Averaging HYSPLIT Outputs

[rotendd]

I am looking to run HYSPLIT for an entire month but looped for 48hr periods. For example: Feb 1st - Feb 2nd, Feb 3rd - Feb 4th, Feb 5th - Feb 6th, etc. but then average them in one output file. Any assistance would be greatly appreciated!

[ariel.stein]

In the GUI, go to Concentration/Utilities/Binary file/Merge

[rotendd]

Excellent! But is there a way to automate the collection of the individual files? For example, is there a batch file that I can use to just click start once and have HYSPLIT output data for 48hr periods?

Thanks again!

[ariel.stein]

Another way do your runs will be to set your concentration averaging time to 48 hours in Concentration/Concentration Setup/Pollutant,Deposition, and Grid setup/ in the last entry of the window use 00 48 00.

[rotendd]

Excellent! But is there a way to automate the collection of the individual files? For example, is there a batch file that I can use to just click start once and have HYSPLIT output data for 48hr periods?

I would like to add on to this previous question. Is there a way to loop HYSPLIT over 48hr periods and RESTART the concentration dispersion every time? It would greatly speed up my project if I could (from just a few clicks) start HYSPLIT for 48hrs (Feb 1st - Feb 2nd), save the file, then RESTART the dispersion and run for another 48hrs (Feb 2nd - 3rd), save an additional file, etc. And finally, in the end merge all of these individual files together.

Specifically, I need a years worth of averaged 48hr dispersions from a point source.

All of the help thus far has been greatly appreciated and has definitely pointed me in the right direction. I'm still searching for a method to fully automate all of the iterations. Having HYSPLIT restart the dispersion every 48hrs is the difficult part for me.

Thanks again!

[ariel.stein]

You can write the particle positions into a PARDUMP file that can be used as initialization for the next 48-hour run.
See the following link for details on how to implement this. http://www.ready.noaa.gov/documents/Tut ... _pard.html
Also, notice that below the page title you have links to scripts that will give you an example of how to program the initialization with a PARDUMP file.

[kenafam]

Excellent! But is there a way to automate the collection of the individual files? For example, is there a batch file that I can use to just click start once and have HYSPLIT output data for 48hr periods?

I would like to add on to this previous question. Is there a way to loop HYSPLIT over 48hr periods and RESTART the concentration dispersion every time? It would greatly speed up my project if I could (from just a few clicks) start HYSPLIT for 48hrs (Feb 1st - Feb 2nd), save the file, then RESTART the dispersion and run for another 48hrs (Feb 2nd - 3rd), save an additional file, etc. And finally, in the end merge all of these individual files together.

Specifically, I need a years worth of averaged 48hr dispersions from a point source.

All of the help thus far has been greatly appreciated and has definitely pointed me in the right direction. I'm still searching for a method to fully automate all of the iterations. Having HYSPLIT restart the dispersion every 48hrs is the difficult part for me.

Thanks again!

I am looking for the same thing, how did you automate the iterations, or did you not end up doing it?

[ariel.stein]

To automate your runs please open the Concentration / Special Runs / Daily
in the GUI

This special menu can be used for concentration
simulations to run automated multiple simulations for one month. The
script will replace the starting day and hour in the CONTROL file with
values generated in the script from the menu shown below. In this
example the model will be run four times a day (at 0, 6, 12, and 18
UTC) for each day from 1 through 31. Note that the start-hour field
is just treated as a string and can be replaced with any sequence of
starting hours. The model output files are named according to the fields
set in the control file and namelist file menus, but are appended with
the four digit day and hour field. This approach differs from a single
simulation with multiple starting times in that each simulation is
independent and a new output file is created for each run.

The model should be configured and run for the first simulation time
through the standard run menu to insure that the simulation is configured
correctly. If part of the simulation requires meteorological data from the
previous month or the next month, these should be included in the base
simulation test.

At first it will appear that after pressing the "Execute Script"
button that nothing is happening. However, as each day's simulation is
completed, the "Start Day" entry box increases in value until its value
matches the "End Day", at which time the script is completed. Note that
concentration simulations are much slower than trajectories and it may
take a while before the start day field is updated.

The dispersion daily run menu contains an additional entry
that determines how the file name is calculated for each run. The default
approach is the same as with the trajectory calculation, where the day and
hour is appended to the base name. Another option is to name the files
sequentially using a three digit number. Using sequential numbering, the
output files can be processed by the probability display programs used
in conjunction with the ensemble calculation.