The arctic winter stratosphere: simulated with a 3-D chemistry transport model

Abstract

During the past two decades, the ozone layer has developed a “hole” each winter and spring above the Antarctic continent. Also in cold Arctic winters substantial stratospheric ozone depletion has been measured, although less than in the Antarctic stratosphere. In the Arctic winter stratosphere, the amount of ozone depletion varies interannually and within one winter, depending on polar vortex stability and temperature. The simulation of transport and chemical conversion of ozone and related species requires a three-dimensional (3D) chemistry transport model (CTM), because of the non-symmetric behaviour of the Arctic polar vortex. This thesis reports on several studies of the Arctic winter stratosphere carried out with such a CTM, using off-line meteorological fields. In Chapter II, chlorine activation and ozone depletion in the Arctic winter stratosphere of 1996-1997 are modelled with the newly developed stratospheric version of our CTM. Comparisons have been made with total O3 columns and ClO concentrations observed by satellites, and with ozone loss rates derived from observations during February and March 1997. ClO concentrations and ozone depletion are somewhat underestimated by the model. Key model parameters have been varied to explain this underestimation. Next to temperature, the formation mechanism of solid and/or liquid PSC particles constitutes the main model uncertainty. The representation of tracer transport is a third uncertain parameter, influencing both ozone and inorganic chlorine. In Chapter III, we have used the CTM with different horizontal resolutions to evaluate this stratospheric transport by simulating the long-lived tracers HF and CH4 during the Arctic winter of 1999/2000. Outside the vortex the model results agree well with the observations, but inside, the model underestimates the observed vertical gradient in HF and CH4. Too strong mixing through the vortex edge could be a cause for these model discrepancies, e.g. associated with the calculated mass fluxes. As later studies point out, the use of a reduced grid around the poles also plays an important role. Another conclusion of this chapter is that a global 6ox9o resolution is too coarse to represent the polar vortex, whereas the 3ox2o and 1ox1o resolutions yield similar results, also when a 6ox9o resolution is applied in the tropical region. Chapter II showed that the treatment of PSC particles in global models can have a large effect on calculated ozone depletion as well. In Chapter IV a new algorithm is implemented in the CTM, which describes PSCs more realistically by mimicking the growth and sedimentation of nitric acid trihydrate (NAT) particles. Simulations were performed for three separate 10-day periods during the 1999-2000 Arctic winter. By implementing this new routine the calculated denitrification increased, which would lead to more modelled ozone depletion. Overall, this thesis has led to an improved model representation of the processes leading to ozone depletion in the Arctic stratosphere. Integrating these results in a full chemistry transport model and further analysis of the past cold winters in the Arctic winter stratosphere will help increase our understanding and predictive capability of Arctic stratospheric ozone.