Search

This paper aims at presenting a model-to-data comparison of the Aerosol Optical Thickness (AOT) and of a few sparse data for Single Scattering Albedo (SSA) over Europe for one year. The optical parameters are computed from a size-resolved aerosol model embedded in the POLYPHEMUS system described in Mallet et al. (2007). The methodology is first described, showing that several hypothesis can be made for several microphysical aerosol properties. The simulation is made over one year (2001); statistics and monthly time series for the simulation and AERONET data are used to evaluate the ability of the model to reproduce AOT and vertically averaged SSA fields and their variability. The relation with the uncertainties of measurements is discussed. Then a sensitivity study with respect to the mixing state of the particle, the way to compute the Aerosol Complex Refractive Index (ACRI) of a mixture and the way to take into account water uptake is carried out. The results indicate that the computation of AOT is relatively stable, while the computation of the single scattering albedo is much more uncertain.

Rapport quadriennal 1999-2003; Human activities are at the origin of emissions of polluted gases that can perturb the chemical composition of the atmosphere and the climate. The measurements of the concentration of atmospheric constituents at global scale is needed to forecast long term consequences of these emissions. Several space missions embarking instruments dedicated to the observation of minor atmospheric species have been launched recently or will be launched in the near future. The measurements of chemical species are obtained at the global scale with vertical, horizontal and time sampling depending upon the instrument. Data assimilation techniques allow to reconstruct in an optimal way 3D global field of atmospheric parameters starting from a limited data set. They are widely used in weather forecast models and starts to develop in the field of atmospheric chemistry. This paper presents the work made within the French scientific community to develop and implement data assimilation methods in chemistry, dynamics and transport models in order to improve our knowledge of the composition of the troposphere and the stratosphere, using available measurements of atmospheric species (from satellites, balloons and aircrafts). The work presented here has been done in the frame of French National Program of Atmospheric Chemistry (PNCA) in a coordinated project " Satellite data assimilation and sources inversion for the study of the atmospheric chemistry ". It is made in interaction with other PNCA projects, " Chemistry at large scale and chemistry-climate interaction " and " Emissions of chemical compounds and aerosols ". The first part describes studies on the chemical composition of the troposphere. The major scientific issues concern the evolution of the oxydation potential of the troposphere, in particular: ■ the impact of changes in emission of trace gases (CO, CH 4 , hydrocarbons, …); ■ the influence of highly industrialized regions on the distribution of chemical species; ■ the effect of biomass burning in the tropics. MOPITT (Measurement Of Pollution In The Troposphere) satellite data have been assimilated in a tropospheric Chemistry Transport Model (CTM) to obtain global fields of CO. MOPITT data will be also used to invert sources of CO and CH 4 by data assimilation techniques. The second part presents development made to assimilate stratospheric ozone and minor species data in several CTM, using both variational and sequential assimilation techniques. The third part concerns methodologic studies for the use of variational methods in data assimilation in CTM.

The transport and conversion of chemically active compounds in the atmosphere is mathematically described by ∂cs∂t+∇⋅(ucs)=V(cs)+Rs(c) (1.1) where cs are the concentrations of chemical species in the atmosphere. The first term denotes the rate of change of c s and the second term describes the transport of species by a given wind field u. The first term on the right hand side represents the sub-grid scale vertical transport that is not resolved on the grid. Finally, the last term describes the chemical reactions among the species. Usually, Eq. (1.1) is discretized in space with finite volume techniques (e.g. Zlatev 1995). The discretized equations are normally integrated forward in time using operator splitting. In this way, small time steps due to the stiff chemistry and vertical mixing operators can be avoided by applying tailored implicit schemes (Verwer et al. 1998).

We present in this article a stochastic algorithm based mainly on [3] and [2] applied to the integration of the General Dynamics Equation (GDE) for aerosols. This algorithm is validated by comparison with an analytical solution of a coagulation-condensation-evaporation model.