Receiver-centric mapping of pedestrian noise annoyance: a cost-effective approach using random forest, psychoacoustic metrics and open-source data

Abstract

This article presents a method for mapping traffic-induced noise annoyance from the pedestrian's perspective using Psychoacoustic [PA] metrics and open-source data. Prioritising the receiver over the source, the approach avoids costly traffic counts and proprietary software by extracting PA values from field-recorded audio and applying Random Forest [RF] modelling to extrapolate these values as a spatial performance indicator with robust predictive ability. The results were propagated using an exponential decay function, and the output is presented as a high-resolution GIS-based noise annoyance map for spatial analysis of pedestrian impacts, offering a receiver-centric alternative to convent..