Convective transport of gases and aerosols
The transport of aerosols and gases to the upper troposphere through deep convection
People: Prof Ken Carslaw, Dr Yan Yin and Dr Doug Parker
Funded by: 3 year NERC grant as part of the UTLS Thematic Programme.
Abstract
There is a lack of basic knowledge on the factors controlling trace gas concentrations and aerosol properties in the upper troposphere (UT) and lower stratosphere (LS). Deep convective clouds, including cumulonimbus and frontal systems, have an important influence on the trace gas and aerosol budget of the UT/LS region by:
- Effectively transporting air masses containing trace gases and aerosols from anthropogenic and natural sources from the boundary layer into the UT in relatively short time scales;
- Altering the population and physical and chemical characteristics of aerosols through chemical and microphysical processes occurring within the transporting clouds and in the cirrus anvils that form in the outflow regions;
- Acting as a source of new particles be enabling gas-to-particle conversion of transported precursor gases, which can occur in the clean regions of the UT.
In this NERC UTLS project we were using a hierarchy of cloud microphysical/dynamical models coupled with descriptions of cloud microphysics and chemistry of varying complexity. The models will allow fluxes of important ozone and aerosol precursor gases into the UT to be estimated, including species such as NOx, a wide range of organics, DMS and SO2. The composition and size distribution of cloud-processed aerosols injected into the upper troposphere will also be computed. The trace gas concentration fields computed at cloud top were used as an initialisation for a chemical box model to investigate the effect of cloud vented air on UT chemistry.
This GIF animation shows the transport of a gas through a cumulus cloud. The gas in the top panel is insoluble, so is transported just like a 'tracer'. In the bottom panel, the gas is assumed to have an effective Henry's law constant of 107 mol dm-3 atm-1. In this case, much of the gas is lost due to partitioning into the cloud water and subsequent rain-out. The blue region indicates the presence of liquid water in the cloud. Notice that the distance axis at the bottom moves with each time step. This is because we have assumed the cloud to form in a sheared flow.
In Waibel et al. (Science, 1999), we used a computer model to simulate denitrification in the Arctic stratosphere for the winter 1994/95. The results showed that the temperature of the 1994/95 Arctic vortex was just cold enough to allow significant polar stratospheric cloud formation and subsequent denitrification. We showed that our calculated denitrification agreed well with the limited observations that were then available. In addition, chemical model simulations indicated that a significant fraction of the observed ozone depletion in that winter could be attributed to denitrification.
Simulating denitrification is not easy. First of all, we don't understand polar stratospheric cloud formation well enough to know which particles are causing it. We therefore don't understand what conditions in the Arctic stratosphere will predispose it to denitrification.
Particle observations from the NASA ER-2 aircraft during the SOLVE campaign have led to a step-change in our understanding of denitrification. Particles as large as 20 micrometres (christened 'rocks') were observed over large regions of the Arctic stratosphere. These particles are much larger than previously thought possible, and require us to completely rethink the way we simulate them in models. The first report of these particles and our preliminary modelling calculations can be viewed in this downloadable publication in Science. In addition, this NASA press release describes the background to the discovery of the rocks. Based on results from the SOLVE campaign, we are developing more sophisticated models of denitrification. These will account for the growth kinetics of the cloud particles and the formation and evaporation cycles of particles as they circle the pole. Our eventual aim is to understand the processes well enough to be able to make reliable predictions of denitrification for future Arctic winters, and hence to more accurately predict ozone losses. Development and application of the model is described in some recent publications [Carslaw et al., Mann et al.]
As part of the EU project MAPSCORE (Mapping of Polar Stratospheric Clouds and Ozone Levels Relevant to Europe) we will compare model calculations of denitrification in recent winters with observations. The model denitrification fields will then be used in chemical transport models to investigate the effect on ozone depletion.
In EUPLEX (European Polar Stratospheric Cloud and Leewave Experiment) we will evaluate our model against new observations from the M55 Geophysica aircraft in the Arctic winter 2002/3. Members of the group will also participate in the field campaign in Kiruna, Sweden, to predict the occurrence of large NAT particles for flight-planning purposes.