The understanding of particulate matter (PM) levels over Europe as a whole is at present limited by the diversity of ground level measurements methods. This study explores the relation between satellite observations of PM to the concentrations at ground level. As such it aims to form an initial step for making PM maps with accurate and consistent concentrations.
To mitigate the harmful effect of particulates on human health, EU wide limit values for concentrations have been set. However, particulate matter (PM) measurements suffer from substantial uncertainty because PM is difficult to measure on a routine basis, which is necessary for monitoring compliance. Moreover, different measurement and calibration methods are used in the many air-quality networks in Europe. Consequently, the understanding of PM levels over Europe as a whole is rather limited. We aim to improve this situation by making use of additional information from satellite observations. As a first step, we present here a comparison for Europe of spatio-temporal variations of PM with those of aerosol optical thickness (AOT) measured by the MODIS satellite instrument, for 2003. The MODIS measurements clearly show the major aerosol source regions in Northern Italy, Southern Poland, and the Belgium/Netherlands/Ruhr area, as well as individual large cities and industrialised valleys (Rhone, Danube). The spatial correlation between yearly average PM10 and AOT is 0.6 for rural background stations.
The seasonal variation of AOT and PM is distinctly different, however. Throughout most of Europe, the AOT as measured by MODIS has a clear minimum in the winter months. The seasonal variation in PM differs across Europe, and at many locations the seasonal variation is less marked than that of the AOT. Consequently, the correlation between one-year time-series of AOT with PM10/PM2.5 is low (0.3). The correlation between PM and AOT is improved when the AOT is divided by the boundary layer height and, to a lesser extent, when it is corrected for growth of aerosols with relative humidity. In that case, the average correlation is 0.5 (PM10) and 0.6 (PM2.5), averaged over rural and (sub) urban background stations. These results indicate that AOT measurements can be useful to improve the monitoring of PM distributions over Europe.