Minimizing Flat Triangles When Building a Tin from Contours

[TroyCarmann]

When I build my tin from contour data (I do NOT have the source data available i.e. point and break lines)  in certain areas 3D Analyst builds a tin that triangulates between three vertices on the same contour therefore producing a flat triangle. This of course is not a very accurate representation of the actual ground surface. Other surface building software address this problem with an option to "minimize flat triangles". This forces the software to triangulate between vertices at different elevations which seems to work adequately. BUT I can not find this option in ArcGIS.

Is there an option to "minimize flat triangles" when building surfaces in ArcGIS/3D Analyst?!

Do you have any other suggestions for building a tin from contour data?

[CarrieDavis]

Hi,

I do not have a solution for your minimize flat triangles issue using the TIN tool.  You may want to submit this idea as an enhancement request to the Ideas website at http://ideas.arcgis.com.   Another way would be to create a continous raster surface from the contours using Topo To Raster interpolation tool.  Then create the TIN using Raster To TIN tool under 3D Analyst > From Raster toolset.

How Raster to TIN works (online help doc for version 10)
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#/How_Raster_to_TIN_3D_Analyst_works/00q...

How Topo To Raster works (online help for version 10)
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//009z0000007m000000.htm

Hope this helps,  Carrie

[TroyCarmann]

Thanks for the input Carrie. If ESRI does not have this capability then it really does need to be on their list of enhancements for their software development. But I am hoping I don't have to wait that long to get the services I am looking for. This feature has been standard for more than half a decade in similar surface building software application. And I am hoping there is someone out there who knows how "minimize flat triangles" in GIS.

ANYBODY?!?

As for your suggestion, making an interpolation (Raster to TIN) of an interpolation (contour to Raster) of an interpolation (Real Points to contours) sounds like a good way to drastically reduce the quality of my data and probably wouldn't be a better than my "flat triangles" in the long run. But thanks for the suggestion.