Making Maps: Lab 7 - TINs and DEMs

I explored Triangulated Irregular Networks (TINs) a little earlier this year but this time we went further into how they are created, different possible variations on visualization in ArcScene and ArcMap, and how they compare to DEMs. As before, I really enjoy looking at and learning more about TINS.

First in lab was seeing how a remotely sensed image or other surface layer can be draped over a TIN to add vertical perspective. Considering I am not as interested in 3D mapping of urban areas, I do enjoy seeing natural areas in 3D. It is exciting to see a flat image of mountains or other areas "come to life," especially if the image is interesting or colorful. Next I looked at using a DEM to develop a suitability map for a ski resort. I created 3 component maps with suitability ratings then combined them through weighted overlay to produce a map highlighting the most suitable areas of 2,501 ft or high, 30-45 degrees slope, and west or south facing slopes.

From there I explored different elements of TINs, including nodes, edges, faces and other symbology. TINs are vector-models built through algorithmic triangulation based on irregularly spaced elevation points (nodes), as opposed to a DEM raster surface that is produced through interpolation of z-value grid cell centers. As TINs are formed with triangles, the appearance of the surface is angular, whereas a DEM is smooth. Both can be extruded in 3D and additional vertical exaggeration can be added to enhance relief. Contour lines can also be added to both types of surface models, although there is some variation due to the different construction algorithms. TINs are more efficient as they can use irregularly spaced points to model a surface, meaning more points can be placed in areas of high elevation variation and fewer points are necessary for areas with little elevation change. Accuracy is dependent on the input data (how it was created, e.g. from LIDAR data, field survey points, digitized contours, etc.), the intended purpose of analysis, and the surface type being modeled. TINs are often more accurate in steep, complex terrain as more points can be added to increase triangulation data, whereas DEMs may be more accurate when produced from grid-based data or with more regularly spaced elevation points.

One of the last parts of the lab was looking at how features can be used to modify a TIN. In this case, I looked at a TIN with a low elevation area where a lake shapefile can be set. At first, the edges are fairly simple, with larger triangles across the basin and smaller triangles as the shoreline terrain increases in elevation. I then "hard-edged" the lake shapefile into the TIN, which created a new edge and points to triangulate off of. The result is a more accurate and detailed terrain model with smaller triangles all along the shore and hills.