A Comparative Study of Severe Thunderstorm among Statistical and ANN Methodologies

Abstract Severe Thunderstorm is the extreme weather convective feature. It causes local calamities in various ways. Proper prediction with lead time is an important factor to prevent such calamities for saving the people. Here both probabilistic and machine learning technique is applied on weather data to get proper prediction. Traditional methodologies are already there for such kind of prediction purpose. But Naïve Bayes and RBFN methodology has been introduced here with some specific weather parameters that has not done before remarkably. A comparative study has been done which are Naïve Bayes, Multilayer Perceptron (MLP), K-Nearest Neighbor (K-NN), and Radial Basis Function Network (RBFN) on weather data. All these data have been procured from Kolkata located in North-East India. The result obtained applying Radial Basis Function Network is better among three methods yielding correct prediction of 95% for severe “squall-storm” and 94% for “no storm”. Prediction can have sufficient lead time of 10- 12hours.