Ideally, a separate LIDAR dataset would be acquired for each posting density under investigation.  This would involve holding as many sensor variables constant as possible (such as scan angle and pulse rate) and changing only one variable (such as altitude) as needed.  This project is taking advantage of an existing high spatial density LIDAR dataset for a study area in Guilford County, North Carolina.  The original posting density for these data is approximately 1m × 1m.  As shown in a LIDAR processing diagram, this original dataset will be systematically decimated by a NASA subcontractor to simulate an array of posting densities.  Bald earth data in various post-density spacings (e.g., 1 m × 1m, 5m × 5m) will be generated.  For each post-density spacing, the LIDAR points will be subjected to the same automated/visual analysis stages typically performed for identifying “ground” returns from over-story returns.

The North Carolina Geodetic Survey (NCGS) will collect in-situ­ positional references for each of the land cover categories in a schema (to be specified by the State of North Carolina).  USC-ARC will collect additional in-situ reference information required for understanding the role of land cover, surface form, and coincident zoning errors.  These references will be used for the development of empirical relationships between land cover categories, LIDAR posting densities, and surface form.  USC-ARC and NCGS will work together in determining the best locations for reference points so that the land cover and surface form parameters are sufficiently represented.

Concurrent with in-situ referencing, flood risk analysis (H&H modeling) will be performed by Dewberry & Davis in conjunction with NCGS.  Each posting density will be modeled separately.  USC-ARC will then analyze the results of both in-situ references and the flood risk maps for each of the posting densities.  The relationship between the in-situ references, the TINs, and the H&H modeling is described in a DEM processing diagram.