Caracterització de l’impacte dels esdeveniments acústics en els nivells equivalents sonors i en la percepció dels ciutadans per a la confecció de mapes dinàmics de soroll

Acoustic pollution has become a serious public health problem, causing various types of disease and disorders in people. According to the World Health Organisation, one million years of healthy life are lost in Western Europe every year due to exposure to environmental noise.
 
In order to evaluate and manage environmental noise in the European Union, Directive END 2002/49/EC requires Member States to prepare and publish updated noise maps and relative action plans every five years. This includes agglomerations of more than 100,000 inhabitants and major roads, train tracks and airports.
Thanks to recent technological advances, the noise map creation paradigm has changed substantially, allowing noise level measurements to be automated using wireless acoustic sensor networks for real-time noise map generation. However, these networks cannot prevent a series of situations that would bias the actual measurement of sound equivalent levels, causing the map not to be true to the reality perceived by the citizen, e.g., the sound of birds, the industry, the claxons, the mermaids, conversations that occur near sensors or weather phenomena such as rain and wind.
 
This thesis studies the characterization of acoustic events for the tailoring of dynamic traffic noise maps. The study begins by presenting the context of the thesis, the LIFE DYNAMAP project, which aims to measure traffic noise levels in two pilot areas and dynamically integrate them into a noise map that is updated in real time. After that, a detailed analysis is presented for the events in the two areas, urban and suburban, and various characterizations are applied. One of the presented measures is the impact on the equivalent sound level (L_eq), which allows the measurement of bias resulting from the presence of certain acoustic events in the making of traffic noise maps. The use of perceptual tests using psychoacoustic metrics is also considered in order to adapt the characterization of these events to citizen perception.
 
The main purpose of the thesis is to characterize the events of urban and suburban environments to offer noise maps more faithful to the reality perceived by the citizen in relation to the sound environment where it is found. And during the thesis, the importance of sound detection on a network of acoustic sensors is shown in order to prevent measurement errors at equivalent levels and the need to train the detection system with data obtained from the same sensors on the network.