A comparative study on VGI and professional noise data

Abstract

The ubiquitous nature of mobile devices and its growing presence in urban areas, turn them up into low cost environmental monitoring platforms. In this field, several authors made different efforts to provide alternatives to Sensor Networks, to assess noise pollution in cities using crowdsourcing techniques. In this sense, citizens might potentially produce large spatio-temporal datasets using their mobile devices to measure noise levels. There are few attempts of assessing the quality of the mobile noise samples on a real scenario and compare them to commercial data to evaluate if they are reliable enough. This contribution reviews the existing applications to collect or assess the quality of noise samples when they are used as sound level meters. Moreover, it presents the results of our experiment: the volunteer noise dataset generated in a ‘mapping party’ on our campus is compared to professional data. Results show that VGI data might be sufficient for multiple daily situations.