As the size of projects grow there are increasing challenges associated with management and analysis of terrain and elevation data, such as DTMs, DEMs and canopy models.

Utilising ESRI’s new Terrain data type available in ArcGIS 9.2, AAM has created Terrain datasets comprising billions of LiDAR points. A recent large project totalled 1.3 billion points alone - over an area of 11,700 sq km. A new project currently underway is expected to total over 3.5 billion LiDAR points.

For ESRI Users who have not yet adopted the Terrain types, elevation data is typically supplied as ASCII Point, ESRI Shapefile Point or ESRI TIN. Some hindrances of these datasets for large projects have been found to include:
 - time consuming to create, load and analyse
 - large storage requirements
       - 1G ASCII = 2.5G ESRI Shapefiles = 2.5G ESRI TIN
 - tile-based approach
       - Generally 2x2 km tiles to limit number of points per file
 - file size recommendations are restricting
       - < 3 million points per shapefile
       - < 12 million points per TIN

ESRI Terrains are multi-resolution TIN-based surfaces designed for large areal extents, where scalability is achieved using Terrain pyramids and the multipoint geometry type. Data sources may include:
 - mass point collections – LiDAR, SONAR, photogrammetric (stored as multipoint)
 - breaklines – stream networks and roads derived from stereo photographs
 - boundaries – internal water, coastlines and survey area extents

These thematic data types and the resultant terrain are all stored in a geodatabase, inside a feature dataset. All ESRI geodatabase types are supported, choice of which is dependent on file size limits, see recommendations below:
 - personal GDB < 2GB (~20 million points)
 - filebased GDB < 200 million points
 - SDE > 200 million points (limited by underlying database)

Some benefits found by employing the ESRI Terrain type include:
 - computation time reduced considerably
 - low disk usage
       - 1G ASCII = 200MB Multipoint = Negligible Terrain
 - seamless dataset
 - improved performance for analysis and display
 - partial editing capabilities.

One limitation found is the difficulty to return single XYZ coordinate values from the source LiDAR points, due to the nature of multipoint storage as a “BLOB.” However, workarounds exist and the immense benefits of the Terrain prevail!

The ESRI Terrain data type has significantly improved access, display, performance and analysis of LiDAR data and the elevation information it represents.