Adding new variables to the Registry

[wjxaly]

I made modifications to the cumulus parameterization scheme and added new variables. I added the new variables in Registry.EM_COMMON, and there was no problem during compilation, and the parallel run works fine. However, I encountered a problem when performing 2-way nesting, where the model fails to continue running.


On the other hand, the model runs smoothly with one-way nesting (feedback=0). So, I would like to know what went wrong? Where do I need to make changes so that I can use my modified parameterization in the 2-way nesting setup?

 

[Ming Chen]

Would you please clarify how you add those new variables?

 

[wjxaly]

These are the new variables added in Registry.EM_COMMON:


This is the constant "talpha" defined in registry.dimspec:


These are the lines for defining these variables in the subroutine in /phys/module_cu_ntiedtke.F/.



I can run the model successfully and obtain the desired results in both no nesting and 1-way nesting setups. However, it fails to run in the 2-way nesting configuration.

 

[Ming Chen]

It seems fine to me that you correctly set 'rhdu' for these new variables. I don't have an immediate answer why it doesn't work for nesting cases.
I will talk to our software engineer and hope he know for sure what is wrong here. It may take some time. Thanks for your patience.

 

[wjxaly]

Thank you so much for your help. I uploaded the file of Registr.EM_COMM for your reference.

 

[islas]

@wjxaly Could you also upload your namelist.input? Additionally, if possible, could you try to run two way nesting with an analogous namelist.input without your registry modifications? Given where the error is coming from, I don't think your registry modifications are the cause - but I am not 100% confident yet so I'd like to try reproducing it on my side.