Replacing SPLINES in updated ROMS – Use of RI_SPLINES vs. KANTHA_CLAYSON/N2S2_HORAVG?

[leilane.passos]

Hi everyone,

I’m trying to compile an old test case (OldTest.h) using the latest version of ROMS. This header originally used the deprecated SPLINES option.

After checking the ROMS forum, I saw that SPLINES has been replaced by:
SPLINES_VDIFF
SPLINES_VVISC
RI_SPLINES (for GLS, LMD, or MY25 mixing schemes)
Reference: Ticket #681 (https://www.myroms.org/projects/src/ticket/681)

Following the updated upwelling.h example, I added SPLINES_VDIFF and SPLINES_VVISC options, and also included the following block in OldTest.h:

#if defined GLS_MIXING || defined MY25_MIXING
# define KANTHA_CLAYSON
# define N2S2_HORAVG
# define RI_SPLINES
#else
# define ANA_VMIX
#endif

Since OldTest.h uses MY25_MIXING, this block activates RI_SPLINES, but also KANTHA_CLAYSON and N2S2_HORAVG.

My question is:
To replicate the old SPLINES behavior as closely as possible, is it enough to use only RI_SPLINES?
Should I also keep KANTHA_CLAYSON and N2S2_HORAVG, or could they introduce changes not intended in the original configuration?
I couldn’t find specific guidance about these two options in Ticket #681, so any clarification or recommendations would be greatly appreciated.

Best,
Leilane

[jcwarner]

RI_SPLINES works in the turbulence closures.
The SPLINES_VDIFF and SPLINES_VVISC work in the tracers and momentum.
So if you want an exact match, you need to use the same (renamed) options that you used previously.

[leilane.passos]

Ticket #681 (https://www.myroms.org/projects/src/ticket/681) provides a good explanation of SPLINES_VDIFF, SPLINES_VVISC, and RI_SPLINES. However, I’m looking for more clarification on the roles of KANTHA_CLAYSON and N2S2_HORAVG. Could anyone explain their purpose and why they are used in the Upwelling Test case?

[jcwarner]

N2S2 is a horiz avg of the density fields. recommended for the GLS and MY25 closure options. i am not sure what the upwelling uses.
The Kantha/Clayson closure descriptions are in our GLS ocean modeling paper:
https://www.sciencedirect.com/science/a ... 0303000702
Ocean Modelling
Volume 8, Issues 1–2, 2005, Pages 81-113
Ocean Modelling
Performance of four turbulence closure models implemented using a generic length scale method
-j