Announcements:
- code freeze targeted for April 1st
- JEDI/Skylab release mid April
Infra (Dom):
- Installing release candidate for spack-stack 1.7: contact through Slack or GitHub w/ questions (Please don't email individuals, particularly ones who may be on spring break next week)
- AWS cost reduction:
- shutdown AWS ParallelCluster
- r2d2 mirrors on US-East 1 buckets removed (all data still available on us-east-2)
- Until further notice, do not start ec2 instances w/o discussing with JCSDA infra team (jedi-infra-support slack channel preferred)
- shutdown legacy CI system. new CI running with one of three compilers
- Fixing issue with cashing
- Contact Evan via the jedi-infra-support slack channel or tag in github issues/PRs (Dom is on leave next week)
- Merged PRs from Ryan Honeyager that improved exception handling
- Standalone app for IODA (upcoming in skylab v8)
- work on IODA obs space data frame model
- Continuing work on r2d2 REST API implementation - follow up to code sprint last week (demonstrator in Skylab v8, but Ewok won't use the new implementation yet)
- more obs types in ingest suite: GOES 16-17-18
Algo (Anna):
- Christian & Tom HTLM w/ high res
- Clementine: Geos & efficient cycling
- Beginning process of training B matrix w/ GFS ens.
- Anna: tuning b-matrix for 2dVar snow
- Francois H: convention for to/from FieldSet & Halo Exchanges - PRs for bug fix w/ small changes needed in Model Interfaces: targeting fix for next release.
- Adoption of Model Variable Naming convention: exploring new functionality for storing/passing metadata. separating obs vs model variables&metadata
MMM (Jake):
- Hybrid 4dEnVar working with MPAS-JEDI
- Question on date-time information in yaml: FieldSets check dates
Navy (Nancy):
- Working on Var BC issues
- Documenting integration with model space diagnostics
UK MO (Phil U):
- No specific updates.
- As mentioned by others:
- Tom Hill & Christian working on high-res HTLM experiment
- Chris Thomas's obs-related PRs (need review)
- Phil at JCSDA working with Infra team on obs space
Obs (Francois V):
- VARBC sprint branches going to be merged week of april 8th (after freeze): See script for updating coefficient files/names in Discussion 118 (https://github.com/orgs/JCSDA-internal/discussions/118) for more information on updating coeff files. Reach out to Hui or Obs team w/ questions
- Related to the VarBC coefficients and covariances, NRL has opened a discussion (#117 https://github.com/orgs/JCSDA-internal/discussions/117) where one could find plots of the convergence of the coefficients and covariances after one month of cycling. The question of the initialization of the coefficients and covariances is also discussed.
- working on Ocean wind surface speed operator w/ UKMO (see Discussion#110 and UFO#3235)
- Space born radar PR this week (UFO#3246)
- Ground based radar (UFO#3234, UFO#3237) and "where" operator (UFO#3258) PRs from UKMO needs review
EMC (Cory):
- Progressing on end-to-end buffer with UFO
- Application specific work on transition to operations
NASA-GMAO (Ron):
- Progress on versions of cycling system w/ GDAS & Swell
- Ricardo opened issues in fv3: can't output increment from 4d-ensemble
CRTM (BenJ):
- Enabling clear/cloudy radiance output simultaneously
- Visible & near IR work progressing; CRTM having issue with a few channels
- Cheng is working on the CRTM optional interface that enables optical profiles as input for radiative transfer calculations (targeting for end of quarter)
SOCA (Kriti):
- Cleaning up soca: deleting lots of code
- 3dVar UFS-marine setup in skylab
- Hamideh: getting OASIM hofx working in skylab for ocean color radiances. Got access to sample NOBM data and re-gridded to match MOM6 bkg in soca and created a PR to read biochemistry bkg from NOBM files, working towards generating 1/4 degree NOBM bkg to put in r2d2.
COMPO (Jerome):
- Finished with clean up in UFO
- Cycling with geos-cf
- debugging geometry issue with 4DVar (Maryam: number variables is possible issue)
ROMS:
- Hernan: work on adjusting resolution in inner-loops for handling islands/coastline and dealing with land-sea masking. Tricky issue, considering using a higher order interpolator near coasts.