Support for thematic data?

[kfoley]

I am working with scientists that generate data on various aspects of climate conditions that occurred during the Pliocene. The data is used as input for climate models. I use netCDF format to distribute data and conform to CF metatdata standards when possible. One dataset we produce is a lat/lon grid of cells, each containing an integer that indicates the nature of ground cover that is predominate in that cell. There exists no continuity between these values -- 7 does not indicate an intermediate form between 6 and 8. Is there support for this thematic data in the CF standards that I have missed or is there some possibility of including it?

Thank you,

Kevin Foley USGS

[jonathan]

Dear Kevin

Yes, this can be done. The answer has two aspects.

1. We have recently introduced a standard name of area_type. A variable with this standard name is string-valued, and the possible values are listed in the area-types table, linked from the standard name page ​http://cf-pcmdi.llnl.gov/documents/cf-standard-names/. We expect that people will propose more types to be added, such as broadleaf trees, needleleaf trees, shrub, grass, etc. Please propose type names that you need to the CF email list (like standard name proposals).

2. The area-type variable can be encoded as integers using the flag_values and flag_meanings attributes. See 3.5 of the CF standard. CF would not standardise the integer codes, but of course you apply your own convention to choose these values.

Best wishes

Jonathan