Skip to content
This repository has been archived by the owner on Oct 23, 2020. It is now read-only.

Default option for frazil ice formation changed to be false #204

Merged
merged 1 commit into from
May 21, 2014

Conversation

pwolfram
Copy link
Contributor

Fixes crashes for runs with linear EOS
causing thickness to become negative and exponential
in ocn_forcing_build_transmission_array to
return a SIGFPE.

Output from gdb can be used to identify this error:

Reading symbols from /Users/pwolfram/Documents/MPAS-pwolfram_fork/ocean_forward_model...rudone.
(idb) run
Starting program: /Users/pwolfram/Documents/MPAS-pwolfram_fork/ocean_forward_model
task 0 of 1 is running
Program received signal SIGFPE
_exp.L () in /Users/pwolfram/Documents/MPAS-pwolfram_fork/ocean_forward_model
(idb) where
#0 0x0000000100f176fe in _exp.L () in /Users/pwolfram/Documents/MPAS-pwolfram_fork/ocean_forward_model
#1 0x0000000100131513 in OCN_FORCING::ocn_forcing_build_transmission_array (mesh= (...), state= (...), forcing= (...), err=0) at /Users/pwolfram/Documents/MPAS-pwolfram_fork/src/core_ocean/shared/mpas_ocn_forcing.F:222
#2 0x0000000100298e12 in MPAS_CORE::mpas_core_run (domain= (...), output_obj= (...), output_frame=2) at /Users/pwolfram/Documents/MPAS-pwolfram_fork/src/core_ocean/mode_forward/mpas_ocn_mpas_core.F:499
#3 0x0000000100001c40 in MPAS_SUBDRIVER::mpas_run () at /Users/pwolfram/Documents/MPAS-pwolfram_fork/src/driver/mpas_subdriver.F:72
#4 0x00000001000016e1 in mpas () at /Users/pwolfram/Documents/MPAS-pwolfram_fork/src/driver/mpas.F:16

fixes crashes for runs with linear EOS
causing thickness to become negative and exponential
in ocn_forcing_build_transmission_array to
return a SIGFPE.
@douglasjacobsen
Copy link
Member

Some more information for people interested.

Frazil ice is formed when the temperature is below -1.8C. In this case, an initial condition mesh is used that has negative temperatures (order -4C). When frazile ice tries to form, it will try to remove enough water to raise the temperature to -1.8C. I'm guessing in this case the cell has less fluid than is needed for this, for the cell is evacuated.

In the future we might want to ensure that when frazil ice is formed, it won't evacuate the cell. But I'm not sure this is a big concern in a realistic simulation where we use a non-linear equation of state and temperatures at not expected to be significantly lower than -1.8C.

Thanks for finding this!

@douglasjacobsen douglasjacobsen merged commit 5097a37 into MPAS-Dev:develop May 21, 2014
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants