grids for defining OCEAN(ng)%ubar

[chenchai]

Hi,

In esmf_roms.F, we have the following.

CASE ('Ubar') ! 2D u-momentum
found=.TRUE.
DO j=JstrR,JendR
DO i=Istr,IendR
romsExpData(ifield,ng)%field(i,j)= &
& OCEAN(ng)%ubar(i,j,Kexp(ng))*scale
Fmin=MIN(Fmin,romsExpData(ifield,ng)%field(i,j))
Fmax=MAX(Fmax,romsExpData(ifield,ng)%field(i,j))
END DO
END DO

Does this mean that grids for OCEAN(ng)%ubar are at the 'u' points, since the 'i' loop starts from Istr? I would think it should be at the 'rho' points, since all romsExpData are at the 'rho' points.

Thanks,

James

[noah.eckman]

Yes, indeed, ubar is defined at u-points, vbar at v-points, just as the 3d momentum are at u and v points. This is verified by the script c_initial.m, which defines the dimensions for each of those.