error date

Submitted by stevefletcher395 on Mon, 08/16/2021 - 18:21
Forums
Initialization

I have managed to run the UFS medium range model and it runs to forecast hour 6 then i received this error message

FATAL from PE    81: diag_util_mod::get_date_dif: in variable t2 is less than in variable t1

application called MPI_Abort(MPI_COMM_WORLD, 1) - process 81
Model ended:     Mon Aug 16 11:49:55 MDT 2021
RESULT: ERROR 1
Not sure what is wrong here?

thanks

Permalink

Submitted by Linlin.Pan on Mon, 08/16/2021 - 19:22

Are there enough BC/IC for the forecast run? It seems that something is missing at hour 6?

Thanks,

Linlin

Permalink

Submitted by stevefletcher395 on Mon, 08/16/2021 - 19:52

It is a global run so i am not sure what boundary conditions would be missing.  I have run this configuration before but not with these initial conditions.  I  have write group at 4 and write tasks per node at 6 but have io_layout at 1,1 with layout at 6,4 but would not run at all if i changed io_layout.

but in the gfs_physics fhzero is set to 6 gut forecast hour is set to 10 this might cause a conflict but it does not say anywhere that forecast hours has to be a multiple of 6?

Permalink

Submitted by larissa.reames on Mon, 08/16/2021 - 20:00

Check the date at the top of diag_table. I've gotten this error when the forecast date/time in model_configure is ahead of the date/time listed on the top 2 lines of diag_table.

--Larissa

Permalink

Submitted by Linlin.Pan on Mon, 08/16/2021 - 21:56

Thanks, Larissa! Yes, incorrect date/time will lead missing data for running.

Permalink

Submitted by stevefletcher395 on Tue, 08/17/2021 - 14:53

thanks for the help. it ran, i changed the forecast time as well as the diag_table file, which i spaced to change, yet it still ran for 6 hours with the wrong date there!!

  • 45 views

Copyright © 2023. All rights reserved.