[wrfems] A Slow Model Run

[Jim Moser]

Ok...after a few more calculations...I've realized the grid is too big.

Current model run has 145x150 gridpoints for a total of 21,750. This 
runs in 50 mins.
Proposed new model has 395x403 gridpoints for a total of 158,790. 
Dividing 158,790
   by 21,750 and multiplying by 50 mins I get a little over 6 hours.

But...might there be other reasons that its taking so long.

Jim



Jim Moser wrote:
> Well...I ran the ems_install.pl script about a month ago and it seemd 
> to go just fine.
> So I have the 3.1 version of the wrf installed. I've worked on 
> configuration  files
> for the last 2 days using the included guide(Thank You Bob!).
>
> Attempted our first run tonight just to see if it would fly. All 
> seemed to go fairly well.
> After almost 2 hours of computation on our 16 cpu modelling box I 
> checked the output
> using the tail command as suggested.
>
> After 1 hr and 51 mins of computation time it had completed 3 hrs 
> (thats right 3 hrs) of the model
> run. The listing showed a basic timestep of 8.6 sec with every sixth 
> timestep taking 20-40 secs.
> To begin with this is a fairly big ARW grid. It goes from CHI to MSY 
> and AVL to FSM. Resolution
> is 3 km. Maybe I've bitten off more than I can chew here. At any rate 
> I'm attaching the run_info.txt
> file for a more accurate description.
>
> If anyone has any ideas about where to start checking...please speak up!
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> wrfems mailing list
> wrfems at comet.ucar.edu


-- 
-----------------------
Jim Moser
WFO Nashville
Warngen Focal Point
----------------------

"The urge to save humanity is almost always only a false-face 
for the urge to rule it."  H. L. Mencken