Environmental drivers of tropospheric bromine and mercury variability in coastal East Antarctica

Bromine radicals released from sea ice, snow, and marine sources play a critical role in the atmospheric chemistry of polar regions. The Chemical and Mesoscale Mechanisms of Polar Cell Aerosol Nucleation (CAMMPCAN) ship campaign conducted in coastal East Antarctica over two 6-month periods in 2017–18 and 2018–19 provides a unique dataset to identify the environmental drivers of bromine variability in Antarctic spring and summer. In this study, we used CAMMPCAN chemical and meteorological observations combined with reanalysis data from the Modern Era Retrospective Analysis for Research and Applications version 2 (MERRA-2) and satellite-based sea ice data from the National Snow and Ice Data Center to select variables that showed statistically significant correlation with bromine monoxide (BrO) partial columns measured during CAMMPCAN. We then used those variables in principal component analysis and subsequent principal component regression to identify dominant modes of Antarctic environmental variability and their impacts on lower tropospheric BrO. Comparing our three dominant Antarctic principal components to those from a similar analysis conducted previously for the Arctic (Swanson et al., 2020), we found only one mode with clear overlap, representing a vertical mixing mode in which low-pressure systems mix BrO and its precursors into the lower troposphere. We also identified an Antarctic mode describing conditions favourable for blowing snow, similar to the combined effect of two modes from the Arctic analysis but more clearly disambiguated here due to the inclusion of sea ice data in our analysis. The third Antarctic mode, attributed to an ocean source (biological activity and/or sea salt aerosol), was particularly important in summer. The principal component regression model developed from these modes showed moderate skill in predicting BrO partial columns in the lowest 2 km of the troposphere (R = 0.51), a significant improvement over the Arctic-based regression model (R = 0.08). Neither model could reproduce the observed variability in BrO in the lowest 200 m. Finally, we applied the same analysis to coincident CAMMPCAN observations of gaseous elemental mercury and found regression of our three dominant modes could explain nearly 50% of observed mercury variability (R = 0.69). Our results reinforce the importance of sea ice and ocean processes in bromine cycling in coastal East Antarctica and highlight the need to consider Antarctic-specific processes in mechanistic models of atmospheric bromine chemistry.