is a new modeling system that produces global forecasts of atmospheric composition at 25km 2 horizontal resolution. • GEOS-CF model output is freely available and offers a new tool for academic researchers, air quality managers, and the public.

We quantify the stratospheric injection of brominated very short‐lived substances (VSLS) based on aircraft observations acquired in winter 2014 above the Tropical Western Pacific during the CONvective TRansport of Active Species in the Tropics (CONTRAST) and the Airborne Tropical TRopopause EXperiment (ATTREX) campaigns. The overall contribution of VSLS to stratospheric bromine was determined to be 5.0 ± 2.1 ppt, in agreement with the 5 ± 3 ppt estimate provided in the 2014 World Meteorological Organization (WMO) Ozone Assessment report (WMO 2014), but with lower uncertainty. Measurements of organic bromine compounds, including VSLS, were analyzed using CFC‐11 as a reference stratospheric tracer. From this analysis, 2.9 ± 0.6 ppt of bromine enters the stratosphere via organic source gas injection of VSLS. This value is two times the mean bromine content of VSLS measured at the tropical tropopause, for regions outside of the Tropical Western Pacific, summarized in WMO 2014. A photochemical box model, constrained to CONTRAST observations, was used to estimate inorganic bromine from measurements of BrO collected by two instruments. The analysis indicates that 2.1 ± 2.1 ppt of bromine enters the stratosphere via inorganic product gas injection. We also examine the representation of brominated VSLS within 14 global models that participated in the Chemistry‐Climate Model Initiative. The representation of stratospheric bromine in these models generally lies within the range of our empirical estimate. Models that include explicit representations of VSLS compare better with bromine observations in the lower stratosphere than models that utilize longer‐lived chemicals as a surrogate for VSLS.

Bromine radicals (Br + BrO) are important atmospheric species owing to their ability to catalytically destroy ozone as well as their potential impacts on the oxidative pathways of many trace gases, including dimethylsulfide and mercury. Using space-based observations of BrO, recent studies have reported rapid enhancements of tropospheric BrO over large areas (so called "BrO explosions") connected to near-surface ozone depletion occurring in polar spring. However, the source(s) of reactive bromine and mechanism(s) that initiate these BrO explosions are uncertain. In this study, we investigate the relationships between Arctic BrO explosions and two of the proposed sources of reactive bromine: sea-salt aerosol (SSA) generated from blowing snow and first-year (seasonal) sea ice. We use tropospheric column BrO derived from the Ozone Monitoring Instrument (OMI) in conjunction with the Goddard Earth Observing System Version 5 (GEOS-5) data assimilation system provided by National Aeronautics and Space Administration Global Modeling and Assimilation Office. Case studies demonstrate a strong association between the temporal and spatial extent of OMI-observed BrO explosions and the GEOS-5 simulated blowing snow-generated SSA during Arctic spring. Furthermore, the frequency of BrO explosion events observed over the 11-year record of OMI exhibits significant correlation with a time series of the simulated SSA emission flux in the Arctic and little to no correlation with a time series of satellite-based first-year sea ice area. Therefore, we conclude that SSA generated by blowing snow is an important factor in the formation of the BrO explosion observed from space during Arctic spring. Key Points: • OMI-observed Arctic tropospheric BrO explosions are associated with GEOS-5 simulated sea-salt aerosols (SSAs) from wind-driven blowing snow • The 11-year time series analysis shows that the Arctic BrO explosion frequency and simulated blowing snow SSA emission flux are correlated • The satellite-based record of first-year sea ice area does not show significant correlation with the Arctic BrO explosion frequency Supporting Information: • Supporting Information S1

Abstract. We report measurements of bromine monoxide (BrO) and use an observationally constrained chemical box model to infer total gas-phase inorganic bromine (Br y ) over the tropical western Pacific Ocean (tWPO) during the CON-TRAST field campaign (January-February 2014). The observed BrO and inferred Br y profiles peak in the marine boundary layer (MBL), suggesting the need for a bromine source from sea-salt aerosol (SSA), in addition to organic bromine (CBr y ). Both profiles are found to be C-shaped with local maxima in the upper free troposphere (FT). The median tropospheric BrO vertical column density (VCD) was measured as 1.6×10 13 molec cm −2 , compared to model predictions of 0.9 × 10 13 molec cm −2 in GEOS-Chem (CBr y but no SSA source), 0.4 × 10 13 molec cm −2 in CAM-Chem (CBr y and SSA), and 2.1×10 13 molec cm −2 in GEOS-Chem (CBr y and SSA). Neither global model fully captures the Cshape of the Br y profile. A local Br y maximum of 3.6 ppt (2.9-4.4 ppt; 95 % confidence interval, CI) is inferred between 9.5 and 13.5 km in air masses influenced by recent convective outflow. Unlike BrO, which increases from the convective tropical tropopause layer (TTL) to the aged TTL, gas-phase Br y decreases from the convective TTL to the aged TTL. Analysis of gas-phase Br y against multiple tracers (CFC-11, H 2 O / O 3 ratio, and potential temperature) reveals a Br y minimum of 2.7 ppt (2.3-3.1 ppt; 95 % CI) in the aged TTL, which agrees closely with a stratospheric injection of 2.6 ± 0.6 ppt of inorganic Br y (estimated from CFC-11 correlations), and is remarkably insensitive to assumptions about heterogeneous chemistry. Br y increases to 6.3 ppt (5.6-7.0 ppt; 95 % CI) in the stratospheric "middleworld" and 6.9 ppt (6.5-7.3 ppt; 95 % CI) in the stratospheric "overworld". The local Br y minimum in the aged TTL is qualitatively (but not quantitatively) captured by CAM-Chem, and suggests a more complex partitioning of gas-phase and aerosol Br y species than previously recognized. Our data provide corroborating evidence that inorganic bromine sources (e.g., SSA-derived gas-phase Br y ) are needed to explain the gas-phase Br y budget in the upper free troposphere and TTL. They are also consistent with observations of significant bromide in Upper Troposphere-Lower Stratosphere aerosols. The total Br y budget in the TTL is currently not closed, because of the lack of concurrent quantitative measurements of gas-phase Br y species (i.e., BrO, HOBr, HBr, etc.) and aerosol bromide. Such simultaneous measurements are needed to (1) quantify SSA-derived Br y in the upper FT, (2) test Br y partitioning, and possibly explain the gas-phase Br y minimum in the aged TTL, (3) constrain heterogeneous reaction rates of bromine, and (4) account for all of the sources of Br y to the lower stratosphere.