TRIGRID co-linearity problem

[Christopher Davey <davey(at)cira.colostate.edu>]

I am trying to produce simple contour plots of temperatures from a
regional network of temperature observations. This should be a simple
problem: taking a grid of irregularly spaced data and interpolating it
to a regular grid. I have used the spherical triangulation option.

The triangulation( ) procedure works just fine, but In my call to
trigrid( ), I get an error that I cannot figure out: "Points are
co-linear, no solution." The source data is good. I have checked and
re-checked that my calls to trigrid( ) all have the correct syntax,
argument order, etc., so the code looks okay. I think one of the
archived messages I saw said something about a "great circle" problem as
a possibility for this error, but I didn't understand what was meant by
that. Is there anyone that might be able to clarify to me what is
causing this error, and how I can get around it? Any good references
that discuss this problem?

Christopher Davey
davey@cira.colostate.edu