simulation data models

[Gerard]

Dear colleagues

The theory interest group has been working on a project to create a data
model describing the results of numerical simulations. The concrete focus of
this model is that it should assist the simple numerical access protocol
(SNAP, another theory IG project) in a similar way as the spectral data
model assists the simple spectral access protocol. As defined in Victoria,
SNAP is supposed to deal with systems existing and evolving in 3 dimensional
space. A scale is not prescribed, the simulations can range from full
cosmological simulations down to formation of planets and everything in
between.

The first goal of the modeling exercise was to create a so called analysis
model, which describes the domain we're interested in, without too much
emphasis on the final use and implementations. We have a skeleton available,
based on the patterns in the domain model I proposed with Pat Dowler some
years ago (see the DM pages). 

To test this model out it would be good to have some concrete data models
that people have used in their work. These may be explicit, such as models
for a database storing simulation (meta-)data, some may be implicit, as in a
description of attributes and data in HDF(5) files.

We would like to get a hold of examples of such models, with the aim of
investigating whether they can be described using the elements in the
current model and to find out whether new features should be added.

So my request is that if you own or know of the existence of such data
models, and these are or can be made public, please mail me about this, so
that we can take those models into account in this exercise.

Thanks in advance 

Gerard Lemson