Tropospheric Ozone Depletion Events in the Arctic Spring of 2009: Modeling and Observations

Ozone is an important atmospheric pollutant in the troposphere due to its high oxidation potential. In the Arctic troposphere, ozone mainly originates from transport and photo-chemical reactions involving nitrogen oxides and volatile organic compounds, resulting in a background mixing ratio of 30 to 50 nmol/mol. During polar spring, so-called tropospheric ozone depletion events (ODEs) are regularly observed, in which ozone mixing ratios in the boundary layer drop to almost zero levels coinciding with a surge in reactive bromine levels on the time scale of hours to days. The source of the reactive bromine is sea salt, i.e. aerosol and deposits on the ice. However, it is not fully understood how the salt bromide is oxidized and reactive bromine is released into the air. The most widely accepted emission mechanism is autocatalytic and termed “bromine explosion”. ODEs strongly change the lifetime of ozone and organic gases, they cause the removal and deposition of mercury as well as the transport of reactive bromine into the free troposphere. In order to model ODEs, the software package WRF-Chem is employed to simulate the meteorology and the emission, the transport, mixing, chemical reactions of trace gases as well as aerosols. For this purpose, the MOZART chemical reaction mechanism coupled with the MOSAIC aerosol model is extended to include bromine and chlorine chemistry. A resolution of 20 km for a 5,000 km x 5,000 km region in horizontal directions is employed, enabling the comparison of the simulation results to satellite GOME-2 BrO with a larger resolution. In vertical direction, 64 non-linear grid cells are used with a finer resolution near the ground. The simulation domain is centered north of Barrow (Utqiaġvik), Alaska and covers most of the Arctic region. The time from February 1 to May 1, 2009 is simulated. Improvements and differences to existing models include more complex bromine chemistry, the inclusion of chlorine chemistry, MOSAIC aerosols, and nudging of meteorological fields to ERA-INTERIM data.The simulations reveal that the first bromine explosions occur in early February in the Bering Sea and then extend to the Beaufort Sea in the middle of February, with further bromine explosions in the Arctic region through the end of the simulation. Simulations results are compared with the GOME-2 BrO measurements and in-situ ozone observations at Barrow (Utqiaġvik), Alaska. The comparison shows good agreement with respect to occurrence and location of ODEs. The simulations indicate that the existence and replenishment of bromine in the sea ice is necessary for the ODEs to occur throughout the observation time. Inclusion of direct release of bromine by the deposition of ozone is essential for the proper prediction of the frequent recurrence of ODEs as found through observations. The largest uncertainty in the model is the strength of the bromine deposition and emission from the ice/snow surface as well as the amount of available bromine in the sea salt, which is varied in a parameter study.