ERA5 as input - SST interpolation issues

[tiziana]

Aloha, I am trying to use the ERA5 data as input for a high resolution WRF run over Hawaii. After searching in this forum I stumbled upon the script era5_to_int to process the ERA5 data (netcdf) at NCAR. I am running WPSv4.5 on Casper. Attched are my namelist.wps and METGRID.TBL. Could, please, anyone explain me the good practice to handle SST, SKINTEMP, SEAICE when using ERA5 (from NCAR netcdf dataset d633000) in WPS and the data from era5_to_int? After running metgrid on the ERA* files created via era5_to_int, I obtained the SST field in the ncview screenshot in attachment, which shows interpolation issues along the coastlines. As suggested in the forum, in METGRID.TBL I used:

name=SST
interp_option=sixteen_pt+four_pt+wt_average_4pt+search
interp_mask=LANDSEA(1)
masked=land
fill_missing=0.
missing_value=-1.E30
flag_in_output=FLAG_SST

I also see issue in the SEAICE field (see attached image). I am also attaching LANDSEA and LANDMASK and SKINTEMP.

On a different note, in the LANDMASK image, I see two tiny dots (water?) in the middle of the Big Island, which should not be there (better visible in the SEAICE image). I also run WRF with GFS for other cases, and when creating the input with WPS and the same geo fields I do not get these features. Just wandering if this is all somehow connected to the use of ERA5.
Thank you for any advice you could provide.

 

[Ming Chen]

We are aware of the issues when using ERA5 to drive WRF.

NCAR RDA recently start to provide NetCDF-4 structured version of ERA5 data, which can be downloaded from:

NSF NCAR Curated ECMWF Reanalysis 5 (ERA5) - Registry of Open Data on AWS

registry.opendata.aws

A python package is provided to process the above ERA5 data and produce intermediate format files for running WRF. Please see

GitHub - NCAR/era5_to_int: A simple Python script for converting ERA5 model-level netCDF files to the WPS intermediate format

Can you use the python script to process ERA5 data? I believe it should work fine for you.

 

[tiziana]

Thank you!

I used the era5_to_int.py script to produce intermediate data before running metgrid. It is during metgrid that these issues arise. I followed this post: How to Make SST Data Match the Coastline in WRF Simulation?.
I simply ran the script as "python3 era5_to_int.py date1 date2 3" and the script did the magic of creating the metafiles for metgrid. This is on Casper. Is not the script fishing the archive you mention? Sorry for my confusion.