In the last section, the simulation ended after 12 hours, at 1200 UTC 14 April 2010. Assume that now we have new information about the release, for instance that the eruption height maximum is now 12,000 m AGL. So we can restart the simulation at 12Z with this new information, keeping all the previous ash particles on the computational domain.
- Open the setup run menu and change the start time from 10 04 14 00 to 10 04 14 12 and open the starting locations menu and change the top height from 6000 to 12000. Save the changes and exit.
- In the previous run we already set the input/output particle file options to output the endpoints every 12 hours. Note that one of the default options is to read the particle positions in a file called PARINIT when the model simulation starts. Therefore, go to the \hysplit4\working directory and rename the file PARDUMP created at the end of the last simulation to PARINIT. Then save the changes and run the model.
- After the simulation completes, open the display menu to view the particle cross-section on the 15th at 00Z. The different release height periods are clearly evident in the display.
- Go through this process one more time, change the start time to 10 04 15 00 and change the release height maximum back to 6000 m. Save the changes, and go back to the working directory and rename PARDUMP to PARINIT. You need to delete or rename the previous PARINIT. Then run the model. If uncertain, you can always confirm that the particle file is being opened by viewing the MESSAGE file. When the run completes, open the display to view the new particle cross-section on the 15th at 12Z. It should be evident that temporally varying release heights can result in quite complex plume structures.
In this section we illustrated how the standard model input procedure using the CONTROL file can be used to create a complex time-varying emission scenario. When the model is restarted, any of the model simulation parameters, either in the CONTROL file, the SETUP.CFG file, or even the meteorology file can be changed. In the next section, we will compare the model results with observational data and after that we will run the volcanic ash simulation again but using an single EMITIMES file to define how the emissions vary in height and time. Do not delete the PARDUMP file because it will be used in the next section.