Validation of the Aircraft Noise and Performance Database Source Spectra

Abstract

Accurate modelling of aircraft noise in different weather conditions is crucial for the reliability of noise predictions and their application worldwide. In best-practice aircraft noise models, such as Doc.29, the change in expected sound level on the ground due to changing atmosphere is modelled with a simplified propagation calculation. The Aircraft Noise and Performance (ANP) database contains several standardised source spectra, known as spectral classes, which are used for these calculations to account for the frequency dependence of the atmospheric effects. The spectral classes consist of Pressure Band Levels (PBL) of 24 1/3rd octave bands. This research focuses on the agreement of these spectral classes with measurements, taken around Amsterdam Schiphol Airport, and quantifies the effect of differences in these spectra for the Doc.29 weather correction. The measurements, taken by an acoustic array close to the runway and by continuous single microphone noise measurement stations (NOMOS) at long range, are propagated to the standard distance of 1000 ft (for which the ANP spectral classes are given) taking into account the geometrical spreading and the actual atmospheric absorption. For the B737-800 and A330, differences in shape are found between the two measured spectra and the spectral class. This is partly due to the low signal-to-noise ratio for the high frequencies in case of large distances between the aircraft and the measurement system. The effect of the application of the measured spectra on the Doc.29 weather correction is found to be smaller than 0.5 dBA for the NOMOS positions, indicating the suitability of the current ANP spectral classes for the weather correction.