Musings on DEMs

 Lidar point clouds can create amazing DEMs, and are increasingly the source data for DEMs from national mapping agencies. USGS 1/3 arc second DEM covering Annapolis.  This is a DTM, but roads, bridges, and cloverleafs create many anomalies.  The arrow points out the most egregious, half of a building.  Guth (2018a) pointed out that castles, or indeed any large buildings, create challenges for DTM creation. Short of lots of editing, are there better alternatives?  Three different DEMs can be created automatically from a lidar point cloud (Guth and others, 2021). DSM, or digital surface models, selects the highest return in each pixel of the DEM, or the average of the first returns in each pixel.  For a small pixel size, the choice does not make much difference. NVS, or non-vegetated surface, selects the lowest return in each pixel of the DEM.  If the density of the lidar survey is sufficient, trees and power lines will be removed, leaving the ground an...

  This morning's tide prediction called for 3.1 feet above MLLW for 5 PM. ( NOAA ) This is my tide model for the 5 roads most likely to flood.  It is created as follows: Get the 2017 county lidar survey. Create an NVS, non-vegetated surface, at 0.5 m and 1.0 m resolution, which will be in meters on NAVD88. The NVS removes the trees, but cars remain. Get the TIGER roads, and extract the 5 roads that control access. Interpolate the roads to 1 m spacing. Compute the elevation at each point on the road in the TIGER extract from the lidar NVS. Convert the elevations to feet, and adjust to MLLW with the offset for the tide gauge. Find the minimum elevation on each road. Get the OpenStreetMap data, and repeat the steps for the TIGER data.  The two data sets do not have exactly the same locations. If the water reached 3.1 feet MLLW, all of these roads would be flooding or just starting to flood in the case of Bowyer Road. The locations that would be flooding at 3.1 feet, with col...