Abstract
Air pollution monitoring and mitigation have become a public and political issue of high priority. According to European Union standards, PM is measured in µm/m3. A measurement that is not sensitive to the presence of the finest given their low masses, while they are more numerous and harmful than the regulated PM. Therefore, an unconventional and alternative approach based on environmental magnetism offers an excellent opportunity to detect the finest particles. This approach consists of measuring the magnetic properties of accumulative surfaces after they have been exposed to polluted air. It allows local mapping of pollutant deposits integrated over the exposure time. But how far can this method go in terms of air quality monitoring?The first part of my thesis made it possible to evaluate the potential of magnetic mapping carried out near a highway by comparing its results with those of a numerical model of pollutant dispersion (CFD). The latter, considered the moment of speed and the turbulence induced by the traffic, which is rarely done in this kind of approach. Mapping of the magnetic parameter exposed an asymmetry in pollutant deposition, where the maximum was located downwind. This surprising observation was confirmed by the numerical model which predicted a recirculation current at this location due to the presence of a noise barrier. This study made it possible to validate the magnetic mapping of pollutants by expressing the values in relative change in order to isolate the source of PM that we want to study.The second part was an opportunity to examine whether magnetism could be used in a study with a citizen science approach and to what extent unconventional data could be used as a lever for political decision-making in a canyon street in Montpellier. Finally, magnetic air quality monitoring seems to be promoted for citizen studies by providing high resolution street-scale data and also be seen as an alternative or complement to conventional monitoring techniques.The third part focuses more fundamentally on what we are able to measure with magnetism since not all pollutants are necessarily magnetic. For this, I magnetically characterized different sources coming from traffic. By combining different techniques, I can say today that the magnetic parameters have a real potential to discriminate between the different sources of pollutant emissions. Finally, in view of the large number of samples, their heterogeneities and the numerous magnetic analyzes carried out, I was not able to directly compare the magnetic properties of the sources with those measured on the accumulative surfaces. Thus, I used Machine Learning with a k-nearest neighbor algorithm to classify the sources into several categories according to their magnetic parameters and to predict the main magnetic source of PM measured on the accumulator surfaces.The last part covers the mitigation of air pollution by urban hedgerows. Today there is still no clear consensus: some affirm their ability to extract these pollutants from the air while others emphasize they hinder the dispersion of pollutants. Using a “homemade” experimental wind tunnel, we carried out several experiments by injecting diesel exhaust into different hedge of endemic species to the south of France to calculate their deposition velocities. This parameter is used to parameterize models on the plant effects on capture or dispersion of pollutants. Unfortunately, we did not find any clear correlation between data from PM sensors and the magnetic measurements.