different simulation results with initial condition from ERA-interim pressure level data and model level data?

[Emily-wrf-support]

Hi,
I am now running the wrf model with version 3.9.1 and I meet a problem when using ERA-interim data as the initial condition to run a real case. The results are different when the ERA-interim pressure level or the model level data are used as the initial condition. The ERA-pressure level (38 vertical levels) and model level (61 vertical levels) data are downloaded from ECWMF website (https://apps.ecmwf.int/datasets/data/interim-full-daily/levtype=sfc/).

The model level data are used the same as the pressure level data in the WPS, except that:

1. link the model level Vtable.

2. run the calc_ecmwf_p.exe with ecmwf_coeffs of 61 vertical levels before running the metgrid.exe.

All goes well but it seems that when plotting the Geopotential height field, they are different from the wrfinput file with pressure level or model level data as initial condition. And the simulation results are different as well.

Did I do something wrong in WPS?

Thank you for your help in advance.

 

[Ming Chen]

Your procedures to process the model level ERA data are correct.
I wonder whether there are large differences between the simulations of WRF driven by the two datasets?

 

[Emily-wrf-support]

Your procedures to process the model level ERA data are correct.
I wonder whether there are large differences between the simulations of WRF driven by the two datasets?

Yes. The simulated geopotential height fields are different even at the first wrfout time. The 500 hPa trough gose faster when using ERA-interim model level data as the initial condition. As a result, the MCS (simulated maximum radar reflectivity) moves faster as well.
Actually, the geopotential height fields are different in the wrfinput files. I wonder, whether the differences come from calc_ecmwf_p.exe when it calculates height from model level data?