Extracting photogrammetric ground control from lidar DEMs for change detection

Research has shown the importance of measuring topography for surface change detection and that the use of remote sensing methods is ideal for this application. A prerequisite for measuring change is a historical data-set that covers the time period of interest. Many remote methods of topographic data collection such as lidar are relatively recent developments and therefore cannot provide records longer than about a decade. Alternatively, aerial photography has been in common use since the early 20th century and archives upwards of 50 years are therefore not uncommon. While photogrammetry is dependent on well distributed, high quality ground control points (GCPs), such data has been successfully applied retroactively. Therefore, contemporary lidar data, which requires little ground-truth data, should be an ideal source of GCPs for controlling historical aerial photographs. This study aims to evaluate the success with which GCPs can be extracted from a high-resolution lidar data-set for controlling aerial photography for digital elevation model (DEM) production. Using a data-set collected in Upper Wharfedale, northern England, GCPs were measured both by means of a lidar data-set and by using traditional field-based methods. The results showed that while the use of lidar-derived ground control produced a DEM of inferior quality, increasing the number of GCPs used in the model produced results comparable to the GPS-controlled DEM. These results are significant especially for surface change detection research in remote areas where high quality ground control is difficult to secure.