Impacts of bromine and iodine chemistry on tropospheric OH and HO 2 : comparing observations with box and global model perspectives

The chemistry of the halogenspecies bromine and iodine has a range of impacts on tropospheric composition, and can affect oxidising capacity in a number of ways. However, recent studies disagree on the overall sign of the impacts of halogenson the oxidising capacity of the troposphere. We present simulations of OH and HO 2 radicals for comparison with observations made in the remote tropical ocean boundary layer during the Seasonal Oxidant Study at the Cape VerdeAtmospheric Observatory in 2009. We use both a constrained box model, using detailed chemistry derived from the Master Chemical Mechanism (v3.2), and the three-dimensional global chemistry transport model GEOS-Chem. Both model approaches reproduce the diurnal trends in OH and HO 2 . Absolute observed concentrations are well reproduced by the box modelbut are overpredicted by the global model, potentially owing to incomplete consideration of oceanic sourced radical sinks. The two models, however, differ in the impacts of halogenchemistry. In the box model, halogenchemistry acts to increase OH concentrations (by 9.8 % at midday at the Cape VerdeAtmospheric Observatory), while the global model exhibits a small increase in OH at the Cape VerdeAtmospheric Observatory (by 0.6 % at midday) but overall shows a decrease in the global annual mass-weighted mean OH of 4.5 %. These differences reflect the variety of timescales through which the halogensimpact the chemical system. On short timescales, photolysis of HOBr and HOI, produced by reactions of HO 2 with BrO and IO, respectively, increases the OH concentration. On longer timescales, halogen-catalysed ozone destruction cycles lead to lower primary production of OH radicals through ozone photolysis, and thus to lower OH concentrations. The global model includes more of the longer timescale responses than the constrained box model, and overall the global impact of the longer timescale response (reduced primary production due to lower O 3 concentrations) overwhelms the shorter timescale response (enhanced cycling from HO 2 to OH), and thus the global OH concentration decreases. The Earth systemcontains many such responses on a large range of timescales. This work highlights the care that needs to be taken to understand the full impact of any one process on the system as a whole.