Terrestrial lidar data from northern Monterey Bay, California, September 2015

Terrestrial lidar data was collected between September 28 and 29, 2015 using a tripod-mounted Riegl VZ-1000 lidar scanner. For each area surveyed, the scanner was placed at several positions which were selected to provide maximum line-of-sight coverage of the area of interest. Scans were typically conducted in panoramic mode, creating a detailed point cloud of all unobstructed surfaces in a 360 degree view of the scanner. Each individual scan typically contains 20 million or more precise point measurements. At each scan position, co-registered photographic imagery was also collected with a scanner mounted DSLR camera. Scanner registration was performed by placing four or more cylindrical or flat reflective tripod-mounted targets in view of the scanner. The positions of these reflectors were measured using survey grade GNSS, and used in post-processing to register each point cloud into a projected coordinate system.

Initial data processing of the terrestrial lidar data was conducted using Riegl RiscanPro software. Each scan position and point cloud was registered using the GNSS-measured positions of the tripod-mounted reflectors. The quality of the registration was visually compared against adjacent registered point clouds, and concurrently collected topographic GNSS data. In the unlikely case that initial registration appeared to be inaccurate either due to poor GNSS precision, or poor reflector geometry, a point cloud fitting process called ?Multi-Station Adjustment (MSA)? was used to adjust the spatial registration to match adjacent point clouds. After registration was completed, the point clouds were colorized using RGB pixel values from the co-registered photographic imagery. The point clouds were then hand-edited to remove erroneous data points below the true ground surface caused by reflections on wet sand or water. Additional noise from water, breaking surf, haze, and objects within the point cloud were also removed by hand editing. The point clouds were then exported to comma delimited (.csv) format and converted into LAS format and processed using the LAStools software suite. A batch script was used to: 1) clip the LAS files to the areas of interest, 2) perform a range filter to remove all points further than 700 meters away from the scanner, 3) thin the point cloud to an average spacing of 5 centimeters to reduce point density, and 4) perform a classification algorithm to detect ground points and remove objects and vegetation from the point cloud. At this point a final visual check on the remaining points was performed to verify and correct any misclassification of ground surfaces. To produce the final output files all classified ground points were thinned using the lowest elevation point in 10 centimeter cells, and the resulting points were exported to a comma delimited (.csv) format.

Edited metadata to add keywords section with USGS persistent identifier as theme keyword. No data were changed.