NumSim Numerical Simulation Metadata

1
NumSimNumerical Simulation Metadata

• Bryan Lawrence1, Charlotte Pascoe1 and Phil
  Bentley2
• 1BADC
• 2Met Office

2
Why NumSim?

• Need to document and discriminate between the
  variety of model simulation data held at BADC.
  Need to do it NOW!
• NumSim is not intended to be THE Solution, we
  will go with Metafor, but we see NumSim type
  information as critical as both human readable
  and human creatable abstracts to a more complete
  solution.
• Not all models will be instrumented to produce
  METAFOR/CURATOR information, so our data
  scientists will need to be able to create this
  stuff on the fly!

3
LINK Project

• Defra (a UK government department) funds the UK
  climate-impacts LINK programme.
• Under the auspices of LINK, we take MetOffice
  model data to make it available to the research
  community
• (particularly outside of the atmospheric
  sciences)
• Because the Hadley Centre generate the model
  data, we need to know what it is!
• Data is flowing now, and we need it documented
  hence NumSim. We expect to start getting NumSim
  documents for all (recent) Hadley Centre model
  data we hold, and all future streams (until
  NumSim is replaced)

4
Overarching Metadata Environment

• Core linking concept is the deployment

of a Data Production Tool
at an Observation Station
on behalf of an Activity
that produces a Data Entity
Activity
DataProductionTool
ObservationStation
Links the metadata records into a structure that
can be turned into a navigable XML using Xquery
or XSLT with any of the record types as the root
element.
Deployment
Each of the main metadata objects has security
data attached to it. This means that this can be
applied to queries on the metadata
Data Entity
5
Moles based model navigation.
6
(No Transcript)
7
(No Transcript)
8
NumSim Schema Concepts - 1
9
NumSim Schema Concepts - 2
10
NumSim Schema Concepts 3
11
NumSim Example
NumSim Example
12
Problems with NumSim

• Over-reliance on external (GML) schemas slows
  parsing and editor startup
• (fixable with think with lazy evaluation)
• Need to include support for XHTML
• (fixed in the next version)
• Too much plain text
• (depending on METAFOR progress, well either use
  what we have by early next year as the basis for
  constructing more controlled vocabs in NumSim, or
  move to an early version of the abstract of th
  CIM)
• Improve citation and outbound references