On Fri, 1 Oct 1999, Matt.Wilkie wrote:
We are creating elevation models from contours and sparse spot heights.
The resultant dems show significant benches at the contour locations.
When watercourses are used as breaklines, the benches are broken up
somewhat but not sufficiently. We do not want to rely on filtering to
"clean" up the data if there are other appoaches out there.
Matt,
Aha! I think that I know the source of the problem. Whatever you are using
to create point data (x, y, z) from the contours is not working well, and
... I think that perhaps your interpolation algorithm is incorrect.
There are several different interpolators, and each works best on
different types of data. That is, an algorithm which works well on
evenly-spaced data will not work well with much more random points.
I'm not looking for information tied to any particular program or
computing platform, but methodologies and pitfalls.
Two things. I've not written (or looked at) interpolation code for about a
decade. But, check out the term "interpolation" on the web, not "dem." Also,
I still use a MapInfo addon called Vertical Mapper. VM is written by a bunch
of geologists who started making more money by writing software. The company
is called Northwood Geoscience, Inc. and it's located in eastern Canada.
<www.northwoodgeo.com> They can interpolate DEMs from point data, and
decompose contour lines into point data for interpolation.
They may give you some guidance on the most appropriate interpolation
method, and some pointers to interpolation algorithms on the 'Net.
Unfortunately, I've no time to work on this with you. I'm overcommitted as
it is, and my wife insists that I have a life away from work and the
computers. <g>
Rich
Dr. Richard B. Shepard, President
Applied Ecosystem Services, Inc. (TM)
Making environmentally-responsible mining happen. (SM)
--------------------------------
2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
+ 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard@appl-ecosys.com