PE and RS PUBLIC October 2011 : Page-964

Graduate student Anna Cohen (University of Washington) surveying at Sacapu Angamuco, Michaocán, Mexico. 2008; Millard et al ., 2009; Challis et al. , 2011; Chase et al ., 2011), previous projects have used lidar either as a test case to see if known architectural remains could be identifi ed, or over a large landscape area to identify fairly large features that were believed to have existed. Lidar had never been used to survey an area similar to the malpais that was believed to be densely covered with architectural remains. Acquiring lidar meant investing funds in a technology that, as far as we could determine, had not been used in this way before. L IDAR P ROJECT S COPE Lidar was collected over an eight-square-kilometer area of the malpaís that encompassed the major core of the ancient settlement with appropriate archaeological permits from the Instituto Nacional de Antropología e Historia . Data collection specifi cations defi ned a density of 16 points per square meter and processing of the data to develop both LAS and ASCII fi les. Two base stations and four ground control points were set for the fl ight. Data was collected to conform to the National Standard for Spatial Data Accuracy (NSSDA), Achievable Contour interval of 0.3m but with the mission’s Lidar mapping abilities as compared to the data collected by GPS ground survey data, results allowed for greater than 0.3m contour intervals. The lidar fl ight plan called for nine passes made along an east-west axis and eight passes made along a north-south axis. The instrument was fl own at an average altitude of 3810 m above mean sea level. The lidar point cloud data was classifi ed into two classes by Merrick for delivery to the Legacies Project — ground and unclassifi ed. Upon receiving the data, Legacies researchers examined the classifi ed point cloud using MARS ® software, developed and provided by Merrick, and recognized that the ”ground” class had many gaps in it. In discussions, it became clear that the ground class in MARS ® displays the points needed to produce a digital terrain model (DTM), a model of the Earth’s surface stripped of all vegetation and man-made structures. The team was very excited when it was determined that the ”gaps” in the ground class were actually ancient architectural remains such as a wall, a foundation, or a house mound. The next step in processing was to use MARS ® software to fi ll in those gaps. This was accomplished by creating a custom fi lter to include not just points from the point cloud that were considered ground, but also points that were refl ected from the architectural remains. In order to do this the team experimented with the MARS ® fi ltering capacity. Through systematic trial focusing on areas where architectural remains had been surveyed during the previous two years, it was discovered that most of the lidar returns between the ground and 1.25 m above ground continued from page 963 computers connected to Trimble ® GeoXH™ and GeoXT™ GPS receivers, was highly accurate but incredibly time-consuming. In two three-month fi eld seasons the project was able to document only over 2,500 architectural features found within 2.5 square km. This was a monumental effort, and in many respects groundbreaking for the use of the mapping technology, but it was quickly realized that at this pace it might take upwards of a decade to document the architecture of the entire site. It was decided to obtain and integrate lidar into the project to increase the speed with which the ground survey could be done. The decision to integrate lidar was not taken lightly. While lidar has been used since the early 2000s on archaeological projects (see, for example, Holden et al., 2000; Rowlands and Sarris, 2007; Challis et al. , 2008; Gallagher and Josephs, Graduate students Jason Bush (Colorado State University) and Andrea Torvinen (Arizona State University) standing on a prehistoric structure at Sacapu Angamuco, Michoacán, Mexico. 964 Oct ober 2011 Photogrammetric engineering & remote SenSing

Previous Page  Next Page


Publication List
Using a screen reader? Click Here
Using a screen reader? Click Here