PHOENICS aims to assess the direct climate effect of aerosols, to quantify and reduce several of the associated main uncertainties using climate models, and to evaluate the impact of European emissions, via aerosol formation, on Europe and the global environment and climate, and the influence of other world-regions on Europe. This information is a prerequisite for the definition of EU directives and international negotiations for protection of the environment and of human health, measures with obvious financial and political impacts.
The optical, chemical and hygroscopic properties of the multi-component mixed aerosol are considered using a size-dependent aerosol dynamic model embedded within a global 3-dimensional atmospheric general circulation model (A-GCM). A thorough validation of the developed parameterisations, the A-GCM results, as well as an evaluation of the uncertainties associated with the calculations of the direct aerosol effect will be based on selected observations and optimal use of satellite data.
To give to the PHOENICS model results and data compilation an international visibility and to enhance the policy-relevance of the PHOENICS results, the PHOENICS project partners have coordinated the Aerosol Model Intercomparison (AEROCOM) initiative. This initiative includes 16 global aerosol models 8 from Europe, and 8 others from the United States and Japan.
Model quality changes from parameter to parameter and from region to region and of course from model to model. The monthly values appear to be robust enough to demonstrate the achieved quality of global models. The different models agree on major features such as the importance of the contrasting spatial distribution of the aerosol which is responsible for regionally important radiative perturbations. Models agree that the water mass associated with the aerosol is globally as important as the dry aerosol mass. Both the humidity fields in the model and the water uptake formulation are important to calculate optical properties. The variation of aerosol extinction is affected by that of the vertical profiles of concentration as much as by the gross optical property mass extinction coefficient.
The distinct effect of absorbing aerosols is enhanced by the fact that the predominant anthropogenic absorbing aerosol component, black carbon, is partly internally mixed with otherwise predominantly scattering aerosol components, such as sulphates and particulate organic matter. Consideration of internally-mixed aerosol results in higher absorption cross-sections than for the simplified externally mixed approach.
For the first time, the aerosol radiative impact has been estimated systematically for each component not only in the shortwave but also in the longwave spectrum. The longwave component is less important than the shortwave one but represents a systematic greenhouse effect of a few tenths of Wm-2 which has to be considered as our quantification of the aerosol radiative impacts is being refined.
Anthropogenic sulphate from fossil fuel burning is estimated to provide almost half of the total anthropogenic aerosol optical depth. The second important contribution is due to biomass burning organic aerosols. The heating of the atmospheric column (difference between surface (SFC) and top of the atmosphere (TOA) forcing: ~0.5 Wm-2 globally) has been shown now to be important in terms of regionally changing the vertical stability of the air column.
The global coverage from satellites is certainly the most valuable check on the realism of the model simulations. Measurement based estimates of the radiative forcing in clear sky areas have been assembled for the first time and challenge the modelling. Recently satellite observations have been refined to provide also size related information that is valuable to differentiate between natural (coarse) and anthropogenic (mostly fine) aerosols. The state-of-the-art MODIS aerosol retrievals combined with TOMS and SSM/I measurements have been used within PHOENICS to estimate the clear-sky direct radiative forcing for the year 2002. For the first time, aerosol direct radiative forcing can be estimated by aerosol type both over ocean and land.
Dedicated model simulations from an ensemble of AEROCOM aerosol models for present and pre-industrial conditions reveal that 25% of the aerosol optical depth is anthropogenic. The resulting direct all-sky (including clouds) anthropogenic aerosol radiative forcing estimate by the PHOENICS models is -0.12 to -0.29 Wm-2 and the all-sky direct radiative perturbation from aerosols (natural and anthropogenic) is estimated at -1.2 to -1.8 Wm-2. Both forcing and radiative perturbation from aerosols for clear sky conditions amount to larger effects than in all-sky (clear and cloudy) conditions. Model estimates show a smaller magnitude than the satellite estimates from observed aerosol optical properties and observed surface albedo. Observation based shortwave radiative forcing include both the natural and anthropogenic signal and in clear-sky conditions amount to -5.1 up to -6.6 Wm-2. Model simulations, in contrast, provide shortwave radiative forcing values of only -3 to -3.3 Wm-2. There is ongoing research is ongoing to resolve this discrepancy.
The response of the global aerosol system to changes in anthropogenic emissions is demonstrated to be non-linear due to the increase in the transfer of insoluble to soluble particles with increasing anthropogenic emissions. The demonstrated non-linearities, the coherence between different aerosol cycles and the deviations from additivity challenge the concept of radiative forcing applied in IPCC (2001) for assessment of aerosol radiative effects. Our study emphasizes the need for integrated emission strategies for aerosols and their precursors comprising the cross-connections of the global aerosol effects.
The largest range of radiative perturbation estimates by the various AEROCOM models exists for black carbon optical depth and both loads and optical depth for the natural aerosols sea salt and dust. The uncertainty is higher for natural sea-salt and dust than for the anthropogenic species, mainly due to the size distribution assumptions of the emissions. This is evidenced by models having different residence times for a given aerosol component for the same total emissions.
A marine primary of natural organic aerosol (or particulate organic matter - POM) has been identified within PHOENICS. Given that the evolution of phytoplankton is driven by environmental change, particularly increasing oceanic temperatures, the production of marine POM represents a newly-identified and potentially important component of the marine-biota and climate-feedback system involving aerosols and clouds. Another important fraction of organic aerosol studied within PHOENICS is the secondary organic aerosol (SOA). The global annual SOA production from biogenic VOC is estimated from 2.5 to 44.5 Tg of organic matter per year with large uncertainties (of about a factor of 20) associated with this estimate. There is consensus that the major precursors of SOA are biogenic volatile organic compounds (VOC). The natural variability of SOA production is evaluated to about 10% and is as high as the SOA formation from anthropogenic precursors. Significant positive and negative feedback mechanisms in the atmosphere are responsible for the non linear relationship between emissions of biogenic VOC and SOA burden. Roughly 1/3 of the total chemical production of SOA is calculated to occur in the upper troposphere/lower stratosphere thus affecting the SOA vertical distribution and inducing different climate impact than surface emissions.
According to the three different PHOENICS model simulations 40-50% of the European Aerosol Optical Depth (AOD) is caused by emissions of anthropogenic aerosol and aerosol precursors over Europe (in 2000). With regard to the individual aerosol components black carbon (BC) is the one affected most by anthropogenic emissions over Europe (75-85% of AOD from BC) whereas only 23% of the AOD from SOA is attributed to these emissions mainly by enhancement of the large natural component of SOA.
The contribution of anthropogenic European emissions to inorganic AOD in the rest of the world (RoW) is larger than for BC and POM, because sulphate is a secondary aerosol that is produced from primary sulphur dioxide (SO2) emissions. The results of the different models are generally quite consistent with some differences caused by the treatment of water-uptake and aerosol dynamics. From pre-industrial condition simulation it is estimated that for the 2000 AOD values over Europe about half is caused by anthropogenic emissions. For the AOD from inorganic, BC, and POM, these fractions amount to 61%, 79%, and 44% respectively.
Dust and sea-salt are mainly of natural origin, while BC, POM, SO4 and NO3 are mainly anthropogenic. The net import in Europe of sea-salt aerosol (from the west) and dust aerosol (from the south) is calculated to be 11.8 Tg yr-1 and 20.4Tg yr-1, respectively. The export of anthropogenic aerosols from Europe is calculated as 0.23 Tg yr-1 BC, 0.53 Tg yr-1 POM, 1.37 TgS yr-1 sulphate, and 0.11 TgN yr-1 nitrate aerosol. The most noticeable aspect is the large contribution of Europe in the global aerosol nitrate budget. Coarser model resolution clearly results in less efficient production of aerosols in particular for NH4NO3, for which the formation is highly non-linear and thus very sensitive to the model resolution. The budgets of other aerosol components present similar behaviour but are less sensitive to model resolution. Wet removal parameterization and emission scenarios are also critical for aerosol budgets.
Current legislation over Europe will probably result in lower emissions of aerosols and aerosol precursors. The relative roles of nitrate and sulphate aerosols may therefore change in the future. For the three major components of inorganic aerosol, significant reductions over most polluted areas are calculated to be expected for the year 2020 based on Eurodelta emission scenarios. For aerosol nitrate, however, a slight increase is calculated at some locations due to the competition of nitrate with SO4 for neutralization of NH3.
The climate of the 20th and 21st century has been simulated with a coupled atmosphere-ocean model including a detailed representation of tropospheric aerosols and their climatic effects. The aerosol model developed in PHOENICS predicts the evolution of an ensemble of interacting internally and externally mixed aerosol populations of different chemical composition. Novel features are also the interactions between the atmospheric dust and sulphur cycles with the marine biosphere. This allows us to simulate not only the climatic response to enhanced levels of greenhouse gases and anthropogenic aerosols of different chemical composition but also the feedbacks arising from the interaction between climate parameters and natural aerosol components. Externally prescribed are anthropogenic emissions of aerosols and aerosol precursor gases, atmospheric greenhouse gas concentrations, and variations in solar irradiance and volcanic aerosol optical depth.
PHOENICS has been successfully met the objectives set at the beginning of the project. PHOENICS developed a host of parameterisations of aerosol formation, and removal processes, which were evaluated by comparison of model results with observations collected and harmonized within the framework of PHOENICS. The current direct aerosol effects have been derived both by A-GCM modelling coupled with the aerosol dynamic module and by satellite observations.
The aerosol direct effect in the longwave has been calculated for the first time and was found to be minor compared with shortwave aerosol forcing. Anthropogenic aerosols increase the outgoing shortwave flux by 2.7 Wm-2 at the top of the atmosphere, in clear-sky on a global, annual average.
The total direct radiative forcing deduced from observations is about -1 Wm-2. This can be compared to the IPCC TAR estimate aerosol forcing of -0.5 Wm-2, with a large uncertainty range of -0.2 to -1.3 Wm-2 (excluding mineral dust).
Dedicated model simulations from an ensemble of AEROCOM aerosol models for present and pre-industrial conditions reveal that 25% of the aerosol optical depth is anthropogenic. The resulting direct anthropogenic aerosol forcing estimate from PHOENICS models is -0.12 to -0.29 Wm-2 and the direct forcing of aerosols (natural and anthropogenic) is estimated at -1.2 to -1.8 Wm-2. Thus, model estimates show a smaller magnitude and the discrepancy between model and observationally-based estimates has not been resolved yet.
PHOENICS models estimate modest direct anthropogenic aerosol radiative forcing of -0.15 to -0.3 Wm-2 at the top of the atmosphere in all-sky on a global, annual average.
The contribution of European anthropogenic emissions to AOD over Europe has been estimated by 3 different PHOENICS models to be 40-50% of the total AOD. Implementation of currently planned emission reductions of aerosol and aerosol precursor gasses indicate a larger future role of aerosol nitrate in Europe.