSM) for prediction of Cyclonic Storm of India. For developing this model, 6 years data (2007-2012) is used, consisting of 14 attributes. TSM of SPSS (Statistical Package for Social Studies) is applied for training and testing the Storm dataset. The model is developed using the training data (2007-2011) and applied using the testing data set (2012). This model is built on ARIMA (Auto Regressive Integrated Moving Average) model of TSM in SPSS 20.0.