January 21, 2014

Agenda

  • 2.3 stable release
  • KISTI SOW
  • 3.0 development

Topics

2.3 stable release

  • Bug list review
  • Schedule

KISTI SOW

We need to provide KISTI with a proposed SOW for 2014. Would like respond to KISTI with draft by end of January. Which of KISTI's requested items are doable? KISTI requested items

  1. Data importers for CAM, GRIMs (GRIB), MPAS
  2. Coupled visualization support: GRIMs & ROMS, WRF & ROMS
  3. Isolines 

Considerations for importers

Current VDC places numerous restrictions on data that may be supported

  1. Structured grid
  2. Single grid
  3. No units
  4. Georeferenced (lat-lon) data must be projected to Cartographic space

Questions that need to be answered when considering supporting a new earth-systems numerical model

  1. Can all variables readily be resampled to a single grid?
  2. Missing values
  3. Units
  4. Map projections provided

Probably too much effort to support MPAS with VDC2.

Coupled visualization

Too difficult with current VDC?

Isolines

Alan estimates 1FTE month for basic support, additional options to be considered.

3.0 Development

Review GUI and VDC prototype, meet and discuss. When can we meet?

Summary

KISTI SOW

MPAS and coupled model visualization are too difficult to support with current data model. We'll limit our proposed SOW to isolines, GRIM data, and possibly CAM data, depending on the expected level of effort required. Isolines will be limited to "basic" capabilities in Alan's proposal.

ECMWF's GRIB API needs to be tested with GRIMs data. The ECMWF metview tool is probably best way to do this.Metview provide a listing of a GRIB file's contents (similar to ncdump -h). GRIB files require a table to map variable numbers to variable name/contents. Each center may use different mapping. Don't have a map for GRIMs yet.

2.3

2.3 bug list is down to a single bug (text cropping) that Alan will fix. We will table further stable release prep (e.g. regression testing) until KISTI SOW is prepared.

3.0

We will review prototypes shortly. More discussion needed on how to proceed with 3.x and 2.x in parallel.

Action items:

  • Evaluate GRIM and CAM (scott, john)
    • ECMWF's "metview' tool may be helpful for GRIB (scott)
    • Get sample data sets from ECMWF (john)
    • Does NCL provide a GRIB to something translator (scott)
  • Send pointers to 3.0 prototypes (alan, john)
  • Fix one remaining 2.3.0  text bug (alan)
  • Schedule 3.0 meeting (john)
  • No labels