Award: OPP-1854462

Mercury (Hg) poses serious risks to environmental and human health. The tendency for mercury to bioaccumulate and bio magnify in the food web is of great concern for the well-being of wildlife and people, and especially for human populations in the Arctic that rely on subsistence fishing and hunting. Air-sea exchange is the dominant source and the major sink for oceanic mercury. The loss of elemental Hg (Hg0) from the surface ocean to the atmosphere prolongs the lifetime of Hg in the biosphere but also mitigates the buildup of Hg in ocean waters, and therefore the concentrations of toxic methylmercury (MeHg) in seafood. Mercury concentrations and speciation within surface waters of the Arctic Ocean are controlled by a complex set of processes including photochemical and microbial transformations, redox reactions, and air-sea exchange of gaseous and particulate Hg species. Previous studies have provided new insights into the factors controlling the deposition and evasion of Hg across the air-sea interface, but the role of air-sea exchange in mitigating Hg inputs to the Arctic Ocean, as well as mediating MeHg levels, is not well understood. In this study, our aim was to estimate the magnitude of volatile Hg fluxes across the air-sea interface and examine the influence of ice cover on this process. By measuring gas fluxes, the relative impacts of chemical and biological processes on mercury distributions within the surface waters can then be deduced. But, while gas exchange velocities (k) in the open ocean have been modeled as a function of wind speed, the parameterization is problematic in the presence of sea ice, which can physically block gas exchange, as well as reduce fetch and dampen waves. Our approach was to use measurements of the inert noble gas Radon-222 to accurately measure k under conditions of varying ice cover, and then apply these velocities to estimate gaseous Hg fluxes across the air-sea interface. Field work was conducted on a cruise aboard R/V Sikuliaq originating in Dutch Harbor on 20 May 2021 and terminating in Seward, Alaska on 14 June 2021. The timing and cruise track allowed for sampling locations at different stages of melt: from fully ice-free conditions on the Bering Sea shelf through the Bering Strait to the retreating ice edge at ~70N on the Chukchi shelf. We found good agreement between the gas transfer velocities estimated from Radon deficits and those calculated from empirical wind speed parameterizations, especially considering that the average relative uncertainty associated with the Radon deficit ks amounted to 30% and the wind speed parameterizations carry an uncertainty of at least 20%. Of the six stations in the Bering Sea exhibiting a Radon deficit, four of the Radon-based gas transfer velocities agree within analytical uncertainty with values predicted by the empirical relationships. The two exceptions were from a) a location influenced by Yukon River outflow, and b) a location where the lack of water column stratification allowed for input of Radon from deeper waters. Turning to Chukchi Sea stations, there was also excellent agreement of Rn-based piston velocities with those predicted from wind speeds, and there was no apparent correlation of k with ice cover. This suggests that sea-ice had a negligible effect on sea-air gas exchange at locations within 10 km of the ice edge. Elemental mercury concentrations in seawater (measured by colleagues from the University of Connecticut) were not significantly higher at the ice covered locations sampled in this study. While ice cover may play an important role in the retention of Hg0 in the upper water column, our results suggest that the accumulation of Hg0 under the ice may need a longer amount of time (i.e. more time under the ice and further away from the ice edge where air-sea exchange processes are more active). Last Modified: 04/03/2024 Submitted by: MarkPStephens