Home‎ > ‎

Plan for CF-netCDF Encoding Standard

Ben Domenico
October 25, 2009
Updated June 8, 2010

Alternative Approaches

The bulk of this document describes the traditional approach to establishing a new OGC standard specification.  However, the OGC is now considering a Fast Track option "for the special case of externally developed de-facto standards ... , a more efficient review and approval processes is required in the OGC."   There is a draft of the proposed Fast Track Policies and Procedures document among the OGC pending documents entitled  "Draft OGC Fast Track Policies and Procedures" at

With that in mind, those of us interested in CF-netCDF standardization are compiling a compendium of documents and metrics that support the fact that netCDF and the CF conventions are in fact a "de facto" standard.   At the moment, this is an information-gathering  work in progress and is available at:

OGC Fast Track Supporting Information.

Traditional Standards Process Outline

The goal of this activity is to establish CF-netCDF as an OGC binary encoding standard. The approach is modeled on that taken for establishing KML as an OGC standard for XML encoding.   One difference is that  CF-netCDF will be standardized with a core and a set of extensions..  The standardization process for the core and each of the extension will involve the following steps:
  • Form a CF-netCDF Standard Working Group (SWG) -- Done
  • Have kick-off telecon on October 14 -- Done
  • Finalize work plan and timeline -- This is it.
  • Get cover letter to the OGC in the next 30 days -- In progress.
  • Edit draft candidate standard -- In progress
  • Release draft candidate standard to the OGC Architecture Board for review. (This can happen at any time but must be before release for public comment. This review usually takes two weeks or less.) -- Nov 12 telecon?
  • After positive OAB review, SWG can then vote to release the document for a 30 day public comment period. SWG works with OGC staff to prepare and press release. -- Dec 3 telecon??
  • After 30 days, collate the comments (if any) and work the comments and edit the document. Each comment must be responded to.
  • When the SWG is ready, vote for a TC adoption vote. You do not need approval from the TC to request an adoption vote.
  • Adoption vote happens. Requires 60 days.
This process is repeated (most likely with some overlap) by the SWG for the core and for each extension  standard.

Core Plus Extensions
The plan is to start with a core standard (the NetCDF classic) and at least one extension for the CF conventions.

1. A core netcdf encoding standard that is essentially the same as the netCDF standard that has been through the NASA ESDS (Earth Science Data Systems) Standards Process Group (SPG) process, namely: 

This will provide the foundation or substrate on which a number of extensions can be built.  Exactly what those extensions will be is not clear in great detail at this time, but CF will have to be among the extensions in order to provide the needed semantic elements.   A draft of the netCDF specification has been submitted to the OGC TC as discussion paper: 09-122

2. A CF extension standard based on the Climate and Forecast (CF) Conventions


The CF Conventions standard document is also being developed for submission to the NASA ESDS Standards Process Group.  So the CF extension will also be a joint OGC / NASA standard.

Thus the core netCDF standard and the CF extension will form the basis for OGC CF-netCDF encoding specification.   Other extensions may be added, but those will be addressed later.


The proposed timetable for establishing the core standard and the CF extension as an OGC encoding specification is ambitious:
  • September TC meeting
    • submit the initial core candidate standard as a discussion paper (done)
    • establish the CF-netCDF SWG (done)
  • November and December SWG telecons 
    • submit candidate standard for the core)
    • submit the draft  for the CF extension standard to the OGC TC as a discussion paper
  • December TC
    • submit the refined core netCDF candidate standard for public comment
  • Spring 2010:
    • submit the CF extension standard for public comment
    • submit the core netCDF as formal standard
  • At each subsequent TC meeting, move each candidate standard extension  to the next step in the process as outlined.   
The idea is that, since the netCDF classic format and the CF conventions for gridded data are already in wide use internationally and are well specified, the process should not require too much discussion of the content.  Hence the main work will be getting the documents into the proper form which  we hope will only take about 3 TC meetings for the netCDF core and the CF extension for gridded datasets.

Future Extensions and Related Efforts

Other possible extensions that have been discussed are the netCDF API(s) and the data model(s) as well as XML/GML extensions that include the semantic metadata in a form compatible with the other OGC standards.
One example of such a possible extension is ncML (netCDF Markup Language) and ncML-GMLencoding as well as encodings for other CF convention data types such as the proposed CF conventions for point/station data.   Initial discussions have been initiated with the CF community and a draft CF standard specification is being developed.  Further discussion will take place at the GO-ESSP (Global Organization for Earth System Science Portals) meeting October 6-8.

The plan is to conduct this CF-netCDF effort in parallel with the work already underway in the WCS SWG to establish CF-netCDF as a specific binary encoding format for WCS.

Charter Members of the SWG



Ben Domenico

Unidata (NCAR)

Stefano Nativi,

Italian CNR-IMAA

Dominic Lowe


Roy Mendelssohn,


Luis Bermudez


Rich Signell


Stefan Falke

Northrop Grumman

Jon Blower

University of Reading

Martin Desruisseaux


Chris Little

UK Met Office,

Marie Francoise Voidrot

Meteo France

Frederic Guilland,

Meteo France

Related Documents:

Ben Domenico,
Sep 29, 2009, 2:52 AM