[GO-ESSP] Re: Update on Progress of Numerical Model Metadata (NMM XML) aka the EarleySuite

[Karl Taylor]

Dear Katherine,
Thanks for sharing your progress.  The prototype pages look quite 
useful, but I must admit, I'm a real novice in this area.  Being able to 
create tables like your sample IPCC numerical properties table could be 
especially appreciated by analysts.

Just to make sure that someone out there is not misled, could I suggest 
the following minor changes in http://cgam.nerc.ac.uk/NMM/ipcc/index.php
1) change "IPCC numerical properties" to "IPCC atmospheric model 
numerical properties".
2) change "IPCC climate model documentation" to "IPCC atmospheric model 
documentation"
3) change "(see original)" to "(for more complete documentation, see 
original)"

My fear is that someone will think that all the IPCC model documentation 
has been captured by your pages, which is not what you intended to do at 
this point.

Thanks for all your work on this.  I am sure an approach like yours will 
lead to a much more systematic and comprehensive documentation of models 
and their simulations in the future.  This is very much needed.

best wishes,
Karl



K.A.Bouton wrote:

>Just an update on the progress of the numerical model metadata schema and
>its implementation.
>
>While it may seem quiet, in fact a quite a bit of progress has been made
>such that we are close to implementing it as an in-house tool to use it 'in
>anger' and see if we can really get out of it what we want - comparing and
>contrasting model run properties/configurations etc. 
>
>Also we have been able to use the schema on the IPCC documentation to be
>able to compare and contrast IPCC model runs.
>
>Details below:
>
>1. due to high demand for the name to reflect more generally the intent of
>the schema - the EarleySuite has been renamed NMM (Numerical Model Metadata)
>XML. The home page and Wiki are reached via http://cgam.nerc.ac.uk/NMM/
>
>2. We have generated an proto-automated process for the Unified Model (the
>model used here) to produce the the most current version xml , then to
>'style' the XML vis xslt into a table.  This is done via a script in the
>local model creation tool which queries a local MySQL database, gets user
>details, machine details, parses the model tool configuration file for
>required fields like science parameters and puts it all into an XML file.
>
>This initial results for this can be seen under implementation-> personal
>notebook. Pretty nifty actually. As usual, what looks really simple took a
>lot of hard work. http://cgam.nerc.ac.uk/NMM/personal_notebook/index.php
>
>The schema is looking much more full--pretty good for the general
>information, technical and science sections, looking better for the
>numerical sections (discussions soon with Balaji on how to implement his
>wonder grid into NMM) and looking better with input/output descriptions.
>
>You can view the schema in various forms --graphical, html, xml, structure
>etc from http://cgam.nerc.ac.uk/NMM/schema.php.
>
>Note - I have fallen a bit behind in updating the implementation of the
>schema via a database so viewing that part of the www site may be out of
>date or not work.  I will be working on xslt to database insert translations
>in the near future.
>
>
>3. We have taken the IPCC Model documentation from the PCMDI WWW site and
>entered it into the NMM schema, to re-create the table they have online, but
>using the XML files. http://cgam.nerc.ac.uk/NMM/ipcc/index.php
>You will notice a lot of blanks, but that is because information is missing
>from the documentation,as expected.
>We especially looked closely at the numerical properties of each IPCC model,
>comparing and contrating them via a table which can be seen at 
>http://cgam.nerc.ac.uk/NMM/ipcc/index.xml
>
>This table looks at the 17 individual NMM XML files (one for each
>experiment) pulls out the Numerical Properties for each model, and formats
>it via xslt.
>Also pretty cool I think.
>
>
>Things we are working on:
>1. Implementing Balaji wondergrid into numerical properties
>2. Local tools for input files identification and description
>3. PRISM/FLUME cooperation
>
>Always something more
>
>Comments, suggestions requested, in fact - demanded (grin)
>
>Katherine
>
>
>  
>