Short-term prediction model of airworthiness in time series

Meteorological events are very important phenomena in aviation field. Aviation activities are interrupted from time to time as a result of meteorological events. They should be considered primarily as they cannot be substituted and controlled such as inventory or human factors. In this study, the importance of meteorological events for flight training organizations is examined on the case of Eskişehir Technical University Pilotage Department (ESTU-P). Airworthiness time series were obtained from meteorological aerodrome reports (METAR) covering a period of twelve years, using the official criteria in the regulations determined by the aviation authorities and the informal criteria based on the instructor pilot experiences. Prediction models for the next day are suggested by using decomposition and deep learning methods in time series. In addition, statistical airworthiness results for the period 2009-2020 have been presented in order to contribute to planning activities and provide resources for researchers who examine flight training organizations.