nesting about fine2coarse

[Tsubono]

Hi,

I tried a test case using DEFINE nesting.

I think there might be bugs about fine2coarse.
Please let me know if these are bugs, if you know well about fine2coase.

I think that line 291-293 in main3d.F

Code: Select all

IF ((istep.eq.Nsteps).and.(RefineScale(ng).gt.0)) THEN
     CALL nesting (ng, iNLM, n2way)
END IF

must be written after "call step3d_t" around line 711 in main3d.F.

and
I think these line 2144,2166 and 2184 in nesting.F

Code: Select all

     &                           OCEAN(rg)%t(:,:,:,nstp(rg),itrc))
     &                          OCEAN(rg)%u(:,:,:,nstp(rg)))
     &                          OCEAN(rg)%v(:,:,:,nstp(rg)))

must be

Code: Select all

     &                           OCEAN(rg)%t(:,:,:,nnew(rg),itrc))
     &                          OCEAN(rg)%u(:,:,:,nnew(rg)))
     &                          OCEAN(rg)%v(:,:,:,nnew(rg)))

. Of course, I set the ratio of the two dt(ng)s odd number.

Or,
It might be good to only change line 229

Code: Select all

IF ((istep.eq.Nsteps).and.(RefineScale(ng).gt.0)) THEN 

to

Code: Select all

IF ((istep.eq.Nsteps).and.(RefineScale(ng).eq.0)) THEN

in main3d.F, if I will use two domains nesting model only.


I have one more tiny question about fine2coarse.

In dogbone_ngc_refined.nc,
the one longitudinal line of Idg(NstrU(2)...) are 5, 8, 11,.... , and one latitudinal line of Jdg(NstrV(2)) are 5, 8, 11,..... .
But, I think that of Idg(NstrU(2)) must be 4, 7, 10,....,. and Jdg(NstrV(2)) must be 4, 7, 10,...... , because the longitudinal u point and the latitudinal v point starts from 1 while the rho point starts from 0.

Sorry for wasting your modeling.

Thank you.