Types of source data supported in terrain datasets

Photogrammetrically derived mass points

• Irregularly spaced
• Surface-specific peaks and pits
• Pseudo-regularly spaced points gathered to achieve a specified minimum sample density/nominal point spacing
• Minimal attribution
• Typically moderate quantities

Photogrammetrically derived breaklines

• Surface-specific breaks in slope along linear features
• Natural landforms (for example, ridges and valleys)
• Terrain features resulting from earthworks (for example, bulldozed)
• Lake shorelines, creeks, river shorelines
• Edge of road pavement
• Delineation of void areas (usually obscured by vegetation)
• Minimal attribution
• Typically moderate quantities

GPS points

• Irregularly spaced
• May or may not represent surface-specific peaks and pits
• Used for control, QA, and benchmarking other data
• Property boundary control points
• Potentially significant attribution per point
• Typically small quantities

Lidar points

• Irregularly spaced
• Not surface specific
• Sampled to required nominal point spacing varying from 1 to 15 feet (approximately 0.3 to 5 meters) depending on application
• Often, but not always, filtered to bare earth points only
• Limited attribution (what's available typically not very useful to end users)
• Typically in large quantities

Sonar

• Similar to lidar in terms of use in defining a surface
• Dense, non-surface-specific, irregularly spaced points

Points, lines, polygons (mass, break, clip, erase, replace)

The thematic data types used to derive a surface are stored in the geodatabase as features with different geometry types. These types are points, multipoints, polylines, and polygons. These geometry types can be incorporated into the terrain surface in different ways. You control how a feature is used by setting its surface feature type (SFType). The following table shows the relationships of different SFTypes, types of feature classes that can be used for each SFType, height source options, and thematic data examples.

Surface feature types in terrain datasets

Surface feature type	Feature class	Z-value source in the feature class	Thematic data type examples
Mass points (x,y,z locations)	Point feature class	Shape geometry; x,y,z per vertex or Attribute column holding one z-value per shape	Spot heights Survey points GPS points
	Multipoint feature class	Shape geometry; x,y,z per vertex or Attribute column holding one z-value per shape	Lidar points Sonar points
	Line feature class	Shape geometry; x,y,z per vertex or Attribute column holding one z-value per shape	Contours
	Polygon feature class	Shape geometry; x,y,z per vertex or Attribute column holding one z-value per shape	Shoreline delineation
Breaklines (hard or soft)	Line feature class	Shape geometry; x,y,z per vertex Attribute column holding one z-value per shape No height source; z-values interpolated for each feature from the surface before being added	Contours Edge of pavement Waterlines (streams, rivers, canals, shorelines)
	Polygon feature class	Shape geometry; x,y,z per vertex Attribute column holding one z-value per shape (such as a shoreline) No height source; z-values interpolated for each feature from the surface before being added	Lake shoreline
Clipping polygons (hard or soft)	Polygon feature class	Shape geometry; x,y,z per vertex Attribute column holding one z-value per shape (such as a shoreline) No height source; z-values interpolated for each feature from the surface before being added	Study area boundary
Erase polygons (hard or soft)	Polygon feature class	Shape geometry; x,y,z per vertex Attribute column holding one z-value per shape No height source; z-values interpolated for each feature from the surface before being added	Lake shoreline Obscured areas (dense forest canopy)
Replace polygons (hard or soft)	Polygon feature class	Shape geometry; x,y,z per vertex Attribute column holding one z-value for all vertices in each shape	Water bodies

Note:

It is best for the sake of terrain performance to place all hardlines together in one feature class. It is understood this might not be possible, for example, if you need to keep road and water features separate. Keep in mind, the fewer feature classes used to define a terrain, the better.

Replace SFType is used to set everything inside a polygon at a constant height. It is used mostly for lakes when there's inadvertently other data inside them, such as lidar points, whose heights are not exactly the same as the shoreline and therefore prevent the water bodies from being flat. Replace SFType does incur more processing cost than normal hard- or softlines, so it is best to avoid using it in a terrain dataset. Ideally, there should not be lidar samples in water bodies (consider adding this as a stipulation in the contract with your data provider), but if there are, you can use the Delete Terrain Points geoprocessing tool to handle them after the terrain dataset is built. Otherwise, you can eliminate any offending points before building your terrain using the Erase Point geoprocessing tool.