Versions Compared

Old Version 1

changes.mady.by.user Cheng Hsuan Lu

Saved on

compared with

New Version Current

changes.mady.by.user Cheng Hsuan Lu

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

General update (Yannick)

  • No special announcement

JEDI Infra (Evan)

  • Spack-Stack 1.8.0 testing is going well.  The official release is planned for the end of month 9/27, release is in-progress and available on some platforms such as discover SCU 16.
  • A new skylab feature to build jedi bundle is being worked on and will be shared soon.
  • Ongoing R2D2 dev involving API changes for users - we are reviewing and updating workflow code and have successfully tested end-to-end.

JEDI Model Interface (Francois H.)

  • Liam is working towards offloading interpolations onto GPUs; current work is on:

    • some refactors in the area of Global Interpolator / local interpolators.

    • looking at offloading sparse matrix multiplication in Atlas.

    • some investigation of Atlas's ECTrans interface.

JEDI Algo (Christian)

  • Planning and preparing for model variable naming convention sprint Oct 7-11 (Steve V)

  • Adding letkf/getkf suites in ewok/skylab (Clementine, PR)

  • Refactor output handling (model and obs output) in oops for code and yaml readability (Clémentine, PR + started working on model PRs)

  • Design for making the file I/O and MPI distribution independent steps in the ioda reader (Steve H, Ron McLaren EMC)

  • Design translation layer between ObsSpace API and ObsDataFrame API (Steve H, Phil Underwood)

  • Testing VarBC for aircrafts (Anna, OBS team, EMC)

  • Ensemble covariances with spectral analytical filter localization for mpas-jedi (including winds) (Nate)

  • Continuous DA: design for shifting obs window; updating observation data vectors between different loops instead of resetting them (Christian)

  • PR for general Pseudo Model in SkyLab (Christian)

SOCA (Travis) 

  • The team is working on name convention.  Only focus on ocean and ice sea, skipping biogeochemical variables.

COMPO (Shih-Wei)

  • Jerome and Maryam are working on clean up compo team experiments for skylab
  • 2 PRs of IODA converters are ready to merge
  • Shih-Wei continues his work on L1 reflectance DA

Obs team (Francois V.)

  • IROWG this week

  • Work with Ricardo on BT Jacobian related to hydrometers and unrealistic increments. Same issues also occur for skin temp and surface wind. While this issue is found for FV3 interface, Ricardo check MPAS interface as well (working fine)

  • TROPICS mission

    • 5 out of 7 smallsats are working

    • prepare for the mission
    • fix CRTM v3 coefficient mismatch, bug fix implemented by Ben J

CRTM (Cheng) 

  • TROPICS coefficients are fixed
  • New features related to radiance operators are added.  This is to prep surface DA and aircraft DA.
  • Radar operator from STAR has been implemented.  Future work is needed to compare it with the current operator.
  • continue the efforts to remove binary coeff files.  The last piece is MW sfc emissivitity..    All files will be in netcdf.
  • Code sprint planned for early Oct for surface emissivity and user interface.

MMM

  • Interface to CRTM is checked and there is no increment issues (see Obs Team) 

Met Office (Phil U.)

  • Interface is almost ready
  • The conversion is good, compared with the baseline system
  • Jedi seems getting noisier increments

Navy (Sarah)

  • Focused on ozone DA
  • Transitioned to github

GMAO (Ricardo)

  • Provide more information on his debugging efforts related to hydrometer increments (See Obs Team note)
  • While the issue is found for FV3, the interface for MPAS and UFS are checked and seem working fine
  • Compare increment for GSI versus JEDI.  For clear sky DA, the increments look healthy.  For cloud DA, the results for IR IASI shows JEDI getting unrealistic increments.  The issues are now narrowed in to aerosol coefficient and water emissivity

EMC (Dan)

  • Work on GEFS V17 implementation, with focus on atmos-ocean coupling
  • For global UFS
    • Chemistry, work on efficiency
    • sea ice: work on B-matrix
    • land: 2d Var
  • For regional system
    • work toward MPAS-JEDI
    • work on static B
  • To-do-list: remove D-grid wind from the background; build cycle for UFS-v18
  • While the issue is found for FV3, the interface for MPAS and UFS are checked and seem working fine
  • Compare increment for GSI versus JEDI.  For clear sky DA, the increments look healthy.  For cloud DA, the results for IR IASI shows JEDI getting unrealistic increments.  The issues are now narrowed in to aerosol coefficient and water emissivity