FORECASTING INTERNATIONAL AIR TRAFFIC USING REAL-TIME FLIGHT DATA: A CASE STUDY OF KYRGYZSTAN AIRPORTS

Abstract

This article examines the problem of forecasting air traffic at international airports in Kyrgyzstan using time series analysis. Bishkek and Osh airports were chosen as the study sites. Initial data on daily aircraft arrivals and departures was obtained from the FlightAware platform. Data collection, cleaning, and preprocessing were performed, including timestamp conversion and missing observation recovery. The Facebook Prophet model, which takes into account trend and seasonal components of the time series, was used for forecasting. Data for the period 2019–2025 was analyzed, allowing us to identify the impact of the COVID-19 pandemic on the aviation industry, the specifics of air travel recovery, and the presence of stable annual seasonality. Based on the constructed model, a forecast of air traffic for 2026 was obtained.