Une atmosphère saine est un besoin élémentaire pour le bien être et la santé humaine. La matière particulaire en suspension (Particle Matter, PM) est bien connue pour avoir un impact significatif sur la santé. Les mesures de PM2.5 et PM10 au niveau du sol reflètent l’influence de la dynamique de la couche limite et du mélange des aérosols locaux ou advectés sur de grandes distances. Le lien entre épaisseur optique en aerosol (aerosol optical thickness, AOT) et PM dépend de la relation entre propriétés optiques et massiques et de la distribution verticale des particules dans l’atmosphère. Nous présentons 3 expériences de terrain dédiées à la caractérisation des aérosols de pollution dans le Nord de la France: la première lors d’un évenement de pollution printanier sur Lille, la seconde durant un événement de pollution hivernal sur Dunkerque et la troisième durant des occurrences de brise de mer sur le littoral Dunkerquois. Nous avons utilisé 2 systèmes Lidar différents, le premier dans le visible (532 nm) et le second dans l’UV (355 nm); un photomètre solaire automatique et des mesures de PM2.5 et PM10 par TEOM. L’altitude supérieure de la couche de mélange (Mixed boundary layer, MBL) est détectée par Lidar et nous avons été capable de suivre le développement classique de la couche limite convective ainsi que des décroissances brutales d’altitude de la MBL dues à la brise de mer. Les profils d’extinction aérosols ont été estimés en utilisant un rapport Lidar de 67 sr à 532 nm à Lille, 77 sr à 532 nm et 30 sr à 355 nm à Dunkerque. Nous avons analysé l’impact du transport grande échelle de masses d’air polluée, du développement convectif de la MBL et du développement de la cellule de brise de mer sur les profils verticaux d’extinction en aérosols. Le signal Lidar dans les premières centaines de mètres est très bien corrélé (coefficient de corrélation supérieur à 0.9) avec les concentrations massiques mesurées au sol dans tous les cas. Il est également montré que l’introduction de la hauteur de la MBL permet une meilleure détermination des PM à partir de l’épaisseur optique. / Clean air is considered to be a basic requirement for human health and well-being. Particulate matter is known to have a significant impact on health. The variability of Particle Matter (PM2.5 and PM10) concentrations recorded at ground-level is influenced by the boundary layer dynamics, local emissions, and advection and mixing of large scale transported aerosols. The link between columnar aerosol optical thickness (AOT) and ground-level PM depends on the relationship between mass and optical properties and on the vertical distribution of aerosols in the atmosphere. We present three field experiments dedicated to the characterization of pollution aerosols in the North of France: the first one during a spring pollution episode in metropolitan area of Lille (50.61°N, 3.14°E), the second one during a winter pollution episode in the industrial coastal city of Dunkerque (51°04'N; 2°38'E) and the third one during summer sea breezes on coastal area of Dunkerque. We have used 2 different Lidar systems, one in the UV (355 nm) and the other one in the visible (532 nm), an automatic sun photometer, and PM2.5 and PM10 measurements with TEOM. The mixed layer (MBL) top altitude is detected from the Lidar signal and we were able to monitor the classical diurnal evolution of the convective continental boundary as well as short-time decreases in the MBL height due to sea breeze occurrences. The aerosol extinction profiles were estimated using a Lidar ratio of 67 sr at 532 nm in Lille, and 77 sr at 532 nm and 30 sr at 355 m in Dunkerque. We have analyzed the impact of long range transport of polluted air masses, convective development of the MBL, and sea breeze development on the vertical profile of aerosol extinction coefficient. The Lidar signal in the first few hundred meters is well correlated (correlation coefficient above 0.9) with the PM concentrations in all cases. It is found that introducing the Lidar derived MBL height enable a better estimation of PM from measured AOT. Clean air is considered to be a basic requirement for human health and well-being. Particulate matter is known to have a significant impact on health. The variability of Particle Matter (PM2.5 and PM10) concentrations recorded at ground-level is influenced by the boundary layer dynamics, local emissions, and advection and mixing of large scale transported aerosols. The link between columnar aerosol optical thickness (AOT) and ground-level PM depends on the relationship between mass and optical properties and on the vertical distribution of aerosols in the atmosphere. We present three field experiments dedicated to the characterization of pollution aerosols in the North of France: the first one during a spring pollution episode in metropolitan area of Lille (50.61°N, 3.14°E), the second one during a winter pollution episode in the industrial coastal city of Dunkerque (51°04'N; 2°38'E) and the third one during summer sea breezes on coastal area of Dunkerque. We have used 2 different Lidar systems, one in the UV (355 nm) and the other one in the visible (532 nm), an automatic sun photometer, and PM2.5 and PM10 measurements with TEOM. The mixed layer (MBL) top altitude is detected from the Lidar signal and we were able to monitor the classical diurnal evolution of the convective continental boundary as well as short-time decreases in the MBL height due to sea breeze occurrences. The aerosol extinction profiles were estimated using a Lidar ratio of 67 sr at 532 nm in Lille, and 77 sr at 532 nm and 30 sr at 355 m in Dunkerque. We have analyzed the impact of long range transport of polluted air masses, convective development of the MBL, and sea breeze development on the vertical profile of aerosol extinction coefficient. The Lidar signal in the first few hundred meters is well correlated (correlation coefficient above 0.9) with the PM concentrations in all cases. It is found that introducing the Lidar derived MBL height enable a better estimation of PM from measured AOT.