<div dir="ltr"><div><div>Dear WRF users,<br></div><div><br>I'm trying to initialize WRFv3.5 with CMC GEM global data and I'm facing
some troubles, so I would like to ask whether anybody has already
tested something similar.<br>
<br>
I have used Vtable.GFS and everything worked (almost) right.<br>
- ungrib: ran without problems.<br>
- metgrid: I had to set SKINTEMP to "<font face="Courier New">mandatory=no</font>" in
<font face="Courier New">METGRID.TBL</font> in order to avoid the "required field" error<br>
- real.exe/wrf.exe: I set num_metgrid_soil_levels to 0 in
namelist.input (because CMC does not have temperature and moisture at
soil layers below ground)<br>
<br>
With that configuration I've managed to correctly run real.exe, but
wrf.exe starts to run and after some iterations it crashes with "<font face="Courier New">WOULD GO OFF TOP: KF_ETA_PARA..</font>." error.<br>
<br>
I'm pretty sure that it must be related with soil variables initalizacion (*), but I don't know how to cope with that.<br>
<br>
Does anyone have already succesfully tested CMC initialization? Any advice on that?<br>
<br>
Thanks a lot in advance! Regards!<br>
<br>
<br>
(*) Related to the soil variables, I've compared 5-layer TSLB field
(wrfinput) from CMC initialization and GFS (which does provide
below-ground temp and moisture), and there are significant differences
(TSLB field (level=5) from CMC initialization is extremely
cold). I'm sending two figures attached where this can be observed<br><br></div>Eduardo Penabad Ramos<br></div>NWP Dpt. MeteoGalicia - Xunta de Galicia<br></div>