Forecasting Carbon Dioxide Emissions from Domestic Flights in the Kingdom of Saudi Arabia: An Artificial Intelligence-Driven Approach

As part of the 2030 vision of the Kingdom of Saudi Arabia (KSA), pivotal action has been taken to reduce greenhouse gas emissions from various sectors. A considerable portion of all greenhouse gases stems from transportation, specifically aerial transportation. It stands that domestic and other short haul flights emit more Carbon Dioxide (CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf>) than longer flights. The first step to the reduction of emissions from these flights is to predict these emissions accurately. Therefore, in this study the CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> emissions from the aviation sector, specifically, domestic flights in KSA, is predicted utilizing Artificial Intelligence (AI) techniques. The models that were used in this research are three variants of the Long Short Term Memory (LSTM) model which are Simple, Stacked and Bidirectional, Convolutional Neural Networks (CNN), and Gated Recurrent Units (GRUs). The prediction accuracy of these models is compared using three different accuracy measures: Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE). The results indicate that the Bidirectional LSTM is the most accurate and suitable model in predicting the CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> emissions released by domestic flights.