Identification of significant parameters for the prediction of pre-monsoonthunderstorms at Calcutta, India

In the present work, the well-known principal component analysis (PCA) tech nique has been applied to study the thunderstorm phenomenon (which evidentl y has a multivariate structure) at Calcutta (India) during the pre-monsoon season (i.e. March-May). Various parameters (both thermodynamic and dynamic ) already identified by different scientists as responsible for thunderstor m occurrence have been considered here for different atmospheric layers. Th e purpose of the study is to reduce the number of parameters and hence iden tify the significant parameters linked up with various layers of the atmosp here for thunderstorm as well as for fair-weather days of Calcutta so that these may be used for parameterization. As the analysis reveals that there is a structural difference between the m orning and the afternoon atmosphere at Calcutta, so the analysis has been p erformed separately for morning and afternoon on thunderstorm as well as fa ir-weather days. Originally, 20 parameters were included in the analysis. T he final result shows that in the morning only four parameters out of the 2 0 and in the afternoon only five parameters out of the 20 are found to be s ignificant (i.e. they are expected to be jointly responsible) for occurrenc e or non-occurrence of thunderstorms at Calcutta during the pre-monsoon sea son. The study has however been confined up to the 500 hPa level. Copyright (C) 1999 Royal Meteorological Society.