Error when restarting coupled WRF-ROMS

[derrickkdanso]

Hi!

I have successfully compiled COAWST and run a coupled WRF-ROMS test application. However, I get an issue when trying to restart the run. It seems ROMS cannot find the surface winds:

DEF_HIS - unable to find variable: Uwind in history NetCDF file: /global/cscratch1/sd/ddanso/ROMS_WRF/COAWST/Projects/Test/output/GOM_ocean_his.nc Found Error: 03 Line: 159 Source: ROMS/Nonlinear/output.F Found Error: 03 Line: 567 Source: ROMS/Nonlinear/main3d.F Found Error: 03 Line: 331 Source: ROMS/Drivers/nl_ocean.h, ROMS_run

I checked the ROMS history file and saw that both Uwind and Vwind have zero values over the whole domain. I previously run and restarted both ROMS and WRF individually in COAWST with no problems, so I am confused. Perhaps I missed something or made a mistake somewhere.

Please could you help me solve this issue? I have attached the full error log and input+header files. Thank you.

P.S: To restart I changed,

ROMS: NRREC from 0 to -1 LDEFOUT from T to F ININAME to ocean_rst.nc

WRF: The start dates restart from false to true

-D

coupling_out.log gom_h.txt ocean_gom_in.txt namelist_input.txt coupling_gom_in.txt

[jcwarner-usgs]

not sure. i usually do not change that LDEFOUT. what if you leave it at T?

you might have to start over, do a short run, then i usually copy the rst file to a new name or move it to a new folder. set the nrrec to -1, and change ininame to where you just moved the old rst. this way you have that rst for future simulations. if you let roms write over it, then the data will be overwritten.

[derrickkdanso]

Thanks John.

Leaving LDEFOUT at T works just fine, although I wanted to append the new ROMS history data to the one created from the first run and keep that in case I need it sometime.

I will just copy the ‘his’ file to a new name just like your suggestion for the rst. Thanks again. -D

[jcwarner-usgs]

i think what you want to do is still achievable with ldefout T. look at your ndefhis settings, etc. if you do a 24 hour run, and do ndefhis to be 6hours, it will create 4 files. then if you restart, it will make a 5th file and keep going. -j

[derrickkdanso]

Thank you but I am getting the same error I had previously.

[jcwarner-usgs]

not sure what you mean. Does it all work if you use LDEFOUT = T?

if it is failing now, can you set NDEFHIS to create at least 2 his files before the end of the run and try to restart with that.

[derrickkdanso]

That is exactly what I did. LDEFOUT is set to T. I have also set NDEFHIS to create a file for each day of the simulation period. So I get multiple history files. But I get this error when I restart:

TIME-STEP YYYY-MM-DD hh:mm:ss.ss KINETIC_ENRG POTEN_ENRG TOTAL_ENRG NET_VOLUME C => (i,j,k) Cu Cv Cw Max Speed

38043240 2017-08-24 18:16:00.00 7.939189E-03 2.324937E+04 2.324938E+04 1.042942E+16 (002,109,16) 3.323992E-04 7.296551E-04 0.000000E+00 2.481089E+00 DEF_HIS - inquiring history file, Grid 01: /global/cscratch1/sd/ddanso/ROMS_WRF/COAWST/Projects/Test/output/GOM_ocean_his_01762.nc

DEF_HIS - unable to find variable: Uwind in history NetCDF file: /global/cscratch1/sd/ddanso/ROMS_WRF/COAWST/Projects/Test/output/GOM_ocean_his_01762.nc Found Error: 03 Line: 149 Source: ROMS/Nonlinear/output.F Found Error: 03 Line: 567 Source: ROMS/Nonlinear/main3d.F Found Error: 03 Line: 331 Source: ROMS/Drivers/nl_ocean.h, ROMS_run