Add a new parameter to control cutoff level for implicit RF

[XiaqiongZhou-NOAA]

Description

A new parameter rf_cutoff_fast is introduced to independently control the cutoff level for implicit Rayleigh damping when fast_tau_w_sec>0, decoupling it from the cutoff level (rf_cutoff) used for explicit damping.

How Has This Been Tested?

It is tested on HERA with intel compiler

Checklist:

Please check all whether they apply or not

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• Any dependent changes have been merged and published in downstream modules

[lharris4]

@XiaqiongZhou-NOAA this is a nice addition, thank you.

[yangfanglin]

@XiaqiongZhou-NOAA Kate, can you add reviewers and get this PR ready for commit as early as possible ? For now we can set the default rf_cutoff_fast to be same as the rf_cutoff.

[bensonr]

While technically this all looks good, I am concerned with the choice of rf_cutoff_fast for the variable name. There is a Ray_fast routine for simple calculation and your new variable name seems to be independent of that logic, but could be mistaken as being related to it.

[lharris4]

@bensonr To be consistent with fast_tau_w_sec, perhaps the best name would be rf_cutoff_w. I admit that there is a namespace clash with RF_fast() although this has largely been deprecated in favor of this w damping.

Also @yangfanglin could we understand some more of the motivation for this change and what positive effects it may have?

[XiaqiongZhou-NOAA]

@bensonr @lucas: Thanks. I will change the parameter name.
Lucas: When we tried to limit all RF above 1hPa, it turns out that the polar vortex has a strong bias. This modification will allow us keep implicit w damping at 1hPa but with lower rf_cutoff to reduce that bias.

[XiaqiongZhou-NOAA]

@lharris4
I would like to add the modify in this PR by changing "trange_bad = (/ 150., 350. /) ! temperature" to (120 to 350) in fv_diagnostics.F90. As our model is up to 80km and C1152 resolution, 150K seems not very warning. Any suggestions?

[yangfanglin]

NASA's TIMED mission and satellite limb sounders have modeled and observed mesopause temperatures as low as 123 K (-150°C). Experimental GFSv17 (C1152) prints a lot of low temperature warnings that slows down model performance and even at times causes model to exceed assigned CPU hours.

[lharris4]

@XiaqiongZhou-NOAA @yangfanglin the low temperature warnings are getting to be annoying in a number of contexts. I would recommend setting trange_bad = (/50., 350.).

A pressure- or otherwise context-dependent temperature check would be ideal, but not a priority at this time (perhaps something better suited for PyFV3).

[FernandoAndrade-NOAA]

Testing on #2730 is complete, please continue with the merge process.

[laurenchilutti]

@FernandoAndrade-NOAA Merge is complete.