Interpolation error after creating WPS intermediate/met_em files from NOAA/CIRES/DOE 20th-century reanalysis v3

[koszutam]

I am trying to use the NOAA/CIRES/DOE 20th-Century Reanalysis v3 netCDF files from RDA as my forcing dataset for WRF. I used a python script that uses pywinter to create WPS intermediate files from netCDF files (see forum post here) and I am able to create met_em files for the event where everything looks fine (i.e., no NaNs, correct orientation/domain, etc.). When I run real.exe to produce the WRF inputs an error related to the interpolation I believe. Of note, which I will highlight with bolding in the error found at the end of the rsl.out.0074 file below, is undefined pressure levels in the log that says they are too close together, which in previous posts it is said to be a warning and not an error. But, immediately following that is a "could not find trapping x locations" fatal error. I am particularly curious because the target pressure (~49.7) presumable fall outside of my data (limited to above 70 hPa) and WRF since ptop is set to 100 hPa. Other posts do not have clear troubleshooting or ways to circumvent this, so if you have any suggestions as to why this error is occurring, please let me know. I have attached all of the rsl, namelist, and real.log files. Thanks!


*** -> Check your wrfinput_d01 file, you might have input pressure levels too close together (********* Pa and ********* Pa) at ( 161, 193) for variable type Z
i,j = 161 193
target pressure and value = 49.73200 47.45673
column of pressure and value = 49.73864 148.3134
column of pressure and value = 49.73864 350.8419
column of pressure and value = 49.73864 557.1318
column of pressure and value = 49.73864 767.4288
column of pressure and value = 49.73864 981.8453
column of pressure and value = 49.73864 1424.970
column of pressure and value = 49.73864 1889.470
column of pressure and value = 49.73864 2378.798
column of pressure and value = 49.73864 2896.740
column of pressure and value = 49.73864 3448.099
column of pressure and value = 49.73864 4037.833
column of pressure and value = 49.73864 4672.494
column of pressure and value = 49.73864 5359.122
column of pressure and value = 49.73864 6106.892
column of pressure and value = 49.73864 6927.901
column of pressure and value = 49.73864 7839.048
column of pressure and value = 49.73864 0.0000000E+00
column of pressure and value = NaN -4.2338584E+25
column of pressure and value = 9.903486 11503.08
column of pressure and value = 9.615806 13355.71
column of pressure and value = 9.210340 15952.12
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 5969
troubles, could not find trapping x locations
-------------------------------------------

 

[kwerner]

Hi,
I'll preface by saying I've never worked with the data you're using, so I'm not super knowledgeable about it. Since everything seems okay with the met_em files, for now let's assume they are correct. I see that you have p_top_requested=10000. Did you set to that level on purpose? Does anything change if you set it to the default value of 5000? Is there any difference if you only run real.exe for one domain? I know that isn't your goal, but just for debugging purposes, I'm curious.

I'm not sure if this will provide any useful information for you, but take a look at Choosing an appropriate number of vertical levels (e_vert).