Trouble Running real.exe: Error Persists Even After METGRID.TBL Changes

[Husile Bai]

I try to downscale CMIP6 to WRF, was able to generate the intermediate files and met_em* files, but I keep getting errors when running real.exe. I’ve attached the rsl.out.0000 log.

As part of troubleshooting, I modified METGRID.TBL at name=GHT by changing


interp_option = sixteen_pt+four_pt+average_4pt
to

interp_option = four_pt+average_4pt

but the error still persists.

---
ERROR
Domain 1: Current date being processed: 1984-01-01_06:00:00.0000, which is loop # 1 out of 123
configflags%julyr, %julday, %gmt: 1984 1 6.000000
d01 1984-01-01_06:00:00 Yes, this special data is acceptable to use: OUTPUT FROM METGRID V4.6.0
d01 1984-01-01_06:00:00 Input data is acceptable to use: met_em.d01.1984-01-01_06:00:00.nc
metgrid input_wrf.F first_date_input = 1984-01-01_06:00:00
metgrid input_wrf.F first_date_nml = 1984-01-01_06:00:00
d01 1984-01-01_06:00:00 Timing for input 0 s.
d01 1984-01-01_06:00:00 flag_soil_layers read from met_em file is 1
Max map factor in domain 1 = 1.01. Scale the dt in the model accordingly.
Using sfcprs to compute psfc
i,j = 27 1
target pressure and value = NaN 105.8165
column of pressure and value = NaN 0.0000000E+00
column of pressure and value = 8.517193 20397.31
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 6302
troubles, could not find trapping x locations
-------------------------------------------



Does anyone know if this issue is more likely related to the intermediate files, or could it be caused by the way WRF was compiled? Any suggestions would be greatly appreciated.

 

[kwerner]

Hi, and apologies for the delay in response. If you're still experiencing this issue, it could be a few different things. First, take a look at this similar post to determine if your issue is the same as that user's. If that's not the issue, check your met_em* files to make sure you don't have any missing data - look at all levels, and make sure everything looks reasonable.

If none of the above suggestions are helpful, I'm a bit concerned because, per your namelist.input setting of num_metgrid_levels, the input data only have 8 levels. I'm not sure if there is really a minimum number of levels, but I've never seen anything less than about 20-25 used. I'll have to ask someone about this, but it may be that you need to decrease the number of vertical levels to adjust for the few number of levels.

 

[Ming Chen]

This is a data issue. Data in the intermediate files are not correct, which leads to NaN and unreasonable pressure when you run REAL.

As a quick check, please use ncview to look at your met_em files.