Tropospheric reactive bromine (Br_y) influences the oxidation capacity of the atmosphere by acting as a sink for ozone and nitrogen oxides. Aerosol acidity plays a crucial role in Br_y abundances through acid-catalyzed debromination from sea-salt-aerosol, the largest global source. Bromine concentrations in a Russian Arctic ice-core, Akademii Nauk, show a 3.5-fold increase from pre-industrial (PI) to the 1970s (peak acidity, PA), and decreased by half to 1999 (present day, PD). Ice-core acidity mirrors this trend, showing robust correlation with bromine, especially after 1940 (r=0.9). Model simulations considering anthropogenic emission changes alone show that atmospheric acidity is the main driver of Br_y changes, consistent with the observed relationship between acidity and bromine. The influence of atmospheric acidity and Br_y should be considered in interpretation of ice-core bromine trends.

We use a global 3D chemical transport model GEOS-Chem (version 11-02d, https://github.com/geoschem/geos-chem/tree/v11-02d-prelim) for historical simulations. The model is driven by MERRA-2 assimilated meteorological fields from the Goddard Earth Observing System (GEOS) (Gelaro et al., 2017), and contains detailed HO_x-NO_x-VOC-ozone-halogen-aerosol tropospheric chemistry (Wang et al., 2021) and fully coupled stratospheric chemistry (Eastham et al., 2014). Details of the modeled bromine chemistry are shown in Fig. S1‒S3.  Sea-salt-aerosol debromination occurs in both open ocean (Jaeglé et al., 2011) and blowing snow (Huang & Jaeglé, 2017) sourced sea-salt-aerosol. Following Zhai et al. (2023), ozone dry deposition velocity onto snow and ice is updated to 0.01 cm s⁻¹, consistent with observations (Simpson et al., 2007). Snowpack bromine emissions (Swanson et al., 2022; Zhai et al., 2023) are not included in the model.

Model simulations are performed under three anthropogenic emission scenarios: pre-industrial (PI, CE 1750), peak atmospheric acidity (PA, CE 1975), and present-day (PD, CE 1999). Anthropogenic and biomass-burning emissions vary between simulations to reflect their impacts on tropospheric bromine. Details of the emission setup can be found in Zhai et al. (2021), and trends of anthropogenic emissions of SO₂ and NO_x can be found in Fig.S4. After a 1-year spin-up, each simulation is run for 1 year using 2007 meteorology and sea-ice extent. By using the same meteorology, we aim to isolate changes induced by anthropogenic emissions. All simulations are conducted at 4° × 5° horizontal resolution and 72 vertical levels up to 0.01 hPa.