Using MUR SST Instead of ERA5 SST in WPS/WRF: Questions on Metgrid Priority and SST Update

[Prolay Saha]

Dear WRF Support Team,

I am using ERA5 data to force WRF. For WPS preprocessing, I run ungrib.exe separately for pressure-level and surface-level ERA5 files:

For pressure-level data:


&ungrib
prefix = 'PLV',
/


For surface-level data:


&ungrib
prefix = 'SFC',
/


Then I run metgrid.exe with:


&metgrid
fg_name = 'PLV','SFC',
/


The ERA5 surface files already contain SST, so the resulting met_em* files include ERA5 SST information.

I would like to replace the ERA5 SST with SST from the Multi-scale Ultra-high Resolution (MUR) SST product. My understanding is that I could convert the MUR SST data into WPS intermediate format and then run metgrid.exe using:


&metgrid
fg_name = 'PLV','SFC','SST',
/


where SST refers to the MUR SST dataset.

I would appreciate clarification on the following points:

1. If both the ERA5 surface files (SFC) and the external MUR SST dataset (SST) provide SST information, will metgrid.exe automatically use the MUR SST and override the ERA5 SST, or do I need to remove the SST field from the ERA5 surface files before running ungrib.exe/metgrid.exe?

2. If the met_em* files contain time-varying SST fields (e.g., from MUR or ERA5), but I run WRF with:


&physics
sst_update = 0,
/


is it correct that real.exe will use only the initial SST field and that WRF will keep this SST fixed throughout the simulation?

3. If the met_em* files do not contain any SST field at all, and I run with:


&physics
sst_update = 0,
/


how does real.exe/wrf.exe handle the ocean surface temperature? Will the model fail during initialization, or will it derive SST from another field?

Thank you very much for your time and assistance.

 

[kwerner]

I'll answer your questions below:

1. If both the ERA5 surface files (SFC) and the external MUR SST dataset (SST) provide SST information, will metgrid.exe automatically use the MUR SST and override the ERA5 SST, or do I need to remove the SST field from the ERA5 surface files before running ungrib.exe/metgrid.exe?

Metgrid will use the data in the order you provide in the fg_name line. Since you have SST listed last, it should use those data (MUR SST) to overwrite the SST found in the ERA5 files. You do not need to remove it from the ERA5 input.

2. If the met_em* files contain time-varying SST fields (e.g., from MUR or ERA5), but I run WRF with:

&physics
sst_update = 0,

is it correct that real.exe will use only the initial SST field and that WRF will keep this SST fixed throughout the simulation?

That is correct. If you have sst_update turned off, WRF will not update SST during its simulation. I've not tested this, but it shouldn't. If you find that it behaves otherwise, please let me know.

3. If the met_em* files do not contain any SST field at all, and I run with:

&physics
sst_update = 0,

how does real.exe/wrf.exe handle the ocean surface temperature? Will the model fail during initialization, or will it derive SST from another field?

If there is not SST data in the input files, WRF will use skin temperature (TSK) for the water points, instead of SST.

 

[Prolay Saha]

I'll answer your questions below:


Metgrid will use the data in the order you provide in the fg_name line. Since you have SST listed last, it should use those data (MUR SST) to overwrite the SST found in the ERA5 files. You do not need to remove it from the ERA5 input.


That is correct. If you have sst_update turned off, WRF will not update SST during its simulation. I've not tested this, but it shouldn't. If you find that it behaves otherwise, please let me know.


If there is not SST data in the input files, WRF will use skin temperature (TSK) for the water points, instead of SST.

Dear Kwerner,
Thank you for the prompt and detail explaination.