5
$\begingroup$

I'm just getting into climate research, and am amazed at how much poorly documented legacy code is used. This kind of thing means that models more or less have to be maintained by the people who wrote them originally. If there was a good open standard for generic data exchange between components of a spatio-temporal model like this, most models could be split up into modules that could more easily be switched in and out, and more easily maintained (you wouldn't have to have such a broad grasp of what was going on inside other modules, and could concentrate on your own area of expertise).

So, are there any such open standards?

Data for climate models might include latent and sensible heat, fluid densities and velocities, vegetation and snow cover, snow and ice depth, etc. etc. It'd be a bit pointless to have a standard just for climate modelling, so ideally the standard would be applicable to any modelling of this type.

| cite | improve this question | |
$\endgroup$
3
$\begingroup$

Interoperable Technologies for advanced Petascale Scientific Computing, or ITAPS, has defined several interfaces in an effort to standardize data communication between large-scale software components. In particular, take a look at Tim Tautges' MOAB software package, which implements the ITAPS iMesh interface.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Hey, thanks, that looks very interesting. There's more about iMesh here: itaps.org/tools/interfaces/index.html. Looks like it's not a complete standard yet, but an excellent start. The software looks interesting too. I was actually wondering if anyone was doing climate modelling with adaptive meshes, but it kind of looks like it's still at the prototype stage. $\endgroup$ – naught101 May 15 '12 at 1:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.