Abstract. More than 3 decades after the discovery of the ozone hole, the
processes involved in its formation are believed to be understood in
great detail. Current state-of-the-art models can reproduce
the observed chemical composition in the springtime polar stratosphere,
especially regarding the quantification of halogen-catalysed ozone
loss.
However, we report here on a discrepancy between simulations and
observations during the less-well-studied period of the onset of
chlorine activation. During this period, which in the Antarctic is
between May and July, model simulations significantly overestimate
HCl, one of the key chemical species, inside the polar vortex during
polar night. This HCl discrepancy is also observed in the Arctic.
The discrepancy exists in different models to varying extents; here,
we discuss three independent ones, the Chemical Lagrangian
Model of the Stratosphere (CLaMS) as well as the Eulerian
models SD-WACCM (the specified dynamics version of the Whole Atmosphere
Community Climate Model) and TOMCAT/SLIMCAT. The HCl discrepancy points to
some unknown process in the formulation of stratospheric chemistry
that is currently not represented in the models. We characterise the HCl discrepancy in space and time for the
Lagrangian chemistry–transport model CLaMS, in which HCl in the
polar vortex core stays about constant from June to August in the
Antarctic, while the observations indicate a continuous HCl decrease
over this period. The somewhat smaller discrepancies in the Eulerian
models SD-WACCM and TOMCAT/SLIMCAT are also presented. Numerical
diffusion in the transport scheme of the Eulerian models is
identified to be a likely cause for the inter-model differences.
Although the missing process has not yet been identified, we
investigate different hypotheses on the basis of the characteristics
of the discrepancy.
An underestimated HCl uptake into the polar stratospheric cloud (PSC) particles that consist
mainly of H2O and HNO3 cannot explain it due to
the temperature correlation of the discrepancy. Also, a direct
photolysis of particulate HNO3 does not resolve the discrepancy
since it would also cause changes in chlorine chemistry in late
winter which are not observed.
The ionisation caused by galactic cosmic rays provides an additional
NOx and HOx source that can explain only about 20 %
of the discrepancy.
However, the model simulations show that a hypothetical
decomposition of particulate HNO3 by some other process not
dependent on the solar elevation, e.g. involving galactic cosmic
rays, may be a possible mechanism to resolve the HCl discrepancy.
Since the discrepancy reported here occurs during the beginning of
the chlorine activation period, where the ozone loss rates are small,
there is only a minor impact of about 2 % on the overall ozone
column loss over the course of Antarctic winter and spring.