The challenge of identifying dust events in a highly polluted Eastern Mediterranean region.

Dust pollution largely impacts our environment, health and well-being. However, there is no agreement on how dust-contaminated days are identified to study exposures, as methods differ across disciplines. Different quantitative thresholds, which rely on ground measurements, are generally used to define these events. In this study, we used ground-based lidar measurements to detect dust layers. The dataset was then compared to methods that are widely used to define the presence of dust on the ground. Our results show that dust layers extend to a height of up to 10 km and a depth of up to 6.3 km. We show that at least 50 % of days that include dust components according to the lidar were not included by any of the methods that we investigated. As a result, these days are not considered in many health-related studies and climate models. Many dust events exhibit a high anthropogenic component and can be misinterpreted: (Ångström exponent>1.2), high-altitude (on average above 1.7 km) and relatively shallow (average depth 1.4 km) dust layers, and low PM on the ground. Mixed pollution (0.8 < Ångström exponent < 1.2) accounts for 45 % of these events. The most accurate dust-detection method considered the aerosol optical depth and Ångström exponent parameters, and provided 60 % of the dust days as determined by lidar. It does not seem to be possible to differentiate between anthropogenic and dust events because most measurements contained dust, resulting in further biased estimations. Our results indicate that there is a need to change our perception of what constitutes a dust day, when studying the impact of dust exposure. We suggest that in arid and semiarid, and in particular Eastern Mediterranean climates, where dust is a frequent and strong meteorological component, a greater number of days need to be included in the analyses or critically evaluated.