More throughput by cloning

Post any questions or comments regarding the Mac OS X version of HYSPLIT. This includes the model execution, GUIs, results, or graphics. Be sure to mention the version of OS you are using.
Post Reply
Posts: 4
Joined: June 16th, 2017, 2:41 pm
Registered HYSPLIT User: Yes
Location: East Coast US

More throughput by cloning

Post by sch576 » June 17th, 2017, 2:05 pm

I am using igor to write CONTROL files and write a shell script to batch process my runs.

looping the sequence below, for example, from the working directory ...
cp <my igor prepared CONTROL> ~/Hysplit4/working
cp ~/Hysplit4/working/qdump <orderly location for archival>
../exec/con2asc -iqdump -oBTX
mv BTX* <orderly location for import back to igor>

If I was to clone the Hysplit4 directory to say Hysplit4b and I was to clone the met files with the attendant redirect in generated CONTROLs, could I run another line in parallel?


Posts: 73
Joined: February 8th, 2016, 12:56 pm
Registered HYSPLIT User: Yes

Re: More throughput by cloning

Post by alicec » June 19th, 2017, 10:24 am

You can run more than one hycs_std at a time. This works best if each one is run in its own working directory. So for instance you could have some kind of loop like this.

Make directory to run hysplit in. Change to that directory. Create CONTROL, SETUP.CFG (if used) and ASC.CFG files in that directory. Call hysplit.

Make another directory to run hysplit in. Change to that directory. Etc….

The directories don’t have to be called working. That is the default working directory but hyc_std will run in whatever directory it is called from. So you could call your directories working1, working2, working3, etc. (or whatever else you want to call them).

You might need to copy the meteorological files if the runs are trying to access the same meteorological files. If they are trying to access the same files at the same time, the run might fail.

You do not need to clone the whole hysplit4 directory. You need only one copy of the hycs_std executable.

There is another trick which can use to run hyc_std in the same directory. You can add a suffix to the CONTROL file CONTROL.001, CONTROL.002, CONTROL.003 etc. Then you can do the same thing with the SETUP files SETUP.001, SETUP.002, SETUP.003. Then if you call hycs_std 001 it will use the CONROL and SETUP files with the 001 suffix. This is very convenient if you want to run a bunch of hys_std sequentially in the same directory without overwriting your CONTROL files. It doesn’t work so well for running hycs_std at the same time because occasionally the hycs_std runs will both be trying to access the ASCDATA.CFG file at the same time and so every once in a while a run will fail. (you would also have to make sure that you name your cdump files something different in the CONTROL file).

So for running multiple hycs_std at the same time we recommend creating a separate directory for each run.

Post Reply