Developing a multivariate model for predicting the noise annoyance responses due to combined water sound and road traffic noise exposure

Abstract

People in an urban environment are exposed to different types of natural and man-made sounds. Human sound perceptions due to exposure to a single noise source, in particular road traffic and aircraft noises, have been investigated for a long time. However, only very few studies have been focused on exposure to a combination of sound sources. Also, there is a lack of multivariate models that can help to predict the preferences or annoyance responses as a result of adding a wanted sound to an unwanted sound. Accordingly, this study aimed at developing a multivariate model to predict the probability of invoking a high noise annoyance response due to combined water sound and road traffic noise exposure. A series of laboratory experiments were performed. Participants were presented with a series of acoustical stimuli before being asked to assign their annoyance ratings. Results suggested that other than acoustical properties like sound pressure levels, personality traits were found to exert considerable influences on the maximum likelihoods of the model prediction and thus should not be excluded from the model specification form. Also, the quality of the acoustical environment could be improved by adding water sounds to road traffic noises at high levels. The capability of stream sound to moderate noise annoyance was found to be slightly stronger than that of fountain sound. In addition, the formulated multivariate model enables to reveal the tradeoff decisions performed by people. An increase in the SPL of road traffic noise by 1dB was considered to be equivalent to a reduction in the SPL of water source by 1.7dB for a given probability value. Results arising from this study should provide valuable insights on understanding how humans respond to the combined water sound and road traffic noise exposure.