Individual versus generic acoustical predictive modeling of aircraft related annoyance

摘要

Aim of the study is the acoustical prediction of annoyance ratings of airport residents measured by the hourly ICBEN scale of groups of subjects and of individuals. Three different cases are considered: 1. Case 1: Generic (sample based) time series modeling is applied to data from 3 European airports (Arlanda, Cologne, Heathrow) 2. Case 2: Individual time series modeling is applied to subjects of Cologne airport. 3. Case 3: Without reference to time the data of individual subjects are predicted using again the data sets from all three airports. Dependent on time the case 1 generic modeling showed Goodnes Of Fit (GOF) - scores~2 with maximal values of r = 0,65. In the early morning hours and in the evening the mean GOF-scores are considerably less than in the time between. Individual modeling allowed a separation of subjects responding more sensitive to aircraft noise than other subjects which seem to be unaffected. Obviously both groups differ in their perception and/or motivation related to aircraft noise. This result is confirmed by the case 3 analysis. The data suggest that motivation plays a crucial role when trying to forecast annoyance by acoustical predictors. If the acoustical airport environment has no impact on the annoyance state of the resident no prediction is possible. Thus, as a conclusion for further predictive studies it seems to be necessary to run pretests to ascertain that varying and sufficient degrees of annoyance are given within the sample of residents.