bartbrashers
New member
With WRF-4.0.x and WRF-4.1, when you try to restart a run that uses Thompson microphysics, an error occurs:
Deleting the file freezeH2O.dat first allows the restart to work OK – WRF runs.
The LUN for opening OBSDOMAINd01 is dynamically set in share/wrf_fddaobs_in.F (line425):
The LUN for opening freezeH2O.dat is hard-coded in phys/module_mp_thompson.F (line 4065):
A possible solution is to simply close the file after reading it in phys/module_mp_thompson.F (starting line 4065):
I have verified with WRF-4.1 that this fix allows WRF to restart without the FATAL error. But I have not followed every possible if-then outcome to verify that is the best solution.
Bart Brashers
Code:
# tail wrf.rsl.error.0000
d03 2016-01-24_12:00:00 Input data is acceptable to use: wrflowinp_d03
d03 2016-01-24_12:00:00 Input data processed for aux input 4 for domain 3
OBS NUDGING: Reading new obs for time window TBACK = -1.000 TFORWD = 1.000 for grid = 3
File open on obs nudging unit ( 63 ) with wrong name
File open on obs nudging unit is named freezeH2O.dat but it should be named OBS_DOMAIN301
This likely means this unit number was opened elsewhere in WRF
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: wrf_fddaobs_in.b LINE: 460
wrf_fddaobs_in: in4dob STOP Obs nudging file name mismatch
-------------------------------------------
Deleting the file freezeH2O.dat first allows the restart to work OK – WRF runs.
The LUN for opening OBSDOMAINd01 is dynamically set in share/wrf_fddaobs_in.F (line425):
Code:
obs_domain_file_unit = NVOLA+INEST-1
The LUN for opening freezeH2O.dat is hard-coded in phys/module_mp_thompson.F (line 4065):
Code:
OPEN(63,file="freezeH2O.dat",form="unformatted",err=1234)
A possible solution is to simply close the file after reading it in phys/module_mp_thompson.F (starting line 4065):
Code:
CALL wrf_message("ThompMP: read freezeH2O.dat instead of computing")
OPEN(63,file="freezeH2O.dat",form="unformatted",err=1234)
READ(63,err=1234)tpi_qrfz
READ(63,err=1234)tni_qrfz
READ(63,err=1234)tpg_qrfz
READ(63,err=1234)tnr_qrfz
READ(63,err=1234)tpi_qcfz
READ(63,err=1234)tni_qcfz
CLOSE(63) ! BAB proposed solution
good = 1
I have verified with WRF-4.1 that this fix allows WRF to restart without the FATAL error. But I have not followed every possible if-then outcome to verify that is the best solution.
Bart Brashers