While running the frierson_test_case.py experiment on ISCA, Derecho HPC, I encountered a fatal error when specifying the convection_scheme parameter in the idealized_moist_phys_nml namelist. The ISCA documentation (https://execlim.github.io/Isca/modules/idealised_moist_phys.html) lists valid values as SIMPLE_BETTS_CONV, FULL_BETTS_MILLER_CONV, RAS_CONV, DRY_CONV, and NO_CONV. However, when I set convection_scheme='NO_CONV', the model crashes with the following error:
FATAL from PE 6: idealized_moist_phys: "NO_CONV" is not a valid convection scheme. Choices are NONE, SIMPLE_BETTS, FULL_BETTS_MILLER, RAS, DRY.
The model runs correctly when I use the runtime-specified values, such as 'NONE' instead of 'NO_CONV'. It seems the accepted namelist values differ from what is documented. Could the documentation be updated to reflect the actual valid inputs? This would help avoid confusion and run-time errors for users configuring convection schemes. Thanks!
While running the frierson_test_case.py experiment on ISCA, Derecho HPC, I encountered a fatal error when specifying the convection_scheme parameter in the idealized_moist_phys_nml namelist. The ISCA documentation (https://execlim.github.io/Isca/modules/idealised_moist_phys.html) lists valid values as SIMPLE_BETTS_CONV, FULL_BETTS_MILLER_CONV, RAS_CONV, DRY_CONV, and NO_CONV. However, when I set convection_scheme='NO_CONV', the model crashes with the following error:
FATAL from PE 6: idealized_moist_phys: "NO_CONV" is not a valid convection scheme. Choices are NONE, SIMPLE_BETTS, FULL_BETTS_MILLER, RAS, DRY.
The model runs correctly when I use the runtime-specified values, such as 'NONE' instead of 'NO_CONV'. It seems the accepted namelist values differ from what is documented. Could the documentation be updated to reflect the actual valid inputs? This would help avoid confusion and run-time errors for users configuring convection schemes. Thanks!