DETERMINING THE APPLICABILITY OF CHANGE DETECTION FOR QUANTIFICATION OF EROSION ON THE NIAGARA ESCARPMENT, HAMILTON, ONTARIO, CANADA

The Niagara Escarpment is a steep faced cuesta composed of Ordovician and Silurian sedimentary rocks that traverses southern Ontario. In Hamilton, the Niagara Escarpment divides the city into the lower city and the upper city and agricultural areas. This division of the city is bridged by 19 access routes that cross the escarpment. Recent activity of the rock face, including major rockfalls onto the roads, has raised major concerns about the stability of the escarpment. To address these concerns and to provide more information on erosional processes active along the escarpment, a quantitative study to determine the rate of erosion using change detection methodologies was conducted. Using an unmanned aerial vehicle (UAV), imagery of a selected rock face on the escarpment was collected and used in a Structure from Motion (SfM) workflow to create detailed 3D models. Following this first survey, blocks of a known size were removed from the site and a second survey was conducted. Agisoft Metashape software was used to create point clouds of the rock face from the two surveys. These were imported into CloudCompare, aligned, and the M3C2 algorithm was used to determine the distance of change detected from the removal of the blocks. The distances computed using the M3C2 algorithm were also used to estimate the volume of change which was compared to the known volume of the rock samples removed from the outcrop. The results of these calculations indicated that most removed blocks were detectable using M3C2, with a single small sample being missed. However, the volume of the change estimates was usually smaller than the measured volume of the removed blocks by 30 – 50%. This study demonstrates that the method does provide a good estimate of the location of eroded blocks from the rock face, but volume estimates show substantial inaccuracy and should be used only as a guide for estimating minimal amounts of volume change that may have occurred.