Advanced Noise Indicator Mapping Relying on a City Microphone Network.

Area of Science:

• Environmental Science
• Acoustics
• Urban Planning

Background:

• Accurate city-wide road traffic noise mapping is crucial for urban planning and environmental impact assessments.
• Existing methods often require extensive traffic data, which can be difficult to obtain.
• There is a need for efficient and accessible noise mapping methodologies.

Purpose of the Study:

• To develop and validate a novel methodology for city-wide road traffic noise indicator mapping.
• To bypass the need for direct traffic data access by utilizing street categorization and a microphone network.
• To integrate deterministic modeling with machine learning for improved noise prediction accuracy.

Main Methods:

• A simplified dynamic traffic model was used as a basis for deterministic noise prediction.
• Sound propagation was modeled by combining aspects of the CNOSSOS and QSIDE models.
• An artificial neural network (ANN) was employed to refine deterministic predictions against measured data from a microphone network.

Main Results:

• The methodology achieved prediction accuracy within 2-3 dB of measured noise levels in Barcelona.
• The number of noise events was predicted with approximately 30% accuracy.
• Noise indicators could be accurately predicted at various time scales, including hourly.

Conclusions:

• The presented methodology offers a robust and data-efficient approach to city-wide road traffic noise mapping.
• This technique enables the inclusion of a wide range of noise indicators in environmental noise impact assessments.
• The integration of deterministic models and machine learning shows significant promise for urban acoustic monitoring.