with the advancement of information technology and their tremendous development, �Numerical Weather Prediction� is used by many meteorological services for predicting weather forecasts. This is available mostly for the welfare of the public. As this needs more scientific knowledge, Global Forecast Model came into existence for prediction of weather development from Numerical Weather Prediction. Data mining Clustering technique is applied in this analysis for forecasting the National Centre for Medium Range Weather Forecasting model which helps in predicting cloudburst. Tamil Nadu recently overcame a dreadful cloudburst on October 2015. Foretelling of cloudburst is exceptionally hard. This could be foretold only a few hours before. In difference to the on top of statement we have a tendency to predict cloud burst two or three days before. In this article rainstorm over Uttarkhand that created a heavy loss has been analysed by k-means algorithm.