You should still import as a 4x4 grid.  Within DX you can treat the data as
grid or scattered -- this way you have the option.  For the missing samples
put in a "missing" value, which you can flag as invalid with Include, but
kept in the field (culling should be off).  You can still use Autogrid or
Regrid on the scattered points, but you can also do Refine and other
operations on the "original" imported grids.

FYI, Paul Huang ([EMAIL PROTECTED]) implemented something similar
several years ago while helping a group from the NY State environmental
agency.  Part of their needs was to reconstruct reservoir bathymetry from
sample points.  He may have some further suggestions.


"Ted Lewis" <[EMAIL PROTECTED]>@opendx.watson.ibm.com on 12/10/2001
06:17:53 PM

Please respond to [email protected]

Sent by:    [EMAIL PROTECTED]


To:    <[email protected]>
cc:
Subject:    [opendx-users] Include module /  thinking out loud



I'm quite new to OpenDX - this is probably a very newbie question.

I recorded temperature, conductivity, transmissivity and depth with a CTD
at
a lake.  Recording was done in at 16 locations in a 4x4 grid.  In OpenDX I
made a .general file that contains X Y Z and "value".  It was imported as
scattered to allow for the occasional sampling site that doesn't conform to
the 4 x 4 grid.  Finally, I used Autogrid & made 2d slabs, isosurfaces,
etc.

I'm happy with the output - except in certain areas there are connections
(&positions?) at depths that the bathymetry should not allow (ie.
Interpolation has continued deeper than the bottom of the lake).  Not
surprising.  The bottom of the lake is very irregular - each sampling site
is a different depth. In the .general file, data at each sampling site
stops
at the lake bottom.

After combing through the User's guide, I think the approach I should take
is to use the Include module & modify my data file.  Rather than having the
data "stop" at the lake bottom, I could make Z at each X Y stop at some Z
that is deeper than the deepest sampling location.  I could make the grid
"aware" of the real bathymetry by adding invalid positions with Include
(ex.
X, Y, Z, flag).  Where the flag is used if Z > lake bottom at position X,
Y.

I thought I'd check to see that I'm on the right track before I go ahead,
since modifying the file to this format will take a fair bit of time.

Thanks in advance,
Ted Lewis



Reply via email to