Questions on WSIs and their inclusion in BUFRs

Questions on WSIs and their inclusion in BUFRs

by WONG Shwei Lin -
Number of replies: 2

Hi all,


We have a question on the WSIs.


As a historical practice, we have all along used the same 5-digit WMO ID for our synoptic upper air and surface station although they are some 11km apart (registration in OSCAR is based on the location of the surface station).


To better represent the metadata, we intend to create a new WSI for our upper air station.  For the new station ID, we intend to:


(1) Indicate the “Issuer of identifier” using the ISO country code (as the new ID will not have an exact corresponding entry in the old Vol A)

(2) To however retain the 5 digit ID value as the local identifier (to enable continuity)

(3) Report the WMO block number/ station number (001001, 001002) fields using the old 5 digit ID in the BUFR message (rather than “missing”)


Can we check if this is the correct approach, and how best to reflect this in OSCAR pls. Thanks. 


In reply to WONG Shwei Lin

Re: Questions on WSIs and their inclusion in BUFRs

by Joerg Klausen -

My suggestion would be to use the ‘issue number’ (block 3 of the WSI) to differentiate. This is what was done during the integration of the original Vol A, where the Subindex Nbr was 0 for surface stations, 1 was used for upper-air. 

In reply to WONG Shwei Lin

Re: Questions on WSIs and their inclusion in BUFRs

by Luis Filipe NUNES -

Dear Shwei Lin,

I'm trying to reach to a conclusion here.

As you know there was a parallel discussion on this by email, which I try to summarize below: 

 - it has been agreed that a new WSI for the upper air station is needed,

 - in terms of assigning a WSI to the upper-air station, your ISO country code should be used as Issuer of Identifier,  as proposed approach described under 1. in your message,

 - regarding your proposed approach described under 2. in your message (to retain the 5 digit ID value as the local identifier) there is still no clear guidance, since there might be advantages (enable some continuity) and disadvantages (might cause confusion for data users).

I hope this correctly summarizes the discussion, however, any additional comments, suggestions are welcome!

Thank you, kind regards

Luis