Natural marine bromoform emissions in the fully coupled ocean–atmosphere model NorESM2

Abstract

Oceanic bromoform (CHBr3) is an important precursor of atmospheric bromine. Although highly relevant for the future halogen burden and ozone layer in the stratosphere, global CHBr3 production in the ocean and its emissions are still poorly constrained in observations and are mostly neglected in climate models. Here, we newly implement marine CHBr3 in the second version of the state-of-the-art Norwegian Earth System Model (NorESM2) with fully coupled interactions of ocean, sea ice, and atmosphere. Our results are validated using oceanic and atmospheric observations from the HalOcAt (Halocarbons in the Ocean and Atmosphere) database. The simulated mean oceanic concentrations (6.61 ± 3.43 pmol L−1) are in good agreement with observations from open-ocean regions (5.02 ± 4.50 pmol L−1), while the mean atmospheric mixing ratios (0.76 ± 0.39 ppt) are lower than observed but within the range of uncertainty (1.45 ± 1.11 ppt). The NorESM2 ocean emissions of CHBr3 (214 Gg yr−1) are within the range of or higher than previously published estimates from bottom-up approaches but lower than estimates from top-down approaches. Annual mean fluxes are mostly positive (sea-to-air fluxes); driven by oceanic concentrations, sea surface temperature, and wind speed; and dependent on season and location. During winter, model results imply that some oceanic regions in high latitudes act as sinks of atmospheric CHBr3 due to their elevated atmospheric mixing ratios. We further demonstrate that key drivers for oceanic and atmospheric CHBr3 variability are spatially heterogeneous. In the tropical West Pacific, which is a hot spot for oceanic bromine delivery to the stratosphere, wind speed is the main driver for CHBr3 fluxes on an annual basis. In the North Atlantic, as well as in the Southern Ocean region, atmospheric and oceanic CHBr3 variabilities interact during most of the seasons except for the winter months, when sea surface temperature is the main driver. Our study provides an improved process-based understanding of the biogeochemical cycling of CHBr3 and more reliable natural emission estimates, especially on seasonal and spatial scales, compared to previously published model estimates.