Climate Chemistry Subproject
A PDF file containing a description of the Tropospheric Chemistry schemes available can be found here: Tropospheric Configuration of UKCA (Fiona M. O’Connor).
Introduction
The Atmospheric Chemistry Modelling Support Unit (ACMSU) at the University of Cambridge, together with the University of Leeds are collaborating with the UK Met. Office on the development of a new community chemistry-climate model. We describe the motivation behind this model here. Dr Olaf Morgenstern is now at NIWA. Background
The role of atmospheric chemistry in the climate system is now recognised as being of central importance. Modelling of the chemistry/climate system needs to be advanced in recognition of this important coupling. In the modelling for the 3rd assessment report of the IPCC, tropospheric chemistry modelling using HadCM3 was performed off-line from other components of the climate model. This is an important limitation; chemistry/climate coupling now needs to be included 'on-line' to maintain a state-of-the-art modelling capability. It is also now recognised that certain aspects of the aerosol behaviour in climate models need to be improved in order properly to capture the effect of aerosols on radiative forcing, particularly the forcing related tochanges in cloud microphysics. Similarly, links with the vegetation, deposition, etc., should also be on-line. To remain at the leading edge of worldclass science it is essential that the UK community continues to develop our modelling capability, towards the objective of a fully coupled climate model; this development is the aim of this project, drawing on the strengths of the Met Office and NCAS communities to build a model for a variety of uses including contributions to future IPCC assessments, advice for government and for detailed process studies.
The MO has relied on a series of different models to carry out different aspects of atmospheric chemistry modelling: the Lagrangian model STOCHEM has been used for tropospheric chemistry modelling, with EUMETRAC being used for interactive stratospheric chemistry modelling and with aerosol modelling being done within the climate model using the default transport scheme. In NCAS, models in use include a range of tropospheric and stratospheric schemes implemented in HadAM3 using a revised transport scheme with the ASAD chemistry module, as well as chemical transport models for the troposphere and stratosphere. Sophisticated aerosol box models and offline global aerosol transport models are also being developed with a view to understanding the evolution of aerosol physical and chemical properties. There is clearly a case for combining the best aspects of these models in a framework to create a flexible model environment based on the UM with the added advantage of being accessible to more users in the research community than the current generation of models.
The Hadley Centre model presently contains options for interactively simulating several varieties of aerosol: sulphate, sea-salt, fossil fuel black carbon, biomass smoke and mineral dust. The direct and (where relevant) indirect effects of these aerosols can be included if desired. However, there are some important scientific issues that the current aerosol code addresses incompletely or not at all. For example, there is only limited freedom for the size distribution to evolve (apart from the case of mineral dust where 6 size bins are used). Also there is no treatment of internally mixed particles (i.e. particles composed of more than one constituent). New aerosol observations and detailed model simulations highlight the importance of aerosol types that are not presently included in the current model; the most important being nitrate and secondary organic aerosol, the latter formed from biogenic and anthropogenic source gases. Some of the limitations of the current model reflect the Hadley Centre's requirement for a scheme affordable in centennial timescale experiments with the computer power hitherto available.
Looking forward, likely increases in computer power and the need to improve the representation of important processes both indicate the need to develop a more sophisticated aerosol code for the Unified Model. It is important now to design a new aerosol scheme that will be as self-contained and "modular" as possible.
The model to be developed will have a range of uses from longterm IPCC assessments through to detailed process studies. It will become the model of choice for the UK atmospheric science community. The scientific argument for the project is that this new model will allow a range of interactions which are known to be essential and which have previously not been fully included. There are also strategic arguments. First, the human resources for the development of atmospheric chemistry and aerosol models of high complexity are currently small in relation to those available in atmosphere modelling. It is important to harness the world class expertise available in the MO-NCAS communities. Second, an increasing amount of data (produced with important scientific and financial contributions from the UK) is being collected from aircraft, balloon ascents and, more recently, from satellites, which needs to be used in modelling in order to improve the science. This new model will maximize UK use of these data.
The model to be developed must be flexible. It must be capable of use at high resolution with short timescales, as well for runs for much longer time periods. While its main use will be in a fully coupled mode, for a range of timescales and with varying complexity (the objective here), the model will also be used outside this specific project for a further set of scientific issues. The model developments here will be an important step towards a fully coupled Earth System model, incorporating all relevant aspects of the earth system. Finally, it must be capable of easy use by non-specialists using UMUI.
Scientific Driver
The model will be a state-of-the art chemistry/climate model, which will perform centennial timescale integrations for IPCC assessments. A primary aim will therefore be to study future states of the atmosphere under different scenarios for the emissions of major greenhouse gases (CO2, methane, CFCc, HCFCs, HFCs, N2O, etc.) and projected changes in aerosol composition. The interaction with changing levels of oxidants and aerosol will be central to these integrations.
Studies of the oxidation capacity of the troposphere, including the two-way interaction with atmospheric aerosol, will have high priority.
A range of chemistry-climate interactions will be considered, including the role of temperature and water vapour changes on atmospheric composition, the impact of changing convection and lightning NOx emissions under changed climates, and the importance of natural variability including various known modes of atmospheric variability.
The model will be suitable for use to explore chemistry-ecosystem interactions, including natural emissions of VOCs and DMS, and deposition of ozone and nitrogen species in gaseous and particulate forms. An important new area for investigation will be biogeochemical cycles involving particulate deposition to the oceans.
Air quality is a major concern during the coming century, with large increases likely in surface ozone and particulates at some locations. Study of the interaction between air quality and global change will be another high priority research area.
The recovery of stratospheric ozone is known to depend on climate change (and this problem intimately connects the Kyoto and Montreal Protocols). We will investigate the interactions between stratospheric dynamics (including the Brewer Dobson circulation, tropopause structure, stratosphere to troposphere exchange, etc.) and chemistry, and climate change. The whole atmosphere chemistry scheme to be implemented will make the technical distinction between troposphere and stratosphere superfluous and will be ideal for, for example, stratosphere/troposphere exchange studies and assessment studies of the UTLS. Aerosols-chemistry-radiation interaction is a developing field for which the model will be ideally suited.
