Horizontal bands in SHF

[a_sengupta@unsw_edu_au]

Hi,
Ive come across some strange horizontal bands in the pop output files for SHF (both CCSM2 - b20.007 and CCSM3 - b30.004 and presumably others). If you take a long term average of a particular month its easy to see. The bands are horizontal in the Southern Hemisphere and become curved, following the pop grid in the Northern hemisphere. The spacing is approx 2.5 degrees, so I presume its related to interpolation from the atmosphere to ocean grid.
Ive put a sample plot on http://web.maths.unsw.edu.au/~alexg/b20.007_SHF_jan.png

The errors appear to be quite large and it looks like the pattern also appears in ocean fields e.g HMXL.

Has this been looked into?
cheers
Alex

 

[rneale]

These patterns have been seen before and it is a consequence of how the atmospheric fields are interpolated to the finer oceanic grid. Fluxes such as SHF tend to be interpolated maintaining the area average (for energy conservation) and scaler fields such as Temperature tend to be bi-linearly interpolated. What you are seeing is the ocean SHF calculated using a bi-linearly interpolated 2-m temperature from the atmosphere giving it the saw-tooth appearance. When this ocean SHF is interpolated back onto the atmospheric grid this pattern tends to disappear.

 

[a_sengupta@unsw_edu_au]

Hi Richard, thanks very much for your reply.
Isn't this quite concerning? If this banded SHF is forcing the ocean, does it not set up spurious gradients at the surface? The saw-tooth pattern in SHF seems to have a rather large amplitude (in some places 10W/m2/degree latitude). Having said that while I can see banding in the HMXL field I can't see any banding in the SST.
cheers
Alex