[Mitgcm-support] Lopped cells

[mitgcm-support at dev.mitgcm.org]

Arne,

Atlantic
========

I'm not quite sure what to do about the lopped cells. I've been going through
various experiments that use lopped cells (including some really turbulent
ones) and it all works fine. The one common effect lopped cells have is to
reduce the stability which is to be expected because we effectively reduce
the grid size. Reducing the time step further for the 1/6th runs is not an
attractive prospect!

Unfortunately, I can't actually run the Atlantic case here. Chris has been
able to set it up because the one machine big enough (our PC cluster) has
been broken for two months and the two people in the world capable of
fixing it (at LCS) can't be persuaded to come over.

Would you mind telling me what the last attempts involved (regards lopped
cells). There were these issues about initialization (switching lopped cells
on suddenly etc.) and I forget exactly what you had and hadn't tried.

Scripts
=======

I like the html document you have up about the data processing scripts.
Are the versions in "internal/mitgcmuv/scripts/" up to date (ie. the
latest versions you use)? I'd like you to put them under CVS on server
(I'll give you access) so that we can start maintaining versions.
hfac.m needs to be brought up to date with the model, for instance.

Caribbean
=========

I have a question from Terry Joyce to you about the deep circulation in
the Caribbean. He pointed to a paper by Holloway where he uses his
Neptune parameterization and with it he gets a deep reaching cyclonic
rim current and not without. Could you make a current map at about
1600m of the Caribbean to show what we get? I suspect we won't
have anything but who knows.

Alistair.