Dimethyl sulfide (DMS), sulfur dioxide, non‐sea‐salt sulfate, and various aerosol properties were measured during three field programs (two airborne and one ground‐based) near Barrow and Deadhorse (Prudhoe Bay), Alaska. The two airborne sampling programs took place in spring and early summer, and the ground‐based measurements spanned an entire summer. DMS concentrations in the Arctic atmosphere ranged from a few parts per trillion by volume (pptv) in spring and fall to higher values in summer (generally a few tens of pptv with occasional peaks of 100 to 300 pptv). In addition, DMS concentrations were measured during the spring near Resolute in seawater below the ice and in ice‐algae and kelp cultures. The seawater samples taken from below the ice in spring had DMS concentrations comparable to those in other oceanic regions. Taken together, these measurements show that the Arctic Ocean is potentially a substantial source of DMS, which likely becomes important as sea ice melts in the early summer. Local atmospheric concentrations increased throughout the summer, peaking in August. In regions where accumulation mode aerosols have been scavenged (e.g., by low‐level stratus clouds, which are common during the Arctic summer), evidence of rapid new particle production was observed. The seasonal cycle of atmospheric DMS closely resembles that of fine particles observed at Barrow, Alaska, and Alert, Northwest Territories, Canada. This finding indicates that DMS is likely an important precursor to the types of particles that dominate the background arctic aerosol in summertime. These results, together with those from several recently published studies of arctic aerosol, are combined to yield a consistent picture of the role of locally emitted DMS in the production of atmospheric aerosols in the Arctic in summer.

Airborne measurements in the srnoke from the Kuwait oil fires in May and June 1991 indicate that the combined oil and gas emissions were equivalent to the consumption of about 4.6 million barrels of oil per day. The combustion was relatively efficient, with about 96% of the fuel carbon burned emitted as CO2. Particulate smoke emissions averaged 2% of the fuel burned, of which about 20% was soot. About two‐thirds of the mass of the smoke was accounted for by salt, soot, and sulfate. The salt most likely originated from oil field brines, which were ejected from the wells along with the oil. The salt accounts for the fact that many of the plumes were white. SO2 and NOx were removed from the smoke at rates of about 6 and 22% per hour, respectively. The high salt and sulfate contents explain why a large fraction of the particles in the smoke were efficient cloud condensation nuclei.

No abstract

Optical properties of the aerosol from the 1991 Kuwait oil fires are calculated using measured aerosol size distributions and a spectral refractive index based on the measured chemical composition of the particulate matter. At a wavelength of 538 nm the calculated light-scattering coefficient agrees well with measurements, but the calculated single-scattering albedo is systematically higher by about 18% than the measured value. Radiative transfer calculations indicate maximum net daytime heating rates of 94 and 56 K d -• for smoke 1 and 3 hours downwind of the fires, respectively. In the upper regions of the plume, where the calculated heating rates decrease with height, a radiative-convective mixed layer developed. There was no significant temperature inversion at the top of this layer, which allowed rapid entrainment of air into the top of the plume, causing it to thicken at an observed rate of ~ 0.1 rn s -•. In addition, radiative heating of the plume as a whole caused it to lift as a unit at a measured rate of ~ 0.1 rn s -• during the first few hours of plume evolution. A theory, based on mixed layer modeling and a scale analysis of the equations of motion, is presented that successfully reproduces the two rates of vertical transport. This model of the dynamics of a radiatively heated plume can be used to predict the evolution and lofting of large composite smoke plumes, such as those from forest fires; it also has implications for the transport, lifetime, and climatic importance of smoke generated on continental scales. 18,80918,810 HERRING AND HOBBS' PLUME DYNAMICS OF KUWAIT OIL FIRES Section 4 contains a brief description of the radiative transfer model that we used to calculate heating rates of the smoke. Vertical profiles of the heating rate are calculated for several different hypothetical smoke distributions to examine how the location and magnitude of the heating was affected by changes in smoke concentration. Heating rates are also calculated for two vertical cross sections of the smoke plume from the Kuwait oil fires and are compared to measurements reported by Pilewsla'e and Valero [ 1992].Solar heating of the smoke from the Kuwait fires played a major role in the dynamical evolution of the smoke plume. In section 5 we examine various theories for the interactions between radiative heating and plume dynamics. Lilly's [1988] analysis of the dynamics of upper tropospheric cirrus clouds heated by terrestrial radiation is found to provide a useful model of the Kuwait smoke plume.One of the most valuable flight patterns conducted with the Convair C-131A was designed to track the smoke plume in a Lagrangian fashion. In section 6 we describe the results from one of these flights in some detail. Not only does this flight provide important insights into the dynamics of the plume but it is also the source of much important data on the chemical evolution of the gas phase and aerosol constituents of the plume. Using our calculated heating rates, we estimate the self-lofting and entrainment rates of the Kuwai...

Total reactive nitrogen oxide (NO,) and NO were measured at the National Oceanic and Atmospheric Administration Geophysical Monitorihg for Climatic Change (NOAA GMCC) site at Barrow, Alaska, for 5 weeks during the spring of 1989. The data demonstrate the existence of a complex array of pollution sources which influence arctic air chemistry. During periods when 10-day back-trajectories indicated southerly flow, median NO, concentrations ranged from 284 to 467 ppt. When the 10-day back-trajectories originated in the Arctic, median NO_ concentrations ranged from 617 to 734 ppt, providing evidence for the long-range transport .r o o o . o of pollutants to the Archc known as 'arctic haze.' The NO. concentrations observed during the spnng ß o o o o J, o background periods are s•gmficanfiy h•gher than prevmus measurements of NO_ at Barrow taken dunng ß o o • j, o summer, regardless of where the a•r mass originated. These observatmns provide evidence for a greatly increased NO. lifetime during the colder months over a large region of the background troposphere as well as the presence of much h•gher NO. concentrations m the Axct•c. During four 12-60 hour events, substantmlly ß o J, .... elevated NO_ were observed •n a•r not •mpacted by enussmns from the town of Barrow. During these four .r o ß o events, peal• NOy concentrahons of 3 to 16 ppb were observed. Other data, including NO, SO 2, SOs, and meteorology, were used to help identify the source region for the elevated NO, levels. These data prosent a consistent picture which indicates that during certain periods, elevated NO. levels are associated with air coming from the Prudhoe Bay industrial area approximately 300 km to the'ESE and demonstrate that this facility is a major source of nitrogen oxides, and possibly other air pollutants, to the Alaskan Arctic.