An analysis of airport noise data using a non-homogeneous Poisson model with a change-point

In this work, a non-homogeneous Poisson model is considered to study the behaviour of airport noise levels. The model is used to count the number of times that the noise level exceeds a given threshold in a time interval of interest. The rate function of the Poisson process is assumed to be of a Weibull type. Two threshold values are considered. One of them, make it evident the change in the data’s behaviour. Hence, it requires the use of a model with a change-point. The models considered depend on some parameters that need to be estimated. After the estimation of the parameters are made, a way of obtaining the probability that the noise threshold is exceeded a certain number of times in a given time interval is presented. Results related to the mean number of exceedances of the threshold are also provided. Those results can be very useful in the study of the population’s exposure to noise produced by air traffic and other types of noise related to the airport operation. The model is also useful to predict, given the current behaviour of the data, the probability of occurrence of high levels of noise in a near future. An application to the airport noise data from Nice International Airport is given.