Re: Databases

> From:  Keith Downing
> Organization:  IKU Petroleum Research, Trondheim, Norway

Hi Keith,

>   Hello.  I work at IKU Petroleum Research as a computer scientist.  We have
> designed several systems for simulating the transport of marine pollutants
> and their effects upon the biota.  We use several different types of wind
> and current files for these simulations, and, frankly, it's become rather
> hairy with all the different formats and such.  The need to maintain
> compatibility with several software systems, not all of which are under our
> control, has made it difficult to establish a standard.  I have recently
> been given the task of creating a database for organizing all these files so
> that, for example, the user can select a lon,lat and time of year and get up
> a list of the currents and winds that are relevant.  I haven't considered
> the problem in detail yet, but on the surface, there appear to be several
> easy but less flexible solutions.
>   A few years back, I visited NCAR and got a bit of an "insider's view".  I
> was very impressed by their data organization system, although I only saw a
> few aspects of it.  I'm thinking that before I sit down and design our
> database, I should take a look at the design of an advanced system like
> NCAR's and see if there are any useful concepts that I could use.  Like I
> said, format constraints prevent us from really doing what we want here, but
> at higher levels of organization, there are certainly standards that we
> could enforce.
>   So I was just surfing the web and came across the netCDF page...and your
> name.  I was wondering if you could send me any info on the organization of
> these types of databases, or point me to another web page?

NCAR currently has many systems that provide access to lots of data in
different ways, so I'm not sure which one impressed you.  Some I know
about (which is probably only a fraction of the systems available) include

    The Scientific Computing Division's data archives: 
            http://www.ucar.edu/dss/index.html
    The Atmospheric Technology Division's field project data:
            http://www.atd.ucar.edu/atd_data.html
    The Research Data Program's Zebra software for integrated displays of data:
            http://www.atd.ucar.edu/rdp/zebra.html
    The VEMAP data project:
            http://www.cgd.ucar.edu/vemap/
    The Climate Analysis Section's Data Catalog:
            http://www.cgd.ucar.edu/cas/catalog/

There is also an NCAR-wide project, just getting under way, to provide a
better-organized access to NCAR/UCAR data resources:

            http://www.coare.ucar.edu/iita/

There's information about software for visualizing netCDF data that is
accessible from the netCDF home page you've already found, and some of these
systems (e.g. Freud, FERRET, GMT, Ingrid, ...) provide interactive access to
data specified by location and time.

Where things may be headed for this sort of system, in my opinion, is
something like the prototype visualizer written in Java that downloads
current data to the client browser, then lets the client interact with the
data at high speed using Java applets.  You can see an example of this at

    http://covis.atmos.uiuc.edu/java/Weather03.html

--Russ

______________________________________________________________________________

Russ Rew                                           UCAR Unidata Program
address@hidden                              http://www.unidata.ucar.edu