Confusion with d_nudgcoef.m script

[lcbernardo]

Dear ROMS users,

Lately I've been looking into creating a file for spatially varying nudging coefficients, to see if this approach can help improve my model results. By looking through some older posts, it seems that modifying the d_nudgcoef.m script is one way to do this, but then I've become confused on how to properly apply the boundary forcing. In my understanding, if a nudging coefficients file is not provided, the values specified in TNUDG, ZNUDG, M2NUDG, and M3NUDG will be used. However, if such a file is provided, the coefficients in the file will be used for the climatology provided in the file specified by CLMNAME. However, I don't usually activate this feature, and my boundary conditions for tracers, momentum, and zeta are all included in a boundary condition file specified in BRYNAME. Does this mean that if I want to apply spatially varying nudging coefficients, I have to provide the boundary conditions as a climatology file instead (specified by CLMNAME)? I would greatly appreciate any help or advice.

Thanks,
Lawrence

[wilkin]

Does this mean that if I want to apply spatially varying nudging coefficients, I have to provide the boundary conditions as a climatology file instead (specified by CLMNAME)?

Yes, if you want to use 3-D nudging to help constrain the solution near to the boundary then you must have a 3-D file of values (CLMNAME) to nudge toward.

The values strictly on the boundary (BRYNAME) are still used in the RadNud boundary condition option. The only change is that ROMS then takes the boundary nudging time scales from the 3-D nudging coefficient values instead of TNUDG etc. in roms.in.

[lcbernardo]

Thank you so much for the clarification Dr. Wilkin! I think that I can indeed use this feature to try what I have in mind.