USING THE NATIONAL ELEVATION DATASET: DOES COST EQUAL ACCURACY?

摘要

The best (most accurate and newest) digital elevation model (DEM) available should be used for any hydraulic analysis including a dam breach model. In many urbanized areas, the best available DEM might be a publically produced, light detection and raging (LiDAR) generated DEM obtainable from the local municipality, county or state. However, in most undeveloped and rural areas, LiDAR-generated data is not readily available and is generally cost prohibitive to obtain for a single dam breach project. In such a case, the default elevation model would first be the USGS 1/9-arc-second National Elevation Dataset (NED) (with about 17% coverage in the contiguous states), followed by the 1/3-arc-second NED (with 100% coverage). This paper evaluates the differences in the modeled breach zones obtained using a county-wide, LiDAR-generated DEM, and the USGS 1/3-arc-second NED, and concludes if using the USGS 1/3-arc-second NED is good enough for a dam breach model. The 30-foot grid size of the 1/3-arc-second NED is too coarse to show any channel features; therefore, as part of the evaluation of the applicability of the 1/3-arc-second NED, it is also necessary to understand the sensitivity of model channelization and channel smoothing (eliminating the run-riffle-pool cycle) on the final breach zone.