[MITgcm-support] coarser cube-sphere grid
Martin.Losch at awi.de
Fri Nov 15 05:32:41 EST 2013
thanks for your PDF document!
I have one comment on my quest to purge the earth from one particular terminology (o:
There is no such thing as a "tri-polar grid" (last line of the first page), even if "le grand modèle d'Europe" claims that.
Or as my more mathematically inclinded colleage puts it: "a sphere has a topological charge of minus two" (or "you cannot comb your hair with just one singularity on you head").
Nemo's ORCA grid still only has two poles, one of which is stretched into a line. Informed people shouldn't call this a "tri-polar grid".
See Murray JOURNAL OF COMPUTATIONAL PHYSICS 126, 251–273 (1996) for details
PS. This is an only semi-serious contribution to this list.
On Nov 15, 2013, at 2:06 AM, "Menemenlis, Dimitris (3248)" <Dimitris.Menemenlis at jpl.nasa.gov> wrote:
> Thanks Gael!
> One question I have is how would someone who is not at MIT
> access the directories you mention in gridsdispo.pdf
> Would it make sense to make some of these files available
> on a publicly-accessible server?
> Annegret, I can vouch for the quality of Gael's grids.
> I have been using the llc1080, llc2160, and llc4320
> and accompanying bathymetries, and they all
> work very beautifully.
> For anyone on support list who wants to give these
> higher-res configs a try, set-up details are here:
> Dimitris Menemenlis
> On Nov 14, 2013, at 2:39 PM, Gael Forget wrote:
>> Hi Dimitris and Annegret,
>> I added the pdf to the cvs repository @ MITgcm_contrib/gael/gridsdispo.pdf
>> In case a 1 degree resolution lat-lon-cap would suit your needs, the easiest
>> would be to get the one I used in ECCO version 4. To that end you
>> can use the script @ MITgcm_contrib/gael/setup_these_exps.csh
>> There is a README in that directory as well.
>> If otherwise, you will find cube spheres and lat-lon-cap grids from very
>> coarse to very high resolution available, In the directory mentioned in
>> gridsdispo.pdf Bear in mind that only few of the available grids have
>> actually been run into MITgcm. They come with no guaranty although, to
>> this day, we have not detected any issue with the ones that have been used.
>> On Nov 14, 2013, at 10:47 AM, Menemenlis, Dimitris (3248) wrote:
>>> Annegret, sorry for breaking link. I moved file to
>>> Gael, is there a more permanent version of this document under CVS somewhere?
>>> Dimitris Menemenlis
>>> On Nov 14, 2013, at 6:47 AM, Annegret Krandick wrote:
>>>> Hi Dimitris,
>>>> I am going to use the MITgcm to investigate the sensitivity of open ocean convection in the North Atlantic to ice sheet melting. We therefore decided to use a Lat/Lon/Cap configuration or a part of it. I want to make myself familiar with this kind of grid and I want to ask if you could give me some information or a description of the LLC grids you mentioned in the MITgcm-support list two years ago (http://mitgcm.org/pipermail/mitgcm-support/2011-December/007557.html). Unfortunately, the link to this description of the LLC grids doesn't work anymore.
>>>> Thanks a lot and with kind regards
>>>> Annegret Krandick
>>> On Dec 1, 2011, at 8:32 AM, Menemenlis, Dimitris (3248) wrote:
>>>> Alan, I have not generated a global CS grid at ~40-km spacing.
>>>> I cc this answer to support list in case someone else has a ~40-km grid readily available.
>>>> If not your options are the following:
>>>> 1. Generate a cube-sphere grid from scratch using scripts in
>>>> 2. Adopt the new lat/lon/cap (LLC) grid for which grids are readily available at
>>>> 4, 2, 1, 1/3, 1/6, 1/12, 1/24, and 1/48 degree grids.
>>>> I recommend you go with option 2 as this is where MITgcm/ECCO effort as a whole
>>>> is heading. A description of the LLC grids and how to obtain them is here:
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
More information about the MITgcm-support