What happened?
There are several gravity wave variables that are written to the history tapes with an ncol dimension, as can be seen here:
https://github.com/ESCOMP/CAM/blob/cam_development/src/physics/cam/gw_drag_cam.F90#L1568
However, these variables are actually dimensioned pcols, and thus using ncol results in non-physical grid imprinting on the values as they are written to the file.
What are the steps to reproduce the bug?
Do a run with the latest CAM development version with certain gravity wave variables, for example 'TAU_MOVMTN_A as reported by @JulioTBacmeister, and examine the values. You will likely see unphysical grid imprinting.
What CAM tag were you using?
cam6_4_184
What machine were you running CAM on?
CISL machine (e.g. derecho)
What compiler were you using?
Intel
Path to a case directory, if applicable
No response
Will you be addressing this bug yourself?
Any CAM SE can do this
Extra info
Credit should go to @JulioTBacmeister for both finding the bug and the method for fixing it.
What happened?
There are several gravity wave variables that are written to the history tapes with an
ncoldimension, as can be seen here:https://github.com/ESCOMP/CAM/blob/cam_development/src/physics/cam/gw_drag_cam.F90#L1568
However, these variables are actually dimensioned
pcols, and thus usingncolresults in non-physical grid imprinting on the values as they are written to the file.What are the steps to reproduce the bug?
Do a run with the latest CAM development version with certain gravity wave variables, for example
'TAU_MOVMTN_Aas reported by @JulioTBacmeister, and examine the values. You will likely see unphysical grid imprinting.What CAM tag were you using?
cam6_4_184
What machine were you running CAM on?
CISL machine (e.g. derecho)
What compiler were you using?
Intel
Path to a case directory, if applicable
No response
Will you be addressing this bug yourself?
Any CAM SE can do this
Extra info
Credit should go to @JulioTBacmeister for both finding the bug and the method for fixing it.