The performance of current numerical weather prediction models for the short term prediction of severe convective storms remains relatively limited. The extrapolation of high resolution precipitation patterns from weather radar is clearly valuable at the small time and spatial scales associated with these events. However, this approach generally fails in predicting storm initiation and evolution. In this study, several years of volume reflectivity measurements from a C-band weather radar are used to study the characteristics of convective storms observed in Belgium. These data are analysed with the TITAN cell tracker which has been recently installed at the Royal Meteorological Institute of Belgium. First encouraging results are obtained from a 3-year dataset. The distribution of different storm properties such as duration, maximum reflectivity, echo-top heights are analysed. Statistics on the kinematics of the storms are shown. The influence of the diurnal and seasonal cycles is presented as well. The final goal of this study is to analyse the relation between storms evolution and some relevant parameters with the aim of improving convective storms nowcasting.