Setting up multiple domains

[Kate Lai]

Is there guideline or procedure detailing how to set up multiple domains in HYCOM? With the small inner domain of a higher resolution than outer domain, And inner domain deriving its boundary condition from outer domain, etc.?

Thanks for help.

[Alan Wallcraft]

HYCOM implements off-line one-way nesting.  

So, first you run the outer model, then you extract the nesting archive files (coarse to fine interpolation), and finally you run the nested domain.

Our example is GOMb0.08 nested inside GLBb0.08.  This is same resolution nesting, but all the procedures are the same for finer resolution nesting. I suggest 3x nesting, which you can setup so that all the u,v,p grids colocate, but any finer grid that is inside the outer grid will work. 

Alan.

[Kate Lai]

Btw, if setting the value of cb positive in namelist, does it mean the bottom friction will be constant everywhere in the domain and all time in the simulation? 

[Alan Wallcraft]

The coefficient of quadratic bottom friction is constant in space and time if it is positive.  Note that the actual friction depend of cb * near bottom speed.

If it is negative, cb.[ab] is read in and the coefficient varies in space but not in time.  This is normally calculated by hycom_botfric, see  cb_07_10mm.csh

C hycom_botfric - Usage: hycom_botfric depth.a idm jdm z0 cbmin cb.a

C
C                 Calculates roughness-based bottom friction coefficient
C
C  depth.a is assumed to contain idm*jdm 32-bit IEEE real values
C   for each array, in standard f77 element order, followed by padding
C   to a multiple of 4096 32-bit words, but otherwise with no control
C   bytes/words, and input values of 2.0**100 indicating a data void.
C
C  depth.a contains the bottom depth in m
C  z0      is the bottom roughness in m
C  cbmin   is the minimum quadratic bottom friction coeficient
C  cb.a    will contain the roughness-based bottom friction coefficient
C
C  Uses NCOM's formula for 2-D cases:
C    cb(i,j) = max( cbmin, (vonk/log(0.5*depth(i,j)/z0))**2 )
C
C  The same scheme, in 3-D, is discussed in Appendix 2 of
C   Oey, L.-Y., 2006:  An OGCM with movable land-sea boundaries.
C   Ocean Modeling, 2006, 13, 176-195.
C  It is using the "Law of the Wall", see:

C   Schlichting, H., 1968. Boundary Layer Theory. McGraw-Hill, 748 pp.

Alan.