why there is inconsistency between WRF Input and Output Files?

[mirabelle]

As we all know, the initial time of the wrfinput file and the wrfout file are exactly the same.

However, when I link the wrfout file as a wrfinput file for running the model, the resulting wrfout file is no longer consistent with the initial field.

Why does this happen?

 

[kwerner]

The initial wrfout* file cannot be substituted for an initial condition file (wrfinput*). If you take a look at the data in each, there are many differences. You can do this using the "ncdump -h file" command for each.

 

[mirabelle]

The initial wrfout* file cannot be substituted for an initial condition file (wrfinput*). If you take a look at the data in each, there are many differences. You can do this using the "ncdump -h file" command for each.

Thanks for your reply.
I want to use the results from wrfout_d01 to drive the WRF model and simultaneously utilize the automatic interpolation feature to obtain results for wrfout_d02 (input_from_file = .true.,.false).

Is there any method to achieve this goal?

 

[mirabelle]

I have two ideas, what do you think of them?

1. Using wrfout as the background field to generate wrfinput. The challenge is that there doesn't seem to be a vtable that can recognize wrfout files.
2. Generating a wrfrst file at the same time as the previous wrfout file and using wrfrst to replace all wrfinput variables. The problem is that some variables have values at both time1 and time2, and I'm not sure whether to use the values from time1 or time2.

 

[kwerner]

Hi,
Apologies for the delay in response. I believe what you want to do can be achieved using the UPP program, but there isn't an option with standard WRF/WPS.