Source term, dispersion modelling and data assimilation

Dispersion modelling

Several atmospheric dispersion models (FLEXPART, SILAM and WRF-Chem) are used in VAST to simulate the dispersion of volcanic ash and SO2 emitted during an eruption. The below movie loop shows an example of the transport of ash and SO2 emitted during the 2011-Grimsvötn eruption.

Figure 1: Modelled dispersion of ash and SO2 from the Grimsvötn eruption in 2011.

The accuracy of the model forecast is highly dependent on the input parameters to the model simulations, such as the accuracy of the meteorological data (wind, temperature, precipitation) and the determination of the emissions from the volcano (the source term).

Source term determination

Using well defined source terms provide more accurate forecasts of the spread and concentration levels of the volcanic emission clouds. The main VAST technique for source term determination is an inversion scheme based on the work of (Seibert, 2000). The inversion method couples atmospheric transport models such as FLEXPART with satellite observations.

The models are firstly used to evaluate how releases of ash or SO2 at different vertical levels above the volcano are transported in different directions (see figure below). The inversion method compares the different model simulations with the satellite data to estimate the most probable source emissions.

Figure 2: Illustration of the basic principle of the inversion method used for source term determination.

The inversion method can retrieve the source term parameters
  • eruption onset and end time
  • the mass eruption rate
  • eruption column height
  • initial vertical distribution
It is assumed that the location of the volcano and the initial ash particle size distribution are known (or approximated). The method estimates an effective source term that not necessarily depicts the true source emissions but is the source term needed for the atmospheric transport model to simulate the downwind emissions clouds in best agreement with observation data.

The VAST inversion method has already been applied and tested for several different volcanic eruptions. The Grimsvötn-2011 eruption is particularly interesting as ash and SO2 were released to different altitudes and transported to different directions;
  • Ash was released mostly to low altitudes andwas transported mostly south-eastwards.
  • SO2 was released to high altitudes and was transported northwards.

Figure 3: The source term for ash and SO2 emitted during the 2011-Grimsvötn eruption as re-constructed using the VAST inverse modelling system. The height of the emissions, the time of the emissions as well as the strength of the emissions were estimated.

The inversion method was able to distinguish between ash and SO2 and retrieve source terms that gave more accurate transport simulations when validating with independent observation data (Moxnes et al, 2013: in prep).

The VAST inversion method was also used to successfully estimate source terms for the eruptions Jebel at Tair-2007 (Eckhardt, 2008), Kasatochi-2008 (Kristiansen, 2010) and Eyjafjallajökull-2010 (Stohl et al, 2010).


1.      Eckhardt, S., et al (2008), Estimation of the vertical profile of sulfur dioxide injection into the atmosphere by a volcanic eruption using satellite column measurements and inverse transport modeling, Atmos. Chem. Phys., 8, 3881–3897, doi:10.5194/acp-8-3881-2008.
2.      Kristiansen, N. I., et al (2010), Remote sensing and inverse transport modeling of the Kasatochi eruption sulfur dioxide cloud, J. Geophys. Res., 115, D00L16, doi:10.1029/2009JD013286&
3.      Seibert, P. (2000), Inverse modeling of sulfur emissions in Europe based on trajectories, in Inverse Methods in Global Biogeochemical Cycles, Geophys. Monogr. Ser., vol. 114, , pp. 147–154, AGU, Washington, D. C.
4.      Stohl, A., et al (2011) Determination of time- and height-resolved volcanic ash emissions for quantitative ash dispersion modeling: The 2010 Eyjafjallajokull eruption, Atmos. Chem. Phys., 11, 4333-4351, doi:10.5194/acp-11-4333-2011.
5.         Moxnes, E., et al (2013) Separation of ash and sulfur dioxide during the 2011 Grímsvötn eruption – source term estimates and transport modeling constrained by satellite observations, In preparation