[MITgcm-support] [EXTERNAL] Treatment of Antarctic grounding zone regions

[Pochini, Enrico]

Hi Yoshihiro and Dimitris,

Thanks for your suggestions and for sending your papers.

In fact I experienced some model crashes, but I thought, since the
velocities close to the grounding line are quite small (at least for the
Ross Ice Shelf), the problem was somewhere else, perhaps close to the shelf
break.
I had to use a timestep of 10 minutes down to 5 minutes.

I will follow your suggestion, I will set a minimum water-thickness value
and after that process the draft and bathymetry; the shallowest GZ region
will be cut out.

Cheers,

Enrico


Il giorno ven 20 nov 2020 alle ore 07:53 Yoshihiro Nakayama <
Yoshihiro.Nakayama at lowtem.hokudai.ac.jp> ha scritto:

> Hi Dimitris and Enrico,
>
> It depends on your aim for your study. If your goal is to represent
> sub-ice shelf cavity circulation as realistic as possible, careful
> attention has to be paid to this.
>
> For my simulations I work in the Ross Sea at 5km resolution and vertical
> layers of a few tens of metres at the depth of the grounding zone. I am
> wondering whether this region should be cut out directly as non-resolvable
> in my vertical discretization, or whether, in a less conservative approach,
> the bathymetry should be deepened in order to resolve the region in at
> least one layer.
>
>
> For this point, I think both approaches are fine considering the
> uncertainty of both depth and draft data.
>
> For example, for my high-resolution study (Nakayama et al., 2019), I did
> not adjust your bathymetry and ice shelf draft much. I simply set the
> minimum water column thickness to be 10 or 20 m as my horizontal and
> vertical grid spacings are high enough.
>
> For a lower resolution study ( Nakayama et al., 2018), minimum water
> column thickness was set to 40-50 m (if I remember correctly), which means
> that ice shelf draft and ocean bathymetry are both adjusted. Too steep
> slopes (both in bathymetry and cavity geometry) will crash your model
> simulation, so you also have to smooth your bathymetry and ice shelf shape
> depending on your horizontal and vertical grid spacings.
>
> Nakayama et al., 2019 : https://www.nature.com/articles/s41598-019-53190-6
> Nakayama et al., 2018 : https://www.nature.com/articles/s41467-018-05813-1
>
> Yoshihiro Nakayama
>
>
> On Nov 18, 2020, at 23:55, Dimitris Menemenlis <menemenlis at jpl.nasa.gov>
> wrote:
>
> Michael and Yoshi, do you have any advice to give for question below?
> What decision did you make for your hi-res simulations, e.g., for:
> https://www.nature.com/articles/s41598-019-53190-6
>
> Thanks, Dimitris
>
> On Nov 18, 2020, at 6:47 AM, Pochini, Enrico <epochini at inogs.it> wrote:
>
> Dear community,
>
> I have a question for the polar oceanographers:
>
> I have noticed that in bathymetric reconstructions available for
> Antarctica (e.g. Bedmap2, RTopo-2) there are regions corresponding to the
> grounding zone of the biggest ice shelves where the water column thickness
> (resulting from the subtraction draft - bathy) equates exactly 1m.
>
> For my simulations I work in the Ross Sea at 5km resolution and vertical
> layers of a few tens of metres at the depth of the grounding zone. I am
> wondering whether this region should be cut out directly as non-resolvable
> in my vertical discretization, or whether, in a less conservative approach,
> the bathymetry should be deepened in order to resolve the region in at
> least one layer.
>
> Employing a preprocessing algorithm (
> https://github.com/knaughten/mitgcm_python
> <https://urldefense.us/v3/__https://github.com/knaughten/mitgcm_python__;!!PvBDto6Hs4WbVuu7!bWCWtDMcJoVTIrR5IAjyeN1eGS1eXJVRuH3eFS2l_JFJ3IhWUV5BBvLC7nnfv5WjG2XalrUIZ7A$>)
> I obtain a grounding-zone water-thickness of ~ 50-80 m, enough to fit ~1-2
> layers.
>
> I am uncertain about which way to take, so I'm curious to hear how you
> treat such thin cavities in high resolution models in other Antarctic seas,
> whether you cut them or keep them, and how you make them resolvable without
> affecting too much the overall setup.
>
> Thanks!
>
> Enrico P.
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
>
> https://urldefense.us/v3/__http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support__;!!PvBDto6Hs4WbVuu7!bWCWtDMcJoVTIrR5IAjyeN1eGS1eXJVRuH3eFS2l_JFJ3IhWUV5BBvLC7nnfv5WjG2XakE4dloY$
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20201120/ab7a1b4a/attachment.html>