Hi Milan,
I am not sure what happened with the descriptor 022039. But I noticed
that this entry, with ref=3D-5000, scale=3D3 and 12bits, cannot
handle the "meteorological residue of tidal elevation". The value of
residue can be positive or negative (see attached example).
If ref=3D-5000 and scale=3D3, 12bits data width can only describe data
ranging from -5m to -0.906m. 13bits would allow a range from -5m to
3.19m. The person who made the Note(4) may know this problem and
suggested to use entry 022040, which got 14bits.
Kelvin who is in Geneva this week (ET-IO meeting) might be able to tell
more about the history of this entry.
Cheers,
Weiqing
--------------------------------------------------------------
Dr. Weiqing Qu ,-_|\ Phone: +61 3 9669 4236
Communications Software/CCSB / Au \ Fax : +61 3 9669 4128
Bureau of Meteorology \_,-._/ Email: W.Qu@xxxxxxxxxx
700 Collins Street, Dockkland, v Web : www.bom.gov.au
Victoria 3008, Australia
--------------------------------------------------------------
From: Milan Dragosavac [mailto:Milan.Dragosavac@xxxxxxxxx]
Sent: Fri 19/09/2008 18:58
To: John Caron
Subject: Re: [bufrtables] More on table versions
Dear Jeff and John,
Thanks for point out 022039 difference. I checked all versions of the
tables I have and the result is 3 -5000 and 12 bits data width.
However, WMO version 13 introduced change to 13 bits which I can not
recall why. In the same time there is a Note(4) which state to use
022040 instead of 022039. the entry 022040 got 14 bits data width.
Obviously there is a conflict. I believe 13 should be set back in the
version 13 to 12 bits. I would like Weiqing to comment on this since
they produce tide data.
Best regards
Milan
Milan Dragosavac
ECMWF
Shinfield Park, Reading, Berkshire, RG2 9AX, UK
Tel: (+44 118) 949 9403
Fax: (+44 118) 986 9450
Telex: 847908 ECMWF G
E-mail: milan.dragosavac@xxxxxxxxx
