Using a Clustering Method to Detect Spatial Events in a Smartphone-Based Crowd-Sourced Database for Environmental...

Area of Science:

• Environmental Science
• Data Science
• Urban Planning

Background:

• Noise pollution significantly impacts health, particularly in urban environments.
• Accurate noise evaluation through tools like noise mapping is crucial for mitigation.
• Classical noise mapping methods have limitations; smartphone-based solutions offer a low-cost alternative.

Purpose of the Study:

• To develop open-science tools for noise mapping via the Noise-Planet project.
• To address data quality issues in smartphone-collected noise data.
• To identify and integrate data from external communities into the NoiseCapture (NC) database.

Main Methods:

• Utilized the open-source smartphone tool "NoiseCapture" for noise measurement and sharing.
• Applied the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm.
• Tested DBSCAN on existing NC party data and then on a global scale.

Main Results:

• The DBSCAN algorithm successfully identified clusters representing organized noise measurement events (NC parties).
• The method demonstrated effectiveness in detecting these events within the larger NC database.
• Different cluster typologies were detected based on DBSCAN parameter variations.

Conclusions:

• DBSCAN is a viable method for identifying high-quality, spatially and temporally dense noise measurement data.
• This approach enhances the creation of comprehensive noise maps by integrating diverse datasets.
• The findings support the improvement of data analysis and the relevance of noise mapping initiatives.