CF Naming Conventions
AQ CF standard names
This is the work space associated with the CF conventions description. While the actual approval process for new standard names happens exclusively via the CF mailing list, you can use these pages to prepare such discussions, monitor progress or ventilate new ideas. For additional information what this is about, please consult the main page on ESIP, the CF home page or directly the latest CF standard document or standard name table.
Discussion of standard name attributes for AirNow and RSIG variables
- AirNow variable list with available and suggested standard_name attributes as well as suggested points for discussion Media:AirNow_Main_Params_mgs.doc
- Commented RSIG variable list (only partially commented due to sheer size) Media:rsig_variables.xls
Brief summary: many quantities contained in either variable set have already been defined in the CF standard name table or they have close counterparts which require only simple additions to this table (for example adding another species name to the available list of "mole_fraction_of_X_in_air"). Some more discussion is needed on the following points:
- aerosol properties and size classes (currently, only pm2p5, nucleation_mode and coarse_mode are known in the CF standard name table)
- handling of source attributions (for example "SOA from benzene" or "anthropgenic aerosol")
- more specific descriptions of technical pecularities concerning (particle) measurements
- description of many variables concerning (CALIPSO) satellite retrievals
Discussion on CF metadata for time aggregates and AQ metrics
Recently, a discussion has started on the CF mailing list, how more complex data aggregates, such as the "daily maximum 8-hour ozone concentration" can be incorporated into the CF standard metadata. The present proposal is to extend the syntax of the "cell_methods" attribute to allow multiple operations. Example o3.cell_methods = "time: mean (period: 8 hours) time: maximum (interval: 1 hour)"..
This discussion is not closed yet, because people from other communities would like to see an alternative approach where the "original" time information could be preserved. In this proposal, one would (have to?) create an additional time axis which would carry the mean time and time bounds for the daily 8-hour intervals.