Jonathan Gregory writes:
>...
> It is also a limitation that the data have to be on the same space-time grid.
>For many analysis operations, however, this is necessary, even if the data were
>not generated like that. In Balaji's case of the AR4 archive, the models do not
>have the same grid. It is necessary to put them on a common grid nonetheless in
>order to produce some familiar diagnostics such as appear in the AR4 and recent
> papers, like the average lat-lon field of surface temperature change, or the
> time-depth field of ocean temperature change averaged over models with
> volcanoes. Even if the archive did not originally hold that data like that, I
> am sure that somewhere along the way the ensemble dimension will be needed.
Jonathan, if I understand the proposal here, it's that experiments will
be archived independently, and when someone constructs an ensemble out
of a set of archived runs, then some software will go in and add the
ensemble dimension?
Do you foresee that at some future date, data centres will take a number
of archived experiments, reorganize some set of runs into an ensemble,
and deliver a packaged ensemble dataset following the rules proposed
here?
I'm still only able to see performance problems, and apologize if I
missed something from earlier on in the discussion. As I've stated,
I do see aggregation as a fact of life. It's possible at small cost to
add attributes, and make minor changes to the netCDF file header after
the fact, but reorganizing data along a new axis is non-trivial.
My preference would still be to use the auxiliary coordinates to
identify individual datasets as belonging to a set or ensemble, but
defer the aggregation step as far as possible to the end-use software.
> ...
--
V. Balaji Office: +1-609-452-6516
Head, Modeling Systems Group, GFDL Home: +1-212-253-6662
Princeton University Email: v.balaji at noaa.gov
Received on Tue Mar 06 2007 - 11:10:22 GMT
