NCAR / DART

Data Assimilation Research Testbed
https://dart.ucar.edu/
Apache License 2.0
184 stars 139 forks source link

Eliminate amp adjustment #604

Closed jlaucar closed 7 months ago

jlaucar commented 7 months ago

Description:

Removed an amplitude adjustment factor from inv_bnrh_cdf. This was a legacy bit of code that was required to bitwise reproduce earlier versions. About half of ensemble sizes continue to bitwise reproduce with this change using Jeff's lorenz96_tracer_mod tests. For the others, the amp_adj was different from 1 at the smallest writeable decimal point and the tests ran producing reasonable, but no longer backward compatible, results.

Fixes issue

Types of changes

Documentation changes needed?

Tests

See description.

Checklist for merging

Checklist for release

Testing Datasets

jlaucar commented 7 months ago

Helen, I removed this, tested a compile, committed and pushed. I didn't like the message I got from the push so not sure if it worked or if I need to update the pull request in any other way. Please let me know if there are problems.

On Mon, Dec 18, 2023 at 11:54 AM Helen Kershaw @.***> wrote:

@.**** requested changes on this pull request.

looks good Jeff. You still have amp_adj in the declarations for inv_bnrh_cdf. You can take that out since amp_adj is not in the subroutine anymore

-real(r8) :: q(ens_size), curr_q, lower_q, upper_q, del_q, fract, amp_adj +real(r8) :: q(ens_size), curr_q, lower_q, upper_q, del_q, fract

— Reply to this email directly, view it on GitHub https://github.com/NCAR/DART/pull/604#pullrequestreview-1787525840, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANDHUIXVFLVSZ2VTN5VDXG3YKCGQFAVCNFSM6AAAAABAZ4QLK6VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTOOBXGUZDKOBUGA . You are receiving this because you authored the thread.Message ID: @.***>