Characterization of synoptic patterns causing dust outbreaks that affect the Arabian Peninsula

Abstract

Dust storms pose serious weather hazards in arid and semiarid regions of the earth. Understanding the main synoptic conditions that give rise to dust outbreaks is important for issuing forecasts and warnings to the public in cases of severe storms. The aim of the present study is to determine synoptic patterns that are associated with or even favor dust outbreaks over the Arabian Peninsula. In this respect, red–green–blue dust composite images from the Meteosat Second Generation (MSG) satellite are used to detect dust outbreaks affecting the Arabian Peninsula, with possible influences in southwestern Asia and northeastern Africa, between 2005 and 2013. The Meteosat imagery yielded a sample of 95 dust storm days. Meteorological fields from NCEP/NCAR reanalysis data of wind fields at 10m and 250hPa, mean sea level pressure, and geopotential heights at 850 and 500hPa were obtained for the dust storm days. Using principal component analysis in T-mode and non-hierarchical k-means clustering, we obtained four major atmospheric circulation patterns associated with dust outbreaks during the study days. Cluster 4 had the largest number of days with dust events, which were constrained to summer, and cluster 3 had the fewest. In clusters 1, 2 and 3, the jet stream favored the entry of a low-pressure area or trough that varied in location between the three clusters. Their most northerly location was found in cluster 4, along with an extensive low-pressure area supporting strong winds over the Arabian Peninsula.The spatial distribution of aerosol optical depth for each cluster obtained was characterized using the Moderate Resolution Imaging Spectroradiometer data. Then, using METAR stations, clusters were also characterized in terms of frequency and visibility.