Abstract Number: 43 Working Group: Source Apportionment
Abstract North-western Europe is submitted to frequent high-pollution PM$_(10) episodes. The exceedance of the existing European ambient air quality standards for PM$_(10) represents a serious health risk for populations living in such impacted areas. This is particularly the case for more than 4 million inhabitants living in the French Nord-Pas-de-Calais region, recurrently submitted to PM$_(10) high concentration episodes. Despite the potential influence of various cofactors, the occurrence rates of respiratory and cardiovascular diseases are more elevated among this population compared to the French averages and suggest a possible link with chronic exposure to elevated atmospheric particulate pollution. In northern France, primary sources of particulate matter are related to a heavily industrial sector, dense urbanization, heavy traffic and intensive agriculture. As atmospheric particles can be transported over long-distances, PM$_(10) mass concentrations measured by the Regional Network for Air Quality Monitoring (RNAQM) are due to both local and long-range sources.
In order to help regional policy-makers to take suitable and efficient measures against air pollution to protect public health, the challenging task of our work was to determine whether local sources of atmospheric particles are involved in these frequent exceedances of the PM$_(10) daily limit value.
In a first step, we developed a robust methodology based on a hierarchical clustering to select monitoring stations with distinct distributions of the hourly averaged PM$_(10) mass concentration measured by a dense continuous monitoring network (27 stations over 5 years, 2009-2012). Secondly, a multi-site concentration field analysis was applied to daily PM$_(10) concentrations at selected stations. This statistical method consists in redistributing the concentrations of PM$_(10) to air mass back trajectories in order to identify potential source areas influencing the receptor site. Additionally, we investigated the matches and discrepancies between these resulting potential source areas and the regional 3 km$^2 grid-based emission inventory.