[MITgcm-support] weak AMOC in LLC90

[gael forget]

Hi Judith,

> The only paper I've found that shows published AMOC from the LLC90 set-up is the Danabasoglu et al
You will also find various MOC plots in the supplement of the paper that documents the 
reference LLC90 set-up and solution (i.e. the 20 year ECCO v4 setup and state estimate)
available @ http://www.geosci-model-dev.net/8/3071/2015/

My general advice would be to start from the ECCO v4 setup (directions to re-run the baseline 20 year solution 
are provided in http://mitgcm.org/viewvc/*checkout*/MITgcm/MITgcm_contrib/gael/verification/eccov4.pdf)
that generally shows reduced drifts in multi centennial simulations thanks to its optimized turbulent 
transport parameters (see http://www.ocean-sci.net/11/839/2015/os-11-839-2015.pdf). One may hope 
that this would lead to a more realistic ocean circulation also when you switch to the Drakkar forcing.

> I haven't yet plotted a time-series of the AMOC as in that paper with synchronous time-stepping, will do.
Good idea. The asynchronous time step approach reduces computational cost but 
it also distorts the model physics so it should only be considered with extra caution.

> is that something that we just have to accept when using truely global configurations
Well … it depends how much time you want to put in model tuning. As you noted AMOC biases 
(often low biases) are a rather common problem that goes beyond MITgcm and llc90.
Many model parameters and settings can influence spurious model drifts and AMOC 
is not necessarily the easiest thing to control — as illustrated by Gokhan’s paper.

> Has any one experience on what possibilities there are to get a stronger AMOC in the global LLC90 set-up? What works? What does not work?

I recall that when I did the CORE2 runs for Gokan’s papers the results were 
quite sensitive to the fresh water balance of the subpolar gyre, and the resulting 
mixed layer depths. Those multi-centennial runs included a surface salinity relaxation 
term to help maintain a somewhat reasonable overturn. You may want to play with that.

Hope this helps.

Cheers,
Gael

On Mar 21, 2016, at 7:35 AM, Judith Hauck <judith.hauck at awi.de> wrote:

> Dear all,
> 
> I have started to use the LLC90 set-up as provided by Gael (http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/gael/verification/global_oce_llc90/?hideattic=0&sortby=author) and noticed that the Atlantic Meridional Circulation (AMOC) is really weak (<10 Sv) after 100 years. 
> 
> That was noticed early on in this conversation on the developer list:
> 
> http://mitgcm.org/pipermail/mitgcm-devel/2007-September/002973.html
> 
> I have used both the CORE2 and the (climatological, daily) Drakkar forcing with asynchronous time-step, but the Drakkar forcing  doesn't make any difference for the AMOC.
> 
> The only paper I've found that shows published AMOC from the LLC90 set-up is the Danabasoglu et al.: North Atlantic simulations in Coordinated Ocean-ice Reference Experiments phase II (CORE-II). Part I: Mean states, Ocean Modelling, 2014,         doi:10.1016/j.ocemod.2013.10.005
> 
> In that paper, the MITgcm AMOC is also weak (but rather at 11 Sv after 100 years). Most models in that comparison have rather weak AMOC, so is that something that we just have to accept when using truely global configurations?
> 
> I haven't yet plotted a time-series of the AMOC as in that paper with synchronous time-stepping, will do.
> 
> Has any one experience on what possibilities there are to get a stronger AMOC in the global LLC90 set-up? What works? What does not work?
> 
> I'd appreciate any feedback.
> 
> Thanks,
> Judith
> 
> -- 
> Judith Hauck
> Alfred Wegener Institute for Polar and Marine Research
> Am Handelshafen 12
> 27570 Bremerhaven
> 
> +49 (0)471-4831-1892
> judith.hauck at awi.de
> 
> http://www.awi.de/People/show?jhauck
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20160321/6b0f42ad/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 1843 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20160321/6b0f42ad/attachment.p7s>