<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7654.12">
<TITLE>Potential utility of 'priority' attribute in CMIP5 netcdf files?</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/rtf format -->
<P><FONT SIZE=2 FACE="Tahoma">Hi folks,</FONT>
</P>
<P><FONT SIZE=2 FACE="Tahoma">A question that came up during recent discussions was whether or not it would be useful to record the priority level (core/tier1/tier2) of a CMIP5 variable as an extra metadata attribute in its corresponding netcdf file. (I'm assuming it's not in there already - I couldn't see it in our latest test netcdf output).</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">As a data producer site, I can think of use-cases where it would be useful to issue queries that report on the core, tier1 and tier2 variables that we've generated. And while I'm not a scientist end-user, I can envision that they may also wish to perform similar queries, e.g. "show me all tier2 simulations|variables produced by UKMO"</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">Is this capability expected to be part of the ESG/CMIP5 web portal offering? It may well be that this will all be handled by the ESG catalogue/metadata solution, and I'm simply not aware of it :-)</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">Anyway, if we did want to encode the priority as an attribute in the individual CMP5 netcdf files it looks like this could be done by carrying over the 'Priority' column (domain = [1,2,3]) from Karl's excel spreadsheets into a corresponding property in the CMOR/MIP tables. From there the CMOR library could ensure that the priority value is trancribed into the netcdf output files.</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">The potential fly in the ointment is the fact that the priority attribute appears to apply to two things: simulations *and* the variables produced as part of those simulations. We know the simulations are either core, tier1 or tier2. In the case of output variables it looks (from the CMIP5 experiment docs) as though there are 'high priority' variables and 'low priority' variables. And some of us (guilty!) have got into the habit of calling the former 'core variables'.</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">I'd be interested to hear how folks envisage this will be handled in the ESG/CMIP5 web portal, and whether we need to encode one or both priority values in the netcdf files (it looks like a higher level cataloging issue to my mind)</FONT></P>
<P><FONT SIZE=2 FACE="Tahoma">Regards,</FONT>
<BR><FONT SIZE=2 FACE="Tahoma">Phil</FONT>
</P>
</BODY>
</HTML>