CMAQ-ADJ (v5.0.2) compatibility with WRF Hybrid Vertical Coordinate without rerunning WRF

CMAQ
1

Hello CMAQ community,

I am currently working with CMAQ-Adjoint (CMAQ-ADJ v5.0.2) for inverse modeling and sensitivity analysis. However, I am encountering a vertical coordinate compatibility issue when processing meteorology generated from WRF.

My WRF meteorology was generated using the hybrid sigma-pressure vertical coordinate (HYBRID_OPT = 2). When running CMAQ-ADJ, I receive the following runtime error:

*** ERROR ABORT in subroutine SYNCSTEP
     Wrong vertical coordinate type

I understand that older CMAQ/MCIP versions may not support hybrid vertical coordinates, and that this support was introduced in later CMAQ releases.

Unfortunately, rerunning WRF without hybrid vertical coordinates is not feasible for this project. Also, CMAQ-ADJ currently requires the older CMAQ framework, so upgrading to newer CMAQ versions is not straightforward.

Is there any recommended or validated approach to:

  1. Convert or remap hybrid WRF output to a vertical coordinate system compatible with CMAQ-ADJ v5.0.2?

  2. Modify MCIP or preprocessing steps to allow hybrid coordinate handling in this older CMAQ version?

  3. Generate ICON/BCON or meteorology files in a way that maintains consistency with CMAQ-ADJ while using hybrid WRF outputs?

I am particularly interested in knowing whether any users or developers have successfully implemented a workaround without rerunning WRF.

Thanks in advance,
Sem

1 Like
2

In theory, CMAQ-ADJ should work with the hybrid generalized coordinate system output. However, recent versions of WRF used more land use categories (11-12), and the code in recent CMAQ versions after 5.0.2 were also modified to use those additional land use categories.

One option would be to use an older version of WRF to generate inputs that would be compatible with CMAQ-ADJ v5.0.2. This would assure that the land use categories between WRF and CMAQ-ADJ are consistent, so that dust emissions would be generated correctly within CMAQ-ADJ.

Shinliu @zhaoshunliu do you have any suggestions or recommendations to help resolve this issue.

Another option is to obtain a more recent version of CMAQ-ADJ that could work with the WRF data that you have, if such a version exists.

3

Hi,

The adjoint should work with the mentioned hybrid vertical coordinates (VGTYP = 2). Probably the VGTYP info is missing from the headers of the meteorology files?

Thanks,
Shunliu

4

No.

VGTYP = 2 is MM5v3-style non-hydrostatic sigma-P. See ioapi/PARMS3.EXT which defilnes the meaning of values for VGTYP (among other things)…

Carlie J. Coats, Jr., Ph.D.
I/O API Author/Maintainer

5

Thanks for the reply. I wonder if upgrading the mixing ratio calculation in vertical diffusion is sufficient for CMAQ-ADJ to use the WRF hybrid vertical coordinate.