metgrid writing zeroes

[simaberson]

I am making slow progress in running WRF with GEFS data from 2021, but am stuck once again, still in WPS. The ungrib step seems to work fine. For each time, my inventory looks like

PRES TT RH UU VV HGT PSFC PMSL SM000010 SM010040 SM040100 SM100200 SM010200 ST000010 ST010040 ST040100 ST100200 ST010200 SEAICE SOILHGT SKINTEMP SNOW
-------------------------------------------------------------------------------
2013.0 O O O O O O X O O O O O O O O O O O O O O
2001.0 X X X X X X O O X X X O O X X X O X O O O
975.0 X X X X X
950.0 X X X X X
900.0 X X X X X
800.0 X X X X X
750.0 X X X X X
650.0 X X X X X
600.0 X X X X X
550.0 X X X X X
450.0 X X X X X
400.0 X X X X X
350.0 X X X X X
300.0 X X X X X
150.0 X X X X X
70.0 X X X X X
30.0 X X X X X
20.0 X X X X X
7.0 X X X X X
5.0 X X X X X
3.0 X X X X X
2.0 X X X X X
1.0 X X X X X

Plus, I get SKINTEMP from GFS

PRES SKINTEMP
-------------------------------------------------------------------------------
2001.0 X

For metgrid, I get

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
! Successful completion of metgrid. !
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

and there are no errors in the metgrid.log files. But the final files are problematic.

Attached, you can find three grib files I am using (pgrb* - (the system does not seem to like these files since they are larger than 100 megabytes, so I uploaded them to https://www.aoml.noaa.gov/ftp/hrd/aberson/metgrid)). I can confirm that the values for surface pressure (not MSLP) are good in each of these files. Ungrib writes out files of the form GFS* that I am not sure how to read. Then, metgrid combines these files with the SKINTEMP file, and creates netcdf files. When I look at the data (met* files at the same URL as the GFS files), I get:

ncdump -v PSFC met_nmm.d01.2021-08-10_06:00:00.nc

PSFC =
100138.8, 100139.1, 100139.4, 100139.7, 100140, 100140.4, 100141, 100142.8,
100144.6, 100146.3, 100148, 100149.7, 100151.4, 100153, 100154.6, etc.

ncdump -v PSFC met_nmm.d01.2021-08-10_12:00:00.nc

PSFC =
98921.52, 98929.91, 98938.39, 98946.98, 98955.66, 98964.44, 98972.34,
98976.48, 98980.64, 98984.8, 98988.95, 98993.12, 98997.27, 99001.44, etc.

ncdump -v PSFC met_nmm.d01.2021-08-10_18:00:00.nc

PSFC =
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,

This happens (third time with zeroes for multiple sets of files I have tried from different months in 2021. Unless there is a bug, I am doing something wrong, but don't know what.

 

[Ming Chen]

I am suspicious this is a data issue. If the code works just fine for most of the data, it indicates the code works as expected. Please check your input data for the time of 2021-08-10_18:00:00, and see whether PSFC exists in the file.