Section 16 contains examples of how to setup more complex emissions scenarios. We will use an example of a simplified volcanic eruption throughout, but the methods are pertinent to a wide range of applications.
In this example we will configure the model to create a simple simulation of the Eyjafjallajokull eruption that started 14 April 2010.
On this page we will describe how to do the following.
To simplify the setup, start by retrieving the example control_volc.txt control file from the \Tutorial\volcano directory. We will review the key elements of the configuration. If this tutorial is being run through the web, it will be necessary to download the 1-deg GDAS meteorological data file for weeks w2 and w3 of April 2010 or the equivalent regional extract apr1420.bin prepared for this example. Any pre-configured CONTROL files used in this example will need to be edited to point to the location of the downloaded data file.
- Create emissions from a vertical line source
- Simulate a simple particle size distribution
Open the starting locations menu and note how the two locations are identical in horizontal position (63.63N 19.62W) but with different height. The model interprets this as a vertical line source and will distribute particles uniformly through the layer 100 to 6000 meters AGL.
Open the pollutant menu and note that four pollutants are defined. Lets assume a mass release rate of 10,000 T/hr which means that if we set the emission rate to the equivalent 1.0E+16 μg/hr then the output units will be μg/m3. Note there are four pollutants defined, each representing a different particle size:
| P006|| 0.6μm|| 1%||0.008E+16|
| P020|| 2.0μm|| 7%||0.068E+16|
| P060|| 6.0μm|| 25%||0.250E+16|
| P200||20.0μm|| 67%||0.670E+16|
We will only run the model for 12 hours at a time and therefore emissions are all set to 12 hours. Check the other three pollutants to insure that the settings follow the table suggestions.
Open the grid definition menu where only one vertical level has been defined resulting in average concentrations in the layer from the ground to 10,000 meters. Six hour average air concentrations will be output because we will focus on the particle display output. However, if you want to create animations, a more frequent interval could be set.
The last menu in this section defines deposition for each of the four pollutants. For instance, in the pollutant #1 (P006) menu, the particle diameter is set to 0.6μm assuming a density of 2.5 g/cc for a spherical particle (1.0). Gravitational settling is computed based upon the previous values. The within cloud (8.0E-05 /s) and below-cloud (8.0E-05 /s) wet removal coefficients will be applied to the local ash concentration and precipitation rate for each grid cell. Except for the particle diameter, these values are also applied to the other pollutants.
Save to close the setup run menu and open the Advanced / Concentration Setup / Configuration menu and retrieve the setup_volc.txt file. There were only two changes to the default values. First the particle number release rate and maximum was raised to 5,000 and 50,000. Second, the input and output of the particle files was set to 12 hours.
Save the changes, run the model for 12 hours, and then open the particle display menu to view the ash plume valid at 1200 UTC 14 April 2010. Note how the particles are evenly distributed in the vertical layer.
To complete the exercise, open the concentration display menu and check the All radio-button to sum the four particle sizes into the same graphic. Also fix the concentration contour intervals (100+50+20+10) and units of ug to get the final plume display.
In the next section, we will continue the calculation for another 24 hours.
For particles larger than about 20 micron diameter the Stokes equation overestimates settling. A more accurate formulation by G.H. Ganser (1993, A rational approach to drag prediction of spherical and nonspherical particles. Powder Technology, 77, 143-152) can be used by setting the shape factor to a negative value (<= -1). The volcanic ash plume simulation shown here is only an approximation of the actual event to illustrate the computational process for multiple particle sizes.
(Richard A. Dare, 2015, "Sedimentation of volcanic ash in the HYSPLIT dispersion model" Centre for Australian Weather and Climate Research Technical Report No. 79 pdf )
Optional Exercise: Go back to the deposition menu and change all the particle shapes from positive to negative in order to use the Ganser equation for gravitational settling. Rerun the simulation and compare the results.
External Meteorology Links: