"to many records"

[ln-x]

Hi,

I could run EMEP with wrf-forcing of about one month length now. Now I am trying to using wrf-forcing (of the same grid and setup but) with 356 days. I set the startdate and enddate in config_emep.nml according to the first and last wrf-time step, but receive a:

STOP-ALL ERROR: to many records in ./wrfout_d02_2020-01-10_00:00:00
Abort(9) on node 0 (rank 0 in comm 0): application called MPI_Abort(MPI_COMM_WORLD, 9) - process 0

It is called in "NetCDF_mod.f90:4582,4535,4770: call CheckStop(SIZE(TimesInDays)<NTime_Read,"to many records in "//trim(fileName))

If I change the end date to a previous date I get the same error.

Heidi

[ln-x]

It runs after making a cdo mergetime operation with the WRF file, which transformed the Times (Time,DateStrLen) variable to XTIME (XTIME). Before the variable was an array of datetime steps (2020-01-01-00:00:00) afterwards an array of integers (minutes since 1990-01-01-00:00:00)

[ln-x]

The simulations completed without any errors and saving the correct timesteps, but the results forced by the adapted WRF-file using the XTIME variable (last year in graph - 2020) look as if something has been read in wrong.
I checked T2, SWDOWN bottom and wind of the WRF-file. I could not find any mistake there.

Any hint where I can look for errors?

[gitpeterwind]

Hi Heidi,
It is difficult to say without looking at the WRF file.
You can check that the time variables are defined with the right attribute.
The time is read by the routine ReadTimeCDF in NETCDF_mod.f90; you could try to print out data from there to see if it is correctly interpreted.
Otherwise, you can try to send me a sample of your WRF file (one or two days is probably enough), so that I can try myself.

Best wishes,
Peter

[ln-x]

Thank you! I am waiting on a job to run on VSC. Maybe now everything is fine. If not I will send you a file of 1-2 days.