Atmospheric Transport Modelling and inverse modelling of trace substance sources


Lagrangian modelling of atmospheric transport and dispersion

The transport of atmospheric trace constituents, but also of air  as such is relevant for many questions and applications, for example

  • air pollution
  • greenhouse gases
  • volcanic eruptions
  • nuclear accidents
  • natural radioactivity
  • synoptic climatology.

The transport process can be divided into the transport by the mean wind and transport by turbulent motions. The latter lead to dispersion. Most substances are transported along with the air, only large particles undergo gravitational settling. All kinds of particles and chemically reactive gases undergo dry and wet removal from the atmosphere with deposition on the ground. Depending on the nature of the substance and the time scale considered, chemical reactions may be relevant.
Complex transport patterns can be modelled with Eulerian models (using a fixed grid) and Lagrangian models (tracking trajectories of computational particles and process acting upon them). 
The group contributes to the development of the Lagrangian particle dispersion model FLEXPART and applies it. FLEXPART is an open-source project started initially in Vienna. It is a highly flexible and efficient dispersion code for regional to global scales.i Current development issues include the parameterisation of wet deposition and numerical issues (interpolation and grids). Some work is also carried out on the base of the trajectory model FLEXTRA.

 

Researchers:

  • Petra Seibert
  • Anne Philipp

Further Reading:

Publications:

Links:


Inverse modelling of atmospheric trace substances

In inverse modelling of atmospheric trace substances, we try to find sources of atmospheric contaminants or constitutents from ambient measurements. Measurements could be, for example, concentrations at a station, or vertical column totals observed from a satellite. This is an optimisation problem: which emissions would, plugged into a transport and dispersion model, optimally reproduce the observations?
We have developed an inversion method for substances which are not chemically reactive (or at least whose reaction rates can be prescribed) that uses source-receptor relationships and (iterated) analytical minimisation of a cost function. In the past years, this development has been in cooperation with Andreas Stohl and his group at NILU. Applications include

  • volcanic ash
  • greenhouse gases
  • reactor accidents
  • monitoring of the Comprehensive Nuclear Test Ban Treaty

 

Researchers:

  • Petra Seibert
  • Radek Hofman

Publications:

Links:

Fig. 1: Inverse modelling of SO2 emitted by the explosive eruption (VEI=4) of Kasatochi volcano (Aleutians) in August 2008. The a posteriori emission profile obtained from the inversion shows a maximum near the tropopause with some overshooting, and a secondary maximum at about 7 km height. (© IMGW)