Dispersion modelling in EU-SENSE

FOI develops model components in the EU-SENSE system to be able to estimate source parameters in relation to hazardous chemical releases. Data from the sensor measurements are used to estimate both the location of the source as well as the source strength. This estimation is then used to calculate concentration fields and risk areas for a better understanding of the areas influenced by the release with respect to first responder operations. We have had the opportunity to talk to research engineer Mr. HÃ¥kan Grahn about the work done at the dispersion modelling group at FOI.

What is the current status of the work?

  • At present, we are focusing on setting up and develop the system interface to our model tools, Mr. Grahn says. In the EU-SENSE system, we will use the FOI Dispersion Engine (DE) and we need to be able to access this tool from the EU-SENSE system. Therefore, we have specified the communication protocols so that DE can be reached externally via the internet.

What kind of model tools are you developing?

  • In EU-SENSE, we are responsible for developing two model prediction tools, Threat Estimation tool (TE tool) and Hazard prediction tool (HP tool). These tools will provide the EU-SENSE system with the estimation of source parameters and calculate risk areas.

As I understand, this is quite a challenging problem. Can you describe the work in more detail?

  • Of course. First of all, we need measurements from the sensor nodes developed in the project. The quality of the measurements is of utmost importance and we need to know if the sensor signal is significant or just a measure of the natural background. When having these quality controlled sensor signals the TE tool is starting to estimate the source parameters via an optimization process. Besides the sensor information, weather and geographical information are needed. Therefore, we are also running a weather module to estimate the best weather input.

It seems that you need quite many parts to be able to produce the results?

  • Yes, therefore we have chosen a software solution based on separate modules communicating with the DE API.

You mentioned that you will calculate risk areas as well, right?

  • Certainly, first responders are mainly interested in this so proper actions can be taken. We will use our Lagrangian Particle model on the urban scale to do these calculations. It is a proven model which at the end of the project can use different geometries.

Can you give an example of work done on source estimation and risk area calculation?

  • We are just about to start with the specific EU-SENSE applications but we have run a study using DE on the regional scale. In April 2019 elevated levels of radiation were measured around Europe. We used backtracking to estimate the location of the source as well as source strength. The location of the source was calculated to be in the Chernobyl area and together with intelligence information which informed us about an ongoing forest fire in that area we draw the conclusion that this was the source. I think this proves the advantage to use backtracking techniques and that these are mature to be used in real decision making. However, success is much dependent on the quality of the sensor data and having the correct flow field. We have made a short movie showing the calculated plume from the forward dispersion calculation using the best source estimate, see the movie further down.

What does the EU-SENSE project mean for you? It means a lot for us, we basically have all bits and pieces already prepared but this is an opportunity for us to put them into a system and to connect our tools to real sensor measurements. I think the EU-SENSE system can be a vital part of future first responder operations. Also, this is a lot of fun, Mr. Grahn concludes.