Documentation on Bottom Variables

[Sandra Slead]

Hello, 

Is there any formalized documentation available on how the _bottom variables are calculated/included in the output for the GOFS 3.1 Dataset (in particular, GLBy0.08 exp. 93.0)? 

I found this forum post, which is helpful for understanding the interpolation from GLBb0.08 to GLBy0.08, but I am wondering if there is additional documentation on what happens in the model between the last standard depth and the seafloor. 

Thank you very much,

Sandra

[Alan Wallcraft]

As the forum post you linked to says:

The GLBy0.08 41-layer archive file is SAMPLED at fixed depths (including 1500 m an d 10m above the GLBy0.08 bottom) using a specified profile type, the 41 layer velocities are first averaged from the cell edges to the cell center (where T & S is defined).

The bottom velocities are SAMPLED at 10 m above the bottom, using profile type PCHIP (piecewise cubic hermite interpolation between layer centers).

In shallow water (certainly shallower than 100n and probably shallower than 500 m) there will be multiple thin layers near the bottom and the velocity will represent 10 m above the bottom well.  We sample at bpttom+10m because our bottom boundary layer for quadratic bottom friction is 10 m.

In deep eater (certainly below 1500 m) HYCOM typically has a very thick bottom active layer and  bpttom+10m represents the average over that layer (e.g. over perhaps the bottom 1,000 m in some cases).

Alan.

[Sandra Slead]

Hi Alan,

Thank you. I am wondering specifically about the computational/native grid. The documentation I have been reading (HYCOM user’s manual and Bleck 2002 paper) suggests that vertical coordinates in the deep ocean are determined using isopycnals only, and that terrain-following coordinates are reserved for coastal regions. How are the vertical coordinates of the bottom boundary layer defined in the native grid? 

Thanks again for the help, 

Sandra 

[Alan Wallcraft]

Here is the long term mean vertical section from 1000 to 6000m at 180W (middle of the Pacific) from the GOFS 3.1 Reanalysis:

This illustrates that the lowest active layer is typically isopycnal for depths below 1000 m and it is usually 100's to 1000 m thick.

I don't have any near surface plots from the Reanalysis, but here is what the coordinate system looks like in the Gulf of America at 25N:

The coordinate is terrain following with 14 layers down to 84m, and in February it is still in fixed depth coordinates (each 8 m thick) down to about 140 m with isopycnals outcropping into the fixed depth layers.  The isopycnal layers are 20 to 50 m thick even this near the surface.

Alan.