Using era5_to_int.py for ERA5 Pressure-Level Input but Getting Requests for ML Files

[Henry18]

Hi,

I’m using the era5_to_int.py tool to process ERA5 data into the intermediate files required for met_grid. I would like to use pressure-level (pl) data as the input. I have collected the pl, surface (sfc), and invariant datasets in my data directory, and I’m running the following command:

era5_to_int.py -p /glade/derecho/XXXXXXXX -i 2023-08-20_00 2023-08-28_00 1

However, I am encountering the following error:

Error: Could not find file e5.oper.an.ml.128_134_sp.regn320sc.2023082000_2023082005.nc needed for ERA5 variable SP

My questions are:

Is this model-level (ml) file actually required for the process?

Is this the only ml variable/file needed (perhaps only for initialization)?

Since I want to use pl data as input (using the -i option), I would prefer not to include ml files, as pl and ml have different resolutions and may affect the results. In a previous experiment, I collected the ml files during the study period in response to this error message, but the resulting WRF output contained irregular discontinuous edges (as shown in the attached images). This time I’m hoping to avoid any influence from ml fields.

Thank you for your help!
Henry

 

[Henry18]

Hi,

This is an update.

I checked the pl files in my data directory, and I believe the following should be the correct file for SP when using pl data as input:

e5.oper.an.sfc.128_134_sp.ll025sc.2023080100_2023083123.nc

And I use these two as the corresponding invariant files:

e5.oper.invariant.128_129_z.ll025sc.1979010100_1979010100.nc
e5.oper.invariant.128_172_lsm.ll025sc.1979010100_1979010100.nc

All of them has the "ll025sc", so I think they match the resolution of the pl data. However, even though these files are present in my data directory, when I run the era5_to_int.py tool to process the ERA5 pl .nc files into intermediate files, I still get the following error:

Error: Could not find file e5.oper.an.ml.128_134_sp.regn320sc.2023082000_2023082005.nc needed for ERA5 variable SP

Is it possible to configure the tool to use "e5.oper.an.sfc.128_134_sp.ll025sc.*" as the input for the SP variable instead of looking for the ml version?

Thank you for your help!
Henry

 

[kwerner]

Hi,
If you are still experiencing this issue, since you're on the Derecho HPC, can you point me to the directory where you're running? Thanks!

 

[Henry18]

Hi,
If you are still experiencing this issue, since you're on the Derecho HPC, can you point me to the directory where you're running? Thanks!

Hi Kwerner,

Thank you for your reply. Below is my WRF running directory:

/glade/derecho/scratch/hhou/Test_ERA5/WRF/test/em_real

I proceeded to run with the (possibly incorrect) met_em.* files anyway. Here is the WPS directory I used to generate the intermediate and met_em files:

/glade/derecho/scratch/hhou/Test_ERA5/WPS

And here is the data directory specified with the -i flag:

/glade/derecho/scratch/hhou/TestData/NCData_ERA5

Thank you for your help!
Henry

 

[kwerner]

Hi Henry,
Thanks for sharing that information. I looked at your files (both the ERA5* intermediate files, and met_em* files), and from what I can tell, they seem to look correct. I'm now trying to figure out your WRF issue (mentioned here). I'll address it in that thread when I have some useful information.

 

[Henry18]

Hi Henry,
Thanks for sharing that information. I looked at your files (both the ERA5* intermediate files, and met_em* files), and from what I can tell, they seem to look correct. I'm now trying to figure out your WRF issue (mentioned here). I'll address it in that thread when I have some useful information.

Hi Kwerner,

Thank you for your help! I'm glad to hear that the intermediate files and the met_em* files look correct. I’ve used these files to run the simulation, and after setting the time_step to 6, the model was able to complete the 8-day run with multiple restarts.

However, when I checked the output using the ncview module, I still noticed some strange box-shaped inconsistencies in the temperature-related variables. I’ve attached both the T2 and TMN fields for d01 and d02. In the simulation I used:

physics_suite = 'CONUS'

These inconsistencies seem like errors to me, since I haven’t seen them in others’ output. Previously, when I forced the same simulation with NARR ungrib data (without using the ‘CONUS’ physics suite), I didn’t see these inconsistencies in T2, but I did notice similar issues in TMN. For your reference, here is the directory of my previous simulation:

/glade/derecho/scratch/hhou/Test_NARR/WRF/test/em_real

Do you have any suggestions on what might be causing this?

Thank you for your time and help!

Cheers,
Henry