MCIP issue about FRC_URB2D

Hi all,
I would like to report an issue that I met when doing the MCIP process. The error message is show below:


*** SUBROUTINE: RDWRFEM
*** ERROR RETRIEVING VARIABLE FROM WRF FILE
*** VARIABLE = FRC_URB2D
*** RCODE = NetCDF: Variable not found


I checked the variable “FRC_URB2D” and only find it in the geo_em.d0x.nc files rather than in the wrfout file. Thus, it looks like mcip is trying to extract this variable from wrfout files whereas this variable is not in it. Am I wrong or the mcip should be updated?
One more question is that whether this warning will greatly affect air quality simulation in CMAQ?
Please note that I am using WRFv3.9.1 with WPSv4.1 and CMAQv5.3.2 with up-to-date MCIP version.
Any comments are highly appreciated!

Regards,
Ryan
@tlspero

Hi,
Do you have IfGeo set to “F” in your run script?
I believe MCIP should run without it, if no geo_em file is provided.

Megan

Hi Megan,
I set the IfGeo to “T” in my run script and I can make sure that the directory for geo_em files is correct.

Could you please find a free moment to look at this issue?
@mmallard
@tlspero

Hi,
Sorry, I thought that you had been able to successfully run with IfGeo=T and a path set to the appropriate geo_em file. Alternatively, I think you can set IfGeo=F and MCIP will run without FRC_URB2D. Are either of those solutions working?

Megan

Hi Megan,
Yes, I can run the MCIP while there is still a warning message
“*** SUBROUTINE: RDWRFEM
*** ERROR RETRIEVING VARIABLE FROM WRF FILE
*** VARIABLE = FRC_URB2D
*** RCODE = NetCDF: Variable not found”
May I ask whether the FRC_URB2D variable will affect the CMAQ simulation results? Shall we just ignore and pass this warning?
Please note that I use the NLCD40 land use in WRF simulation and my domain covers California.
Thanks in advance!

I believe you can ignore the warning and proceed (assuming that’s the only error message and that MCIP successfully creates all needed input files). As for your question on the effect of using the FRC_URB2D field, I haven’t done any sensitivity tests to assess that and it would probably depend on the several factors in your model configuration. This part of the question might be best addressed under the “CMAQ” topic area in these forums.
Megan

I believe the FRC_URB2D field from WRF is used by MCIP to compute PURB (percent urban), and then that is used by CMAQ depending on the value of the KZMIN environment variable. If PURB is not available, then I think you will need to set KZMIN to N.

Since FRC_URB2D is a static field, if it is not available in your WRF output but you want to generate it, you should not need to rerun the entire simulation. You just need enough output to be able to run MCIP on to generate the PURB in your GRIDCRO2D file.