Re: basic topology and the Data Model with geo-locations

"From: Bill Hibbard " wrote:
> 
> Hi Doug,
> 
> > I'm partial to (time) -> ((lon, lat, alt) -> (obs)). This makes
> > resampling by time or space rather convenient. VMET assumes such an
> > arrangement.
> 
> Well, at any given time, the aircraft is only at one location,
> so the domain sampling of ((lat, lon, alt) -> obs) would be
> a SingletonSet with manifold dimension = 0.  That is, to
> precisely represent the aircraft track.  So you might as
> well use (time -> (lat, lon, alt, obs)).  It only takes a
> few lines of code to translate between:
> 
>   (time -> (lat, lon, alt, obs))
> 
> and:
> 
>   ((lat, lon, alt) -> (time, obs))
> 
> However, if you want to graphically represent the aircraft
> track as an animating streamer, then you would produce a
> Field with the type you described:
> 
>   (time -> ((lon, lat, alt) -> obs))
> 
> where the domain Set of each ((lon, lat, alt) -> obs) is
> a segment of the aircraft track between the current time
> [i.e., domain sample of (time -> ...)] and current time
> minus some constant.
> 

For the single purpose posed in the original message, I agree that
(time -> ((lon, lat, alt) -> obs)) would be overkill. But VMET is trying
to integrate various data sets into a common framework and I have found
that this works well. I can pull out a given time step from all my data
sets and display that or I can take the spatial domain set of the flight
path and interpolate my 3D model output to the path,....

Doug

*----------------------------------------------------------------------*
| Doug Lindholm, Software Engineer          |  E-mail: lind@xxxxxxxx   |
| Research Applications Program             |   Phone: 303-497-8374    |
| National Center for Atmospheric Research  |                          |
| P.O. Box 3000                             |     There's no place     |
| Boulder, Colorado 80307-3000              |        like $HOME        |
*----------------------------------------------------------------------*