05/06/15 Notes

Participants : Sam Bacharach (OGC), Dave Cuddy (JPL), Glenn Cunningham (JPL), Siri Jodha Khalsa (NSIDC), Ken McDonald (NCSA), Bob McGrath (NCSA), Tim Smith (LPDAAC), Keith Stark, Ross Swick (NSIDC), Rich Ullman (NASA), Jingli Yang, Kathy Fontaine, Bill Teng, Ken Casey, Scott Lewicki, Jim Frew (UCSB), John Moses, Dennis Grosssman (Natureserve), Tony Zhang (rep Ming), Helen Conover (UAH), Robin Pfister (ESDIS), Yonsook Enloe, Frank Warmerdam, Michael Goodman (MSFC), Ananth Rao, Oscar Casteneda (TRFC), Rama (ESDIS), Elaine Dobinson (JPL), Mike Burnett, Keith Wichmann

Introduction:

Q: Connections between this group and other similar groups (e.g., IOOS, GEOS)?

A: Informal talks, others interested in adopting our process. However this group is about NASA, not national standards, so not appropriate to export our standards

Comment: If harmonization with other standards is a core goal, that becomes all you do.

Discussion: Large gray area between developing new standard and adopting well accepted standards of practice. Shouldn’t shy away entirely from the gray area. If we identify an area where there’s no standard practice but one is needed, then what? Should we flag and communicate that?

• Invite a tech note to discuss?
• Still the community’s responsibility to develop and document a common practice.
• SPG members (and REASoN projects) do have responsibility to look out for things to bring forward.
• Section of web site for concepts that community may want to develop into tech note or standard?

Presentations about current and potential RFCs:

Comment: For proposed standards, want to know:

• Community of practice and leadership
• Implementation and operational experience
• Clear benefits to wider use

SPG Action: Maybe make template for future meetings with these points

Q: What about competing technologies?

A: This should be TWG responsibility.

James Gallagher’s presentation (DAP)

SPG Action: Send this presentation to Martha (and maybe put endorsement on web site)

Comment: Concern that these standards will be orphaned if NASA ends support for SPG.

Next Steps (SPG): Consider process improvements recommended in James’ presentation

Frank Warmerdam (GeoTIFF)

Discussion: References EPSG (European Petroleum Standards Group) database of map projections and projection parameters, but users can also explicitly define projection parameters. EPSG does not define all projections commonly in use, only those important to petroleum community.

Comment: Concern about standard being orphaned again.

Comment: Is SPG the appropriate place to discuss potential improvements to standard (GeoTIFF 2.0)?

Discussion: What about additional information that’s needed with ES (as opposed to GIS) data? LP DAAC has non-standard common practice we could also document and adopt. Maybe don’t need to make it completely robust for ES data, because most GeoTIFF users just want to visualize in GIS system.

Next Steps:

• GeoTIFF 1.1 spec : Geotiff 1.0 with ambiguities resolved – Frank can get that to the SPG at the end of the summer
• Consider documenting other things (potential improvements, best practices for ES data, etc.) in tech notes. http://www.remotesensing.org/geotiff

Michael Goodman (level 2 precipitation science content standard)

IPWG (International Precipitation Working Group) - at Oct 04 meeting- endorsed strategy to work with GPM science team to make agreement on the level 2 Precipitation content standard. Michael to talk to Eric Smith and Chris Kummerow.

Discussion: Could document as Tech Note now, to use as basis for discussion within community. Want to get GPM buy-in first.

Discussion: Underlying data format? Not really the issue here.

Q: working group to contain data producers and consumers?

A: both (but modelers interested in L3 data, not covered)

Comment: Does this comply with ECS model or FGDC model? If so, can create profile of ECS or FGDC model for this content standard? GPM team should do this work.

Discussion: SPG role? Advocate benefit to GPM.

Next Steps: Work with GPM to build community interest.

Bob McGrath (NetCDF4: HDF5 and NetCDF)

NetCDF4 (Strengths of NetCDF3 and HDF5) – netCDF-4 beta : SEPT 05. Joint development by Unidata and NCSA & University of Illinois. Source and binary available for free. Packaging TBD. HDF5 standardization – effort with DOD and with ANSI NCITS (just starting, not sure what will be standardized in HDF5).

Both NetCDF and HDF5 widely used in the Earth science community. Want to bring something in to SPG sometime next year. What should be brought in? What is being standardized in other areas? Don’t want to repeat? No sense in bringing in HDF5 – maybe common practice doc for HDF5. But NetCDF4 not on any standards track anywhere else.

Discussion: What to standardize? API or data format? What’s not being standardized already elsewhere? SPG may help define what aspect to standardize

• maybe profile of NetCDF4 for NASA ES use. Still in development, so don’t expect to start working this for a while.
• Looking to tech note describing relationship between OPeNDAP 4 and NetCDF4/HDF5.

Q: Shouldn’t there be a standard for HDF-EOS which is a profile of HDF5? Should describe the existing data HDF format. We have national assets in HDF5. Can’t depend on one implementation of a library.

A: ECS toolkit user guide describes using this library to read the data. Can also use regular HDF software on any HDF-EOS data.

Q: What about long-term preservation?

A: Need to document HDF format – done. Need to document HDF-EOS format and metadata – also done? Tech Note to describe relationship between HDF, HDF-EOS? Need to preserve HDF and HDF-EOS libraries. May need to define archive format, which may or may not be HDF.

Next Steps: Consider RFC for HDF4. Wait for operational experience with HDF5 to bring HDF-EOS forward. Wait for community to bring netCDF4 forward.

Scott Lewicki (Aura HDF5 Guidelines)

Built on HDF-EOS 5 swath. Profile developed and documented by Aura science team. Would like this document reviewed and advertised by SPG.

Q: Any relationship to Aqua HDF4 profile?

A: Did not look at.

Discussion:

Action item : respond to Scott Lewicki

• SPG will identify RFC editor for the AURA HDF guidelines doc.
• Issue : who does CM for the doc? Need to rewrite the doc or can reference their document?
• What happens when AURA HDF guidelines doc gets updated periodically? Don’t want to redo approval process for each update. Only want to redo approval after a bunch of updates.
• Need guidelines on updates – updates that don’t rep new standard. Updates that rep additions to the spec.

Need to have website such that the approved standards can be found/accessed multiple ways. Sometimes hard to find a particular type of std. Also helpdesk to answer if certain type of std exists.

Is this a formal enough document to reference in RFC (Aura group maintains ownership), or do we need them to recast as RFC (SPG maintains ownership)? SPG needs to figure out the general case for profiles or specs that are externally documented, but not maintained by a standards body. How to handle evolution, new features? Need guidelines on updates that enhance (but don’t fundamentally change) a standard.

Next Steps: Develop RFC.

Robert Wolfe (Surface Reflectance Product content standard)

Robert Wolfe gave talk on surface reflectance product science content std – in a year or so, bring this to the SPG with some kind of std. Need to talk to scientists like Eric Vermote, and others. This would be a best practice when making a product of this kind.

What should SPG be doing wrt this? SPG waits for community interest in this.

Discussion: Determine which aspects are (external) metadata, which are metadata within each data file, which are recommended/required data content. Note data provenance is important issue in general.

Comment: Consider specifying data field names as well.

Comment: SPG needs to get very efficient at putting together science content standards (or tech notes). Need to publicize these as well.

Next Steps: Plan to present to science working groups and build community interest.

Sam Bacharach (OGC WMS)

NASA Earth-Sun Gateway http://esg.gsfc.nasa.gov

Q: Why WMS rather than another OGC standard?

A: Look for whole suite (W*S) over time. This is a good starting point, widely used already.

Q: What’s the benefit of approving WMS – already well-specified and widely used?

A: Allow NASA to claim interoperability with Geospatial One-Stop, DHS, etc. Encourage more people to use this technology, and to use this rather than a competing one. Might end up WMS servers integrated with DAAC Data Pools.

Next Steps: Currently working on RFC, should be ready by mid-July.

Siri Jodha Singh Khalsa (GLIMS data model)

http://glims.colorado.edu/

GLIMS concept – metadata model is candidate for standardization. Shape files. NASA profile of shape files for glacier info?

Adoption as community std would facilitate meta analysis and comparisons. SJSK – needs to figure out what parts of the GLIMS should be standardized and who would benefit from it.

Next Steps: Continue work within glacier community – gaining momentum. Determine which aspect to standardize on – XML schema?

Dave Cuddy (software guidelines)

Discussion: Different type of standard than what we’ve talked about – software engineering rather than interoperability. Can benefit mission science teams putting together processing software. Appropriate for this group to consider? * Delivered Algorithm Package is part of dataset that gets archive. May benefit long term archive – Life Cycle Working Group should consider.

• May be of interest to Software Reuse group, but they’ve not done anything with coding standards so far.

Next Steps (SPG): Pass action on to Life Cycle (Ken McDonald) and Reuse groups (Robert Wolfe).

Denny Grossman (NLCD vegetation classification)

Looking for endorsement of standards committee – encourage NASA to use these FGDC vegetation standards for Land Cover Classification products (or map NASA Land cover products to the FGDC standard).

Discussion: Relationship to FAO LCC (used internationally)? Need to review.

Different classification schemes are better for different uses. Relationship between land cover, wetlands and vegetation communities within FGDC. MODIS using several classification schemes – one not really sufficient. Research community looking toward continuum instead of classification – but classification really needed for application use.

Next Steps: Denny write RFC. SPG / NASA advocate for revitalized multi-agency vegetation working group under FGDC.

Robin Pfister (ECHO)

Looking for guidance on what to submit as standard. Probably Client API – based on programmer’s manual. Develop separate RFCs for each API (client, data provider, service provider). Model on what’s described in DAP RFC.

Q: NOAA or others planning to adopt ECHO?

A: JAXA and DLR interested in interoperating. Looking toward federated registries in future (ECHO instances or similar).

Next Steps: ECHO team telecon with SPG for guidance on what/how to document.

Glenn Cunningham (cryospheric data products)

From REASON work. Sea ice motion. Very small community looking at this. Technical note appropriate?

Rama ; how big is your user community?

Glenn : dozens

How small is too small of a community?

Small community interested in sea ice motion. Only this one group creating this product. Worth writing standard to describe data product?? Product content standard? Probably tech note more appropriate.

Next Steps: ???

Additional notes:

• Any part of website that has info - request/concept for standard that has not been met yet? Then can see if there is any interest. Concept may not be fully developed yet. SPG needs inputs from the ES stakeholders on this.
• Could be lots of science content stds. Bob McGrath – science content stds should be able to be reviewed very quickly. Community isn’t that big yet. Tech note? How to get tech notes publicized across communities?
• Siri Jodha found the Oasis UDDI website – defines the difference between Tech Note and Best Practices documents in terms of intent and specificity.