Hi,
On the September Unidata community survey, in answer to
Question 21: If you have any suggestions or comments regarding Unidata
Data Management tools, please enter them below (e.g. enhancement or
future services):
a user commented
This isn't Unidata's issue per se, but we need to converge on a
method of handling irregularly gridded (e.g., FEM) data within
netCDF and CDM.
We share this user's desire for a way to store and access irregularly
gridded data that could be represented within the netCDF data model
and the Common Data Model. Unidata has participated in several efforts
to address this issue.
Back in 2006, NOAA and Unidata held a workshop with a goal to develop
community standards for unstructured grids
http://www.unidata.ucar.edu/events/2006GridWorkshop/index.html
Two results of that workshop were a small working group exploring
options for representing unstructured grids that might be proposed for
the CF Conventions
https://cf-pcmdi.llnl.gov/trac/ticket/6
https://cf-pcmdi.llnl.gov/trac/ticket/7
and UGRID, still under development, considered a leading candidate for
an unstructured data model built on the netCDF classic data model.
There is currently a Google Group on UGRID Interoperability that is
moving the UGRID proposal forward. According to a developer who is a
member of that group, NOAA's IOOS is funding a modeling testbed to
implement things like subsetting and interpolation of unstructured
grids inside of netCDF-Java.
In an unrelated effort, EXODUS II
http://sourceforge.net/projects/exodusii/
is a finite element model that uses netCDF for its finite element
grids. To handle larger meshes, it requires raising the current
limits significantly on the number of dimensions and variables per
file (to 65536 and 524288, respectively!).
The NSF-funded Ocean Observatories Initiative has also published
drafts that mention netCDF and DAP representations for unstructured grids.
Currently, netCDF lacks resources and expertise to help directly with
any of these efforts, but would be interested in hearing about
progress and other approaches for representing unstructured grids.
--Russ
