Abstract
Significant variations in measured overflight noise levels are observed from airport monitor networks for similar aircraft types flying the same departure and arrival procedures. Various factors, including aircraft configuration, airline, weather, and flight procedure, as well as their interactions, have the potential to impact recorded noise levels. To assess these impacts, various multifactor analysis approaches are applied to coupled operational surveillance data, the National Oceanic and Atmospheric Administration weather data, and noise monitor recordings. A year of Boeing 737-800 departure surveillance data and associated noise monitor recordings from the Port of Seattle Aircraft Noise Monitoring System is obtained for the analysis. Noise monitor recordings are assessed via multivariate linear regression, gradient-boosted decision trees, and clustering to determine which factors and their interactions correlate to increasing or decreasing noise levels. The analysis of Boeing 737-800 noise monitor data reveals that altitude, position, true airspeed, thrust, and airline significantly impact noise levels.