Towards Measuring and Forecasting Noise Exposure at the VELTINS-Arena in Gelsenkirchen, Germany

Urban environments often subject residents to noise levels that pose significant health risks, including hearing loss, sleep disturbances, and decreased cognitive function. To address these concerns, we deployed five sensor devices for acoustic monitoring to measure and analyze noise around the VELTINS-Arena in Gelsenkirchen, Germany, a major venue for sports and concert events. This paper presents a detailed analysis of long-term noise data collected via installed acoustic sensors, focusing on identifying noise patterns and predicting future noise levels, as well as investigating the dependence of noise exposure on audible sound sources. Our research evaluates various time series modeling techniques, both traditional and deep learning-based, to forecast noise levels effectively. The study identified repetitive patterns in urban noise levels influenced by daily rush hour traffic and spectator movements, while local weather conditions, especially wind, significantly impacted noise measurements. Event-related noise, particularly from speech, cheering, and mechanical noises, varied significantly with the type of event and sensor placement, highlighting the need for strategic sensor positioning.