Horizontal and Vertical Datums in Operational Meteorology

Tommorow afternoon I'll talk about the idea below in OGC Met-Ocean DWG.  Hope that makes sense for them.

---

Volume A, Publication 9 of WMO is a catalogue of observing stations, compiled from submission from WMO's member states.  That, of course, includes latitude, longitude and hight above MSL (mean sea level), which are measured from some datum.  Which datum?  That is the question.

Honestly I don't know.  And I'd like to be honest.  So I'd like to make it clear that the datum is unclear.

Recommendation 1, CBS-Ext.(06) recommended that WGS 84 and EGM-96 are the references i.e. datums for coodinates and hight measurement.  But I saw not so much coordinates in the catalogue has changed since then.  So it is highly likely that many member states do not recognise what the recommendation is, and the coordinates/hight measured from past national datums are still listed.

For horizontal coordinates i.e. latitude/longitude that is no problem.  Volume A used to have one arc-minute precision.  The format has been changed to have one arc-second precision, but there are still a bunch of entries with zero-zero seconds.  One arc-minute is roughly one kilometer on the earth, and that is the precision of meteorological analysis.  If there is some coordinates data from different datum, the difference is usually less than a kilometer, so we can live with the situation.

Vertical datum is different.  Some national datums have difference more than one meter from EGM-96.  And we want 0.01 m precision (even the old format had one meter precision). One meter error means 0.1 hPa error in pressure, which is significant in operational meteorology.

So we need to be clear that the datum is unclear.