[MITgcm-support] Solution blows up

[Shevchenko, Igor]

Hi Matt, Jody, and Michael,

I would like to thank you for helping me through the instabilities of the initial configuration
and teaching me how to cope with them. I have taken all your advices onboard.
It is happened that MITgcm is rather sensitive to initial configurations, and filling some values 
with defaults (it was just a few in my case to pay that much attention to them) is really not a good idea even 
when using small time steps (I dropped my dt from 200s to 2s and it didn't help). 
It happened that my interpolation function did interpolate using just a few land points, which I filled with defaults.

Now, the normal functionality is restored! Thank you very much again!

Best regards,
Igor

________________________________________
From: MITgcm-support <mitgcm-support-bounces at mitgcm.org> on behalf of Matthew Mazloff <mmazloff at ucsd.edu>
Sent: 04 March 2020 17:18
To: mitgcm-support at mitgcm.org
Subject: Re: [MITgcm-support] Solution blows up

Its due to how you interpolate 1/12 to 1/6. You have to use caution.
Do the 1/6 and the 1/12 have the same number of z levels? If not you have to use caution here. LIkely interpolate in horizontal and then in depth. As you go down extrapolate deeper in a stratification stable way.

Some notes
- Obviously there can be no extrapolation from land, and I assume you did that correct.
- I would not fill with defaults at depth - this is possibly your problem - why do you even get blanks as I would assume the 1/12 would have at least 1 of its four grid cells in the 1/6 wet cell. Try bin averaging the 1/12 to the 1/6

-Matt




> On Mar 4, 2020, at 8:55 AM, Shevchenko, Igor <i.shevchenko at imperial.ac.uk> wrote:
>
> Hi Jody,
>
> Thank you very much for your answer.
> Attached is STDOUT.0000 . Does the CFL look suspicious?
>
> Best regards,
> Igor
>
> ________________________________________
> From: MITgcm-support <mitgcm-support-bounces at mitgcm.org> on behalf of Jody Klymak <jklymak at uvic.ca>
> Sent: 04 March 2020 16:44
> To: mitgcm-support at mitgcm.org
> Subject: Re: [MITgcm-support] Solution blows up
>
>
> *******************
> This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender.
> If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address.
> *******************
> Hi Igor,
>
> 1. you could have density overturns in the interpolated field, and those will manifest as high vertical velocities that violate the CFL criterion.
>
> 2. you likely have some fast adjustment that is violating CFL and may need to use a smaller timestep for a little while until the model adjusts.
>
> The fact you tried “different time steps” says that maybe the problem is more complicated, but it would be good to look at STDOUT.0000 and see what the cfl criterion is saying.
>
> Cheers,   Jody
>
> --
> Jody Klymak
> http://web.uvic.ca/~jklymak/
>
>
>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
> <STDOUT.0000>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support

_______________________________________________
MITgcm-support mailing list
MITgcm-support at mitgcm.org
http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support