-
Notifications
You must be signed in to change notification settings - Fork 153
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fix a bug in EnKF that can cause segfault in a diagnostic print #12
Conversation
@RussTreadon-NOAA - this should probably get included in the GFS v16 branch since this bug could lead to a crash. |
Hi Jeff, Since this is only a single commit, I can move forward with the change. Please create a new GitHub issue detailing the issue that you corrected, then I can merge the pull request. Thanks, Mike |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Undefined array, mem_ob, removed from print. Looks good to me.
Closes issue #15 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree with Russ. Removal of unallocated array, mem_ob, from the print is fine. Approving change and merging pull request.
…variable for regional FV3 DA (NOAA-EMC#12) *Bug fix for EnKF analysis steps when using t (tv) as a control vector variable for regional FV3 DA. *Whitespace and indentation clean-ups were made in src/enkf/gridio_fv3reg.f90. Notes: This PR addresses issue NOAA-EMC#13, which was reported by Jonathan Poterjoy (UMD). And this bug should have affected all the EnKF steps (mean, update, recenter, etc.) The bug fix is based on the diagnoses and discussions among @BinLiu-NOAA and @TingLei-NOAA (EMC), @XL-OU, Jonathan Poterjoy (UMD), Jason Sippel (HRD), etc.
…nd (#12) 2m temperature is sensible temperature now. The convert from 2m potential temperature to 2m temperature happens during background IO stage.
use stack ncdiag, update hera.gnu hpc-stack, update to head of develop
in src/enkf/mpi_readobs, if the ensemble spread in observation space exceeds a value of 1000 a print statement is triggered that tries to print an array (mem_ob) that has already been deallocated. This causes a segfault on Orion. The fix is simply to not print mem_ob.
Does not affect results, but protects against a possible segfault if ensemble spread exceeds the threshold value.