Note that the "conventions" being talked about here, which add agreed domain semantics to
the netCDF syntax, are effectively an informal style of information model standardization. The
effectiveness of netCDF as a vehicle for interoperability has depended on these, which enable
interoperability within carefully scoped sub-communities within the fluid earth sciences. In the
ISO/OGC context, these are referred to as "Application Schemas".
Simon Cox
From: owner-galeon@xxxxxxxxxxxxxxxx
[mailto:owner-galeon@xxxxxxxxxxxxxxxx] On Behalf Of Ben Domenico
Sent: Wednesday, 25 July 2007 12:41 AM
To: Tom Whittaker
Cc: Unidata GALEON
Subject: Re: weather station observation data in netCDF
Hi all,
Tom makes an excellent point. Even within our own community, it is crucial to establish
and observe "conventions" for data encoding forms -- e.g. CF conventions for
netCDF. Of course it is even more important when we are attempting to agree on
international standard access protocols/interfaces. My hope for GALEON is to develop a
mechanism whereby we can establish and evolve the conventions and protocol standards in
tandem.
For those of you who are not from the meteorology community:
-- METAR is short for MEteorological Terminal Aerodrome Report, the WMO
standard form in which surface weather station observations are transmitted
-- RAOB is RAdiosonde (or sometimes RAwindsonde) OBservation, the output of a
vertical upper air atmospheric sounding
These are the fundamental, long-standing, in-situ observations of the
meteorological community. There are thousands of METAR, and hundreds of
RAOB, reporting stations around the globe. METAR stations produce reports
several times per hour whereas RAOBs are launched only once or twice per day.
The METAR reports especially are important in many of the OGC and GEOSS demo
scenarios -- along with the weather forecast model output. And, as noted
earlier, the METAR observations are very similar in nature to ocean buoy
observations, river gaging station reports, air quality monitoring reports and
many others. So for interdisciplinary research projects and operational
scenarios where data from different sources must be integrated ( e.g. flood
situtions or contaminant plume dispersion), it is absolutely essential that we
establish a common set of conventions and protocols in order enable each
community to access and understand the datasets from the other communities.
My own personal desire is to minimize rather than proliferate the number of
conventions and protocols needed to effect the useful exchange of data between
communities of practice.
-- Ben
