Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

metgrid.exe error: Field (PRES, HGT, HGTTROP, TTROP, PTROPNN etc.) has missing values at level 200100

Y

Yongxin

Member
I have been trying to run WPS (version 4.2) for nested (3km -> 1km) Colorado Front Range domains driven by GFS Final Analysis (FNL) data. The met_em files for the 3km Domain01 have always been created successfully but no met_em files have been created for the 1km Domain02 and the error messages are always something like the following:

Processing domain 2 of 2
Processing 2013-09-10_00
GFS
WARNING: Field PRES has missing values at level 200100 at (i,j)=(21,5)
WARNING: Field GHT has missing values at level 200100 at (i,j)=(21,5)
WARNING: Field HGTTROP has missing values at level 200100 at (i,j)=(21,5)
WARNING: Field TTROP has missing values at level 200100 at (i,j)=(21,5)
WARNING: Field PTROPNN has missing values at level 200100 at (i,j)=(21,5)
WARNING: Field PRES has missing values at level 200100 at (i,j)=(27,5)
WARNING: Field GHT has missing values at level 200100 at (i,j)=(27,5)
WARNING: Field HGTTROP has missing values at level 200100 at (i,j)=(27,5)
WARNING: Field TTROP has missing values at level 200100 at (i,j)=(27,5)
WARNING: Field PTROPNN has missing values at level 200100 at (i,j)=(27,5)
.......

I have tried with the 1-degree and 0.5-degree GFS FNL data and also the 0.25-degree ERA5 data and the error messages from the three runs are similar although the (i,j) where the error messages occur are different. I searched the forum for the similar issues and the reason seems to be that my input data did not cover my nested domains but I used the global GFS and ERA5 data. Very strangely, when I changed the nested domains from 3km->1km to 1km -> 333m (keeping the domain sizes roughly the same) the met_em files were created successfully for both domains.

I attached here the namelist.wps, metgrid.log.0000, and runtime info (run_ungrib_metgrid_GFS.out and run_metgrid_ERA5.out) files for both the GFS- and ERA5-driven WPS runs. For ERA5, I first used the era5_to_int utility to convert from netCDF to WPS format and then ran metgrid.exe. I am wondering if any of you ran into the same issue or may have suggestions for a solution.

Thanks,
Yongxin
 

Attachments

  • namelist.wps.GFS.txt
    1.2 KB · Views: 1
  • namelist.wps.ERA5.txt
    1.2 KB · Views: 0
  • metgrid.log.0000.GFS.txt
    650.5 KB · Views: 0
  • metgrid.log.0000.ERA5.txt
    3.6 MB · Views: 0
  • run_ungrib_metgrid_GFS.out.txt
    493.1 KB · Views: 0
  • run_metgrid_ERA5.out.txt
    2.4 KB · Views: 0
Hi Yongxin,

I repeated your case, using your grid settings. I only change the time of GFS input data.
WPS-v4.6.0 works just fine. Please see the namelist.wps I used. Can you try with my namelist.wps, and change the date for your specific case? Let me know if you still have any issues.

&share


wrf_core = 'ARW',


max_dom = 2,


start_date = '2019-10-01_00:00:00','2019-10-01_00:00:00','2025-08-20_18:00:00','2025-08-20_18:00:00',


end_date = '2019-10-01_06:00:00','2019-10-01_06:00:00','2025-08-20_18:00:00','2025-08-20_18:00:00'.


interval_seconds = 21600


io_form_geogrid = 2,


/


start_date = '2020-06-26_00:00:00','2006-08-16_12:00:00',


end_date = '2020-06-27_00:00:00','2006-08-16_12:00:00',





&geogrid


parent_id = 1, 1,


parent_grid_ratio = 1, 3,


i_parent_start = 1, 16,


j_parent_start = 1, 29,


e_we = 70, 52,


e_sn = 70, 31,


geog_data_res = 'usgs_30s+default','usgs_30s+default',


dx = 3000,


dy = 3000,


map_proj = 'lambert',


ref_lat = 40.0,


ref_lon = -105.0,


truelat1 = 30.0,


truelat2 = 60.0,


stand_lon = -105.0,


geog_data_path = '/glade/work/wrfhelp/WPS_GEOG/'


/

&ungrib


out_format = 'WPS',


prefix = 'GFS',


/

pmin=100.

&metgrid


fg_name = 'GFS'


io_form_metgrid = 2,


/
 
Hi Ming,

Thank you very much for looking into this issue. My WPS version is 4.4. I used your namelist.wps and tried with 1) starting from 2013-09-10_06 instead of 2013-09-10_00 and 2) using a different date of 2018-09-10_00 instead of 2013-09-10_00 but I still got the same missing value issue. Since I have tried with 1-degree FNL data, 0.5-degree FNL data, and 0.25-degree ERA5 data and all gave me the same missing value problem, it seems to me that something may be wrong with my configuration but I could not figure that out. I am wondering if you could take a look at my namelist.wps file located under /glade/derecho/scratch/zhangyx/WRF-Hydro/MPAS-Hydro/WPS-domain-nested on Derecho.

I also tried with the pre-compiled WPS V4.6 that I copied from /glade/work/wrfhelp/derecho_pre_compiled_code on Derecho but somehow I got "ERROR: Error in ext_pkg_open_for_write_begin" error. I will try to compile my own WPS V4.6 on Derecho.

Thank you Ming.

Yongxin
 
Hi Yongxin,

Your namelist.wps looks fine to me.

I copied your namelist and repeated your case in /glade/derecho/scratch/chenming/wpsv4.6.0. The only difference is that I changed the time period from 2013 to 2019.

Can you copy my namelist.wps and GFS data, then rerun under your account? Let me know whether it works.

If you still get the same error, then I am suspicious that either your codes are not compiled correctly, or your environment settings are wrong.

Please keep me updated how it works.
 
Hi Ming,

I copied over the pre-compiled WPS V4.6 as I mentioned earlier and somehow geogrid.exe did not work successfully complaining about "ERROR: Error in ext_pkg_open_for_write_begin" but ungrib.exe and metgrid.exe worked using your namelist.wps file and generated the met_em files for both Domain01 (3-km resolution) and Domain02 (1-km resolution) for the time period that I will need. This is the first time that I got the met_em files for the nested Domain02. Thank you so much for your help with this!

I have been trying to build my own WRF v4.7.1 (or v4.6.0) and WPS v4.6.0 using "cmake" but I keep running into problems with something like "Building Fortran object CMakeFiles/WRF_Core.dir/external/RSL_LITE/feedback_domain_em_part1.F.o". I will send a separate post on the Focum for this.

Thanks,
Yongxin
 
Yongxin,

It is weird that the precompiled geogrid.exe of WPS V4.6 doesn't work for you. You can see that in my test, I actually run WPSV4.6 and I cannot repeat your issue.

Let me know if you finally figure put what is wrong in your compiling/environment settings, --- I never see such weird issues before.
 
Hi Ming,

Thank you so much for your messages!

I copied over your namelist.wps file and your GFS data to /glade/derecho/scratch/zhangyx/WRF-Hydro/MPAS-Hydro/WPS-domain-nested and ran metgrid.exe and I got the same missing value errors as before. This indeed suggests that either my WPS compilation or my environmental settings are not right. I will try to figure this out and will let you know.

Thanks again,
Yongxin
 
You must log in or register to reply here.
Top