Analysis of atmospheric VSLS measurements during the DRIVE campaign in the tropical East Atlantic

This diploma thesis investigates variations in mixing ratios of very short lived substances (VSLS) during the DRIVE (Diurnal and Regional Variability of halogen Emissions) campaign in the tropical East Atlantic conducted during P399 2/3 from May to June 2010. The main focus lies on the diurnal and regional variability of three halocarbons influenced by meteorological factors at six 24 h stations. For this reason, regular ship measurements of temperature, wind, air pressure and humidity were complemented by radiosonde launches and air samples for the trace gas investigation. According to the radiosonde measurements a changeover between tropical and extra tropical air masses is observed at about 30° N. In contrast to the dominating trade wind regime with northeasterly winds, the ship cruise was mainly exposed to north-northwesterly winds with moderate weather conditions. In addition, the height of the atmospheric boundary layer is determined from the atmospheric profile, ranging from 500 m – 1700 m over open ocean. At coastal areas, especially at the Mauritanian upwelling, the boundary layer stays at the surface. For an evaluation of the wind measurements, wind speed and direction are compared to a high resolution ERA-Interim data set and, to NCEP/NCAR Reanalysis Project 1 Data (NNRP-1). In situ ship measurements show higher correlations for the wind speed (ERA-Interim: r = 0.91, NNRP-1: r = 0.79) than for the wind direction (ERA-Interim: r = 0.69, NNRP-1: r = 0.67). Recalculating the correlation coefficients for the same temporal resolution (6-hourly) results in an improvement of the correlations, indicating a good agreement between the observed and simulated wind. For the determination of potential source regions, the trajectory model HYSPLIT is used to investigate the origin of the observed air masses. The air mass history is analyzed 315 h backwards, indicating an air mass origin mainly above the North Atlantic during leg 2 but also above the Arctic Ocean during leg 3. During the 24 h stations close to the coast predominantly younger and local sources seem to influence the observed trace gas mixing ratios (mean CH2Br2/CHBr3 ratio: 0.4) with a partly strong dependency on the wind direction (r > 0.81). Especially the methyl iodide mixing ratios show a strong relation to air masses originating at the Banc d’Arguin National Park. In contrast to that, bromoform and dibromomethane show an increase of the mixing ratios connected to trajectories passing the coastal areas of Mauritania and Western Sahara. Increased mixing ratios of the observed trace gases concentrations with highest values of 8.9 ppt (bromoform), 3.31 ppt (dibromomethane) and 1.25 ppt (methyl iodide) are observed in combination with lowest boundary layer heights. With correlation coefficients of r = -0.90 (bromoform), r = -0.91 (dibromomethane) and r = -0.63 (methyl iodide) especially the mixing ratios of the longer lived bromocarbons seem to be strongly connected to the boundary layer height. However, the highest atmospheric bromoform mixing ratio of 9.8 ppt was not observed at the Mauritanian upwelling (leg 2), but close to Lisbon (leg 3) with the Rio Tejo as a potential source region. This implies the significance of regional sources as well as the consideration of the current lower atmospheric state for the variations of the VSLS mixing ratio