Analysis of aircraft gas emissions for diyarbakir airport using artificial intelligence methods

One of the significant sources of greenhouse gas emissions is the aviation sector, where significant amounts of pollutant gases are released into the atmosphere due to the fuel consumed during the landing and take-off cycle (LTO) of airplanes. To determine the gas emissions released into the atmosphere by airplanes during flight, it is necessary to know the aircraft model and the type of engine used in that model. With this information, it becomes possible to determine the gas emissions originating from aircraft. In this thesis study, it is aimed to determine the gas emission amounts from the airplanes using Diyarbakır Airport between January 2022 and March 2024. For this purpose, registration information obtained from the General Directorate of State Airports Authority was used, and engine types for each aircraft were identified based on queries made using open sources. Taxi times at Diyarbakır Airport, published by the European Organization for the Safety of Air Navigation based on data obtained directly from airlines, were used as the basis for calculating gas emissions. Using the Aircraft Engine Emissions Databank (EEDB), which is regulated by the International Civil Aviation Organization (ICAO) according to engine types, the amount of fuel consumed by each aircraft during the taxi phase was determined. The identified fuel amounts were multiplied by the emission factors available in the emissions databank, and the amounts of CO, HC, and NOX emissions from each aircraft were calculated. Using this data, al-powered prediction models were proposed. It was observed that the deep learning-based LSTM method achieved higher prediction performance in the conducted forecasting applications.