4-Dimensional Variational Assimilation Of Satellite Data
into a Chemistry Transport Model

A contribution to ACCENT-TROPOSAT-2, Task Group 2

Hendrik Elbern

Rhenish Institute for Environmental Research at the University of Cologne (RIU)
Aachener Strasse 201-209, D-50931 Köln
Tel.:  +49 221 400 2220
Fax:  +49 221 400 2320

An increasing wealth of trace gas retrievals of the troposphere will be available for assimilation into complex chemistry transport models. Namely sensors like SCIAMACHY, GOME(-2), AATSR, and MOPITT have demonstrated that they a valuable data sources for assimilation. Species to be assimilated include O3, NO2, SO2, formaldehyde and potentially various classes of aerosols.

Space time data assimilation algorithms aim to upgrade level 2 satellite retrievals, scattered in space and time to level 3 products, giving a comprehensive and continuous picture of the troposphere, commonly presented synoptically on a regular grid at the user´s disposal. In the proposed work the European scale complex University of Cologne EURAD (EURopean Air pollution Dispersion) model with its adjoint components will be taken to form a three- and four-dimensional variational data assimilation ( 3- and 4-D-var) system by ingesting available and envisaged tropospheric satellite data. The 4D-var EURAD assimilation system is presently the only one worldwide to also run in a multiple nesting mode, with grid resolution varying from 125 to 6 km. Further it has shown the potential not only to estimate the chemical state of the troposphere, but also provide assessments of emission strengths.

The 4-D-var method is one of the very few techniques to attain a real synergistic use of observation sets, combined from heterogeneous instrumentation. This work will build on present experiences with tropospheric NO2 columns and artificial neuronal network based ozone profile retrievals.

The scheduled work includes combinations of satellite data with in situ observations, identifying the important boundary layer portion of the tropospheric columns. While it could be demonstrated that surface data assimilation is beneficial for the predictive skill, satellite data assimilation will be evaluated for an added performance increment of air quality forecasts. A further issue, extremely important for data assimilation algorithms, is the estimation of radii of influence associated with retrieved data, to achieve a better formulation of the background error covariance matrix.

The mature data assimilation system is also expected to fit into routine GMES service type tasks

Time schedule

 

2004

2005

2006

2007

2008

2008

Activity 5

**

**

**

**

**

**

 Approximate manpower and cost

 

2004

2005

2006

2007

2008

2008

Personnel / man-years

0.5

1.

1

1.

1

1

Yearly cost (kECU)

30

60

60

60

60

60

Likely funding agencies

CEC; ESA, BMBF, German Science foundation

Co-worker

Achim Strunk
Rhenish Institute for Environmental Research at the University of Cologne (RIU)
Aachener Strasse 201-209
50931 Köln
Tel.:  +49 221 400 2220
Fax:  +49 221 400 2320

 

 


 Return to the top of the page