Search

Atmospheric models are used both to assist in the retrieval of information about tropospheric composition and to interpret the results. The models themselves encompass much of our knowledge and understanding about trace components in the atmosphere and so they themselves must be tested against data. Care must be taken not to mix the dual role so that one can be sure of the conclusions drawn, and also be alert to any new phenomena that may be revealed by the data. These issues are dealt with and illustrated here. Chapter 9 goes on to discuss inverse modelling which is used, for example, to deduce emissions from regional and global data sets. Another important topic is data assimilation in which data and models are combined to derive improved concentration fields and, for example, to make forecasts. A summary provides some perspectives for future work. There is also a useful appendix on the principles of inverse modelling. [ABSTRACT FROM AUTHOR]

The EURAD-IM chemistry transport model and its 4d-var inverse model extension is applied to one summer and one winter episode, in order to identify the benefit of tropospheric NO2 column retrievals for estimating near-surface nitrogen dioxide concentrations. Initial values and emission rates are jointly optimised by assimilating tropospheric NO2 data from the OMI, while European ground based observations are used for impact evaluation. Results show a moderate improvement of surface level nitrogen dioxide estimates during the summer episode, successfully sustained after the assimilation period through emission adjustments. In the winter case, the OMI data is of limited value due to lower boundary layer heights and thus smaller impact on emission rates. [ABSTRACT FROM AUTHOR]

The chemistry transport model system EURAD-IM and its variational data assimilation implementation are applied to air quality assessment problems. The facility of joint initial value and emission rate optimisation together with a nested 4d-var module is employed to the measurement campaign BERLIOZ. To emphasise the benefits of computationally efficient data assimilation for monitoring and simulating air quality, results of an operational 3d-var implementation are also given. [ABSTRACT FROM AUTHOR]

Combined emission rate and chemical state optimization by fourdimensional variational data assimilation is presented. Using an adjoint nesting technique, a model setup focussing on the Berlin area is employed, enabling to analyze constituents with strong spatial gradients like NO and NO2. Possible requirements for satellite observations to be used for tropospheric chemical state estimation are shortly discussed. [ABSTRACT FROM AUTHOR]