Short range fog forecasting by applying data mining techniques: Three different temporal resolution models for fog nowcasting on CDG airport

Forecasting fog is an important issue for air traffic safety because adverse visibility conditions represent one of the major causes of traffic delay and of the economic loss associated with such phenomena. In such context the present work illustrates a Data Mining application for the fog forecast on a short time range (1 hour, 2 hours and 3 hours) on Paris Charles de Gaulle airport. Indeed three predictive models have been built using an historical dataset of 17 years of fog observations and other relevant meteorological parameters collected in the SYNOP message and by applying a BayesNet algorithm. The performances evaluation show that the best model for the fog forecast is that on one hour time range, presenting a percentage of correct classified instances of 97% and a true positive rate of 88%. The other implemented models show slightly worse performances with a percentage of correct classified instances of about 96% and 95% respectively and true positive rates of 80% and 71%. The work has been carried on according to the standard process (CRISP-DM) for Knowledge Discovery in Meteorological Database Process.