In the present study, both parametric (peak over threshold: mean residual life, dispersion index, threshold choice) and non-parametric (percentiles indices 95% and 99%) statistical techniques are employed, aiming at the identification of rainfall thresholds above which a precipitation event can be characterized as extreme. The analysis is based on 45 years (1960–2004) rain gauge daily records from 65 meteorological stations over the European region. According to two climatologically based criteria that were introduced in the study, it was found that a combined peak over threshold methodology has been shown to yield higher threshold values above which extreme precipitation events occur, in comparison to the 95th percentile indices. Overall, concerning northern Europe, it was found that in the majority of the stations, the threshold values vary from 20 to 30 mm, while the results concerning the Mediterranean region are less coherent and the selection of extreme precipitation thresholds differs from region to region. Stations over eastern Mediterranean appear to have thresholds higher than 30 mm, while stations located over the main cyclone trajectories and the cyclogenesis zone of Mediterranean are those with the higher extreme precipitation thresholds (higher than 45 mm).

Identifying the dominant atmospheric scale of motion over a given airshed can be of great importance for many applications such as air pollutants dispersion, pollen dispersion and wind energy. In this work, a method is proposed for differentiating between the synoptic, meso and local scales that can be applied on a daily basis from data collected at several wind-measuring sites in a given airshed. The method is based on the daily mean quantitative measure of wind recirculation measured at each site, and the mean and standard deviation between sites. The method is applied to a 5-year time series of wind observations at seven sites in the Tel Aviv airshed on the East Mediterranean Sea. The high frequency of synoptic-scale dominant flows is found during winter and December in particular, whereas high dominance of mesoscale flows is found during October. The coastal location of the airshed causes the land–sea breeze to overrule local-scale variations. While NOx and SO2 concentrations are shown to increase as the dominant scale is smaller, O3 concentrations decrease as a result of titration with NOx on the one hand and the contribution of long-range transport on the other.

This study was aimed at evaluating dus t distributions pred icted by modified DREAM model with a more detailed set of ei ght dust particle size classes, in order to better understand the model’s ca pabilities for providing re liable dust forecasts. Provided by the Israeli Mini stry of Environment Protec tion, the PM10 measurements for three different sites in Israel (Bee r-Sheva (31.3N, 34.8E), Tel-Aviv University (32.1N, 34.9E) and Carmiel (32.9N, 35.3E) were used for model evaluation. The period under investigation was that of maximum dust activity in the Eastern Mediterranean, from February to May 2006.