Page 1 of 1

Numerical instabilities (spikes in the ice surface) at grounding line

Posted: 24 Mar 2021, 12:17
by chenry
I'm running a simulation with a transient grounding line and am observing numerical instabilities in the form of spikes on the upper ice surface and inverted spikes on the lower ice surface. Has anyone else experienced this and has a solution to fix these numerical instabilities?

Re: Numerical instabilities (spikes in the ice surface) at grounding line

Posted: 24 Mar 2021, 12:23
by gagliar
Helo,

Can you be more specific? Are you running a Stokes or a SSA simulation? 2d or 3d? From which setup did you started?

Regards,
Olivier

Re: Numerical instabilities (spikes in the ice surface) at grounding line

Posted: 24 Mar 2021, 13:44
by chenry
Hi Olivier,

Thanks for your reply. I am running a full Stokes 3D simulation and the grounding line in question is in a collision zone. The basal friction coefficient is set to 7.624e7 and the basal melt rate is dependent on the ice thickness and distance from the GL (Favier et al., 2016). Our current thinking is that the numerical instabilities could be caused by the resulting high melt rates in the collision zone. I have attached a couple of photos for reference.

Cheers,

Clara
UpperView.jpg
(48.39 KiB) Not downloaded yet
LowerView.jpg
(102.25 KiB) Not downloaded yet

Re: Numerical instabilities (spikes in the ice surface) at grounding line

Posted: 27 Mar 2021, 13:32
by gagliar
Hi Clara,

Did you try to decrease your time step?

Also, having a water pressure dependent friction law might help by removing the discontinuity at the GL.

Hope it helps
Olivier

Re: Numerical instabilities (spikes in the ice surface) at grounding line

Posted: 19 Jul 2021, 11:04
by chenry
Thank you for your help, Olivier. I was able to fix the problem in the end by avoiding a BMB discontinuity at the grounding line.

All the best,

Clara