2026/03/01 by Shane V. Lympany, Michael M. James, Matthew F. Calton +4 · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · Health Professions · #Animal Vocal Communication and Behavior #Music and Audio Processing #Noise Effects and Management
paper · doi:10.1121/10.0042979
openalex publication_date 2026/03/01 · openalex created_date 2026/03/17 · openalex updated_date 2026/06/11
This paper presents ambient | global, an ambient soundscape model developed to predict global ambient sound levels from all anthropogenic, biological, and geophysical sources. The soundscape model adopts a geospatial approach by modeling the ambient sound level as a function of geospatial features at a location. The soundscape model consists of an ensemble of four machine learning regression models fitted at acoustic measurement sites where both the geospatial features and ambient sound levels are known. The fitted model is then applied to predict ambient sound levels at any location where the geospatial features are known. The results quantify the spatial, temporal, and spectral patterns of ambient sound levels across the world under various scenarios. This paper presents maps of the existing ambient sound levels across the world in terms of the daytime overall A-weighted L50, or median sound level, and partitions the existing sound levels into their natural and anthropogenic constituents. Ultimately, the soundscape model will enable research into the impacts of humans and nature on the ambient soundscape and the impacts of ambient sound levels on humans and nature across the world.