Running `ungrib` on complete ERA5 global atmospheric reanalysis sigma levels

[bigveech]

Hi everyone,

I’m working to initialize an MPAS run using initial conditions from the full ERA5 reanalysis dataset. Since my simulations target much higher altitudes (>100 mb), I’m using the 137 model levels rather than the standard 37 pressure levels to achieve higher resolution in these upper layers.

I noticed a Vtable.ECMWF_sigma file in the Variable_Tables directory, but it seems outdated and causes the following error:

ERROR: Vtable does not contain Grib2 decoding information. 11 or 12 columns of information is expected. *** stopping parse_table ***

I also tried Vtable.ECMWF, but that didn’t work since it’s configured for pressure levels.

Does anyone have suggestions for creating WPS intermediate files from this dataset or a Vtable compatible with these model levels? Any tips would be greatly appreciated!

 

[Ming Chen]

Would you please upload a single complete ERA5 data for me to take a look?

We didn't process ERA data from Copernicus before and I am curious how it would work with MPAS.

Thanks.

 

[bigveech]

I was able to initialize MPAS using the ERA5 model level data. I'll document the steps I took.

1. Download the data from Copernicus using CDSAPI

c = cdsapi.Client()

# Model levels
c.retrieve('reanalysis-era5-complete', {
    'date': '2023-10-16',
    'levelist': '1/to/137/by/1',
    'levtype': 'ml',
    'param': '129/130/131/132/133',
    'stream': 'oper',
    'time': '12',
    'type': 'an',
    'area': 'global',
    'grid': '0.25/0.25',
    'format': 'grib',
}, '20231016.era5.t12z.global.ml.grib')

# Surface levels
c.retrieve("reanalysis-era5-complete", {
    "date": "2023-10-16",
    "levtype": "sfc",
    "param": "31/34/39/40/41/42/43/67/129/134/139/141/151/165/166/167/168/170/172/183/235/236",
    "stream": "oper",
    "time": "12",
    "type": "an",
    "area": "global",
    "grid": "0.25/0.25",
    "format": "grib2",
}, "20231016.era5.t12z.global.sfc.grib")

2. Run `ungrib` twice, model levels and surface levels. Note: as of today (release 4.6.0) the `Vtable.ECMWF_sigma` file is out of date in the WPS repository. I've attached an updated version here. I don't know if it's fully correct but it's been working for me. You should now have 2 WPS intermediate files:

HYB:2023-10-16_12
SFC:2023-10-16_12

3. Concat files to form single dataset:

cat HYB:2023-10-16_12 SFC:2023-10-16_12 > ALL:2023-10-16_12

4. Run `calc_ecmwf_p` WPS util program. More details here: WRF Guide: WPS Utility Programs. I was able to generate my own `ecmwf_coeffs` file for the 137 model level data using the `n`, `a[Pa]`, and `b` columns from this table. I'll attach mine in case anyone wants to do this using the 137 model levels (remove `.txt` extension). This will output an additional WPS intermediate file `PRES:2023-10-16_12`.

5. Concat pressure intermediate file with rest of outputs:

cat ALL:2023-10-16_12 PRES:2023-10-16_12 > ERA5:2023-10-16_12

I was able to use this final intermediate file to initialize the MPAS model using `init_atmosphere`.

 

[Ming Chen]

@bigveech

Thank you so much for the kind information ! I guess many Copernicus users will benefit from your work. Thanks again !