The impact of varied nominal posting density LIDAR data on DEM accuracy, hydraulic modeling and flood zone delineation

LIDAR data has become a primary source of digital terrain information for use in hydraulic modeling and flood mapping. However, the accuracy of digital elevation models (DEMs) derived from such data has not been widely documented under various conditions. Therefore the factors that influence this accuracy are only generally understood. Based on established relationships between sampling intensity and error, nominal posting density, or the average distance between collected points, likely contributes significantly to the error budget as areas between each ground post must be interpolated. Nominal posting density is also a major cost determining factor of LIDAR data collection. Often more flight time is required as higher posting densities usually require some combination of the following: (a) lower altitude, (b) additional flight lines, or (c) a higher pulse rate. In addition to data collection requirements, the additional data volume incurred through a higher posting density requires extra storage capacity, as well as substantially greater computer and manual processing time. This research presents a methodology for establishing a relationship between nominal posting density and its effects on hydraulic modeling for flood zone delineation. LIDAR data collected at a high posting density (∼1 to 2 m) over a study area in North Carolina piedmont was systematically decimated to simulate datasets with sequentially lower posting densities. Using extensive first-order ground survey information the relative accuracy of each DEM derived from these LIDAR datasets was assessed and reported. A series of hydraulic analyses was performed utilizing standard engineering practices and modeling software (HEC-RAS). All input variables were held constant in each model run except the topographic information from the decimated LIDAR datasets. The results were compared to a hydraulic analysis performed on the un-decimated reference dataset. The sensitivity of the primary model outputs (i.e., base flood elevation and flood zone boundary) to the variation in nominal posting density is reported. The results indicate that although general error patterns are apparent visually they are not statistically significant over the posting densities tested. The flood zone extent is found to be sensitive to variations in posting density.