Radical chemistry and ozone production at a UK coastal receptor site

Robert Woodward-Massey, Roberto Sommariva, Lisa K. Whalley*, Danny R. Cryer, Trevor Ingham, William J. Bloss, Stephen M. Ball, Sam Cox, James D. Lee, Chris P. Reed, Leigh R. Crilley, Louisa J. Kramer, Brian J. Bandy, Grant L. Forster, Claire E. Reeves, Paul S. Monks, Dwayne E. Heard*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


OH, HO2, total and partially speciated RO2, and OH reactivity (kOH′) were measured during the July 2015 ICOZA (Integrated Chemistry of OZone in the Atmosphere) project that took place at a coastal site in north Norfolk, UK. Maximum measured daily OH, HO2 and total RO2 radical concentrations were in the range 2.6-17 × 106, 0.75-4.2 × 108 and 2.3-8.0 × 108 molec. cm-3, respectively. kOH′ ranged from 1.7 to 17.6 s-1, with a median value of 4.7 s-1. ICOZA data were split by wind direction to assess differences in the radical chemistry between air that had passed over the North Sea (NW-SE sectors) and that over major urban conurbations such as London (SW sector). A box model using the Master Chemical Mechanism (MCMv3.3.1) was in reasonable agreement with the OH measurements, but it overpredicted HO2 observations in NW-SE air in the afternoon by a factor of 1/4 2-3, although slightly better agreement was found for HO2 in SW air (factor of 1/4 1.4-2.0 underprediction). The box model severely underpredicted total RO2 observations in both NW-SE and SW air by factors of 1/4 8-9 on average. Measured radical and kOH′ levels and measurement-model ratios displayed strong dependences on NO mixing ratios, with the results suggesting that peroxy radical chemistry is not well understood under high-NOx conditions. The simultaneous measurement of OH, HO2, total RO2 and kOH′ was used to derive experimental (i.e. observationally determined) budgets for all radical species as well as total ROx (i.e. OH + HO2 + RO2). In NW-SE air, the ROx budget could be closed during the daytime within experimental uncertainty, but the rate of OH destruction exceeded the rate of OH production, and the rate of HO2 production greatly exceeded the rate of HO2 destruction, while the opposite was true for RO2. In SW air, the ROx budget analysis indicated missing daytime ROx sources, but the OH budget was balanced, and the same imbalances were found with the HO2 and RO2 budgets as in NW-SE air. For HO2 and RO2, the budget imbalances were most severe at high-NO mixing ratios, and the best agreement between HO2 and RO2 rates of production and destruction rates was found when the RO2 + NO rate coefficient was reduced by a factor of 5. A photostationary-steady-state (PSS) calculation underpredicted daytime OH in NW-SE air by 1/4 35 %, whereas agreement (1/4 15 %) was found within instrumental uncertainty (1/4 26 % at 2σ) in SW air. The rate of in situ ozone production (P(Ox)) was calculated from observations of ROx, NO and NO2 and compared to that calculated from MCM-modelled radical concentrations. The MCM-calculated P(Ox) significantly underpredicted the measurement-calculated P(Ox) in the morning, and the degree of underprediction was found to scale with NO.

Original languageEnglish
Pages (from-to)14393-14424
Number of pages32
JournalAtmospheric Chemistry and Physics
Issue number22
Publication statusPublished - 21 Nov 2023

Bibliographical note

Funding Information:
We thank the science team of the ICOZA project. Robert Woodward-Massey and Danny R. Cryer are grateful to the NERC for funding PhD studentships. Robert Woodward-Massey, Lisa K. Whalley, Danny R. Cryer, Trevor Ingham and Dwayne E. Heard would like to thank the University of Leeds' electronic and mechanical workshops. Robert Woodward-Massey is grateful to Hans Osthoff (University of Calgary) for the provision of Igor functions and to Chunxiang Ye (Peking University) and Samuel Seldon (University Of Leeds) for useful discussions. We thank Lloyd Hollis and Roland Leigh (University of Leicester) for assistance with the spectral radiometer and chemical ionisation mass spectrometer. We thank Sam Cox for his help with the AtChem modelling framework.

Funding Information:
This research has been supported by the Natural Environment Research Council (grant nos. NE/K012029/1, NE/K012169/1 and NE/K004069/1).

Publisher Copyright:
© Author(s) 2023.

Cite this