Opened 2 months ago

Closed 7 weeks ago

#2620 closed help (fixed)

vn10.9 AMIP run PI control

Reported by: admg26 Owned by: um_support
Priority: normal Component: UKCA
Keywords: UKCA, grid point storm Cc:
Platform: Monsoon2 UM Version: 10.9

Description

Hello,

Monsoon user: almin
Archive: /crum/
UM-UKCA vn10.9 Trunk
AMIP run using 1850 PI control CMIP6 emission files and start dumps.

Suite: u-ax344

Run crashes after 12 years with the following

????????????????????????????????????????????????????????????????????????????????
???!!!???!!!???!!!???!!!???!!!       ERROR        ???!!!???!!!???!!!???!!!???!!!
?  Error code: 100
?  Error from routine: set_thermodynamic
?  Error message: A total of     3 points had negative mass in set_thermodynamic. This indicates the pressure fields are inconsistent between different levels and the model is about to fail.
?  Error from processor: 158
?  Error number: 38
????????????????????????????????????????????????????????????????????????????????

I am confused as to why it ran for so long before falling over.

Cheers,
Alison

Change History (7)

comment:1 Changed 2 months ago by admg26

I am going to re-run the model with a copy of the suite. I have an idea of what might be wrong.

comment:2 Changed 2 months ago by willie

Hi Alison,

Is this working now - can we close this ticket?

Regards
Willie

comment:3 Changed 2 months ago by admg26

Hi,

A copy of the suite with a few tweaks is still running but since it took 12 years to crash the last time, I am inclined to wait a bit more (it has done 9 years so far) to see if it is actually working. Would it be okay to leave the ticket open for the week? I could also reopen it if the run fails.

Cheers,
Alison

comment:4 Changed 2 months ago by willie

Hi Alison,

I'll leave it open.

Regarsd,
Willie

comment:5 Changed 8 weeks ago by admg26

Hello,

I made a copy of the suite, make a few changes and ran it (u-bb932). This still fails with the same error after 12 years.

Many thanks,
Alison

Last edited 8 weeks ago by admg26 (previous) (diff)

comment:6 Changed 7 weeks ago by admg26

Hi,

I have followed the instructions for perturbing theta when there is a grid point storm and then restarted from the point where it failed. The runs appears to be going okay so far. You can probably close this ticket.

Cheers,
Alison

comment:7 Changed 7 weeks ago by willie

  • Keywords UKCA, grid point storm added; UKCA removed
  • Resolution set to fixed
  • Status changed from new to closed
Note: See TracTickets for help on using tickets.