Measuring Ozone Deposition to the Ocean Surface and Assessing its Biogeochemical Controls

Dry deposition of ozone to the sea surface represents a significant portion of global tropospheric
ozone loss. It introduces considerable uncertainty in global models due to limited understanding of the
reactivity of iodide and organic material in the sea surface towards ozone. This is particularly true of
organic material due to its variable composition.
This thesis details ozone flux and associated measurements at and around the Penlee Point
Atmospheric Observatory (PPAO) on the UK south coast from 2018 until 2021 where coastal ozone
flux was calculated via eddy covariance. Monthly median deposition velocity was 0.007 – 0.033 cm
s−1 across all fieldwork, similar to the values reported from ship-based measurements: 0.009 – 0.034
cm s−1.
Iodide and dissolved organic carbon (DOC) concentrations in the water within the flux footprint were
~50 to ~100 nmol dm−3 and 1.3 – 2.2 mg dm−3. While iodide increased to a peak in July (coinciding
with phytoplankton blooms), DOC conversely peaked in November. These measurements were used
with a 1-layer and a 2-layer model to compare deposition observations to predictions. The 1-layer
model in the absence of DOC reactivity typically gave values closest to observations and showed a
similarly strong variation with friction velocity. Inclusion of the DOC-ozone reaction with a rate
constant of 3.7 × 10−6 dm3 mol−1 s−1 caused both models to overestimate, but also mimic some
variation between months suggesting its contribution was overestimated, but still important.
Liquid chromatography-mass spectrometry was used to identify compounds in the water near the
PPAO. Double bond equivalence decreased following exposure to 500 ppbv ozone, while
dicarboxylic acid concentrations increased, potentially due to unsaturated fatty acid ozonolysis.
Several dicarboxylic acid concentrations in PPAO samples fell from November – April, similarly to
DOC concentrations. The potential for some introduction of dicarboxylic acids as contaminants from
the sampling method remains a possibility.