Advancing the characterization of urban acoustic environments through multimetric analysis

Urbanization has intensified the complexity of acoustic environments, necessitating a more comprehensive understanding to support urban acoustic planning for healthier living spaces. Traditional noise monitoring, primarily based on sound pressure level indices, is insufficient for capturing the full scope of these environments. This study investigates whether a diverse set of acoustic metrics can improve the characterization of acoustic environments and examines their stability across different land use types. We analyzed 1 year of time-series data from acoustic monitoring stations in Bochum, Germany, calculating psychoacoustic, ecoacoustic, and complex network indices. Our goals were to: (1) identify interdependencies among selected metrics, (2) uncover temporal patterns in acoustic measurements, and (3) relate them to their respective locations. Methods included correlation analysis, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering, principal component analysis, and descriptive statistics with diurnal aggregation. The findings demonstrate that acoustic indices of eight distinct dimensions, along with eight individual metrics, reveal crucial temporal, spatial variations of the acoustic environment and the interplay of individual sound sources overlooked by conventional sound pressure level (SPL) metrics. In particular, the study identifies the maximum Sharpness (Aures method), Link Density, the Bioacoustic Index, and the Amplitude Index as the most effective predictors of land use types, achieving the highest Adjusted Rand Index values (0.25, 0.17, 0.13, 0.13). Incorporating such indices into acoustic monitoring practices offers a refined, site-sensitive framework to identify more nuanced qualities of the acoustic environment, therefore, potentially laying the groundwork for targeted urban interventions that could promote health.