Authors: Christopher Scarpone*, Ryerson University, Environmental Applied Science and Management, Andrew A Millward, Ryerson University, Geography and Environmental Studies
Topics: Environmental Science, Quantitative Methods, Remote Sensing
Keywords: urban forest, ecological restoration, machine learning, UAV, drone, LiDAR
Session Type: Paper
Start / End Time: 10:00 AM / 11:40 AM
Room: Bourbon Room, Astor, Mezzanine
Tommy Thompson Park (TTP), along Toronto’s waterfront, is a spit of anthropogenic origin that has evolved into a vibrant ecosystem rich in flora and fauna. TTP is home to a population of double-crested cormorants (DCC) Phalacrocorax auritus believed to be the largest colony within North America. Nutrient enrichment from DCC guano, coupled with destructive nesting habits (bark stripping, leaf harvesting) have caused rapid decline and mortality of early successional urban forest, as well as alterations to soil properties. So as to maintain ecological integrity in TTP, a restoration plan is required where a precise representation of the scalar ecological processes is critical for understanding the immediate and long-term effects of DCC nesting. Quantification of current ecological conditions is integral to enhancement of the prediction capabilities of modern machine learning techniques. Leveraging machine learning approaches in ecology requires spatial data that can complement traditional, plot-based, measurements. Advancements in drone and LiDAR technology have allowed for small spatial-scale projects to be acquire rich spatial and spectral datasets with modest budgets. This study utilises drone-acquired, 2 cm ground resolution, imagery that covers the RGB and NIR spectrums along with complementary LiDAR data. Comparisons of topographic, soil and vegetative indices, along a continuum of spatial resolutions (i.e., using pixel resampling), were conducted to investigate the comparative value of higher spatial-resolution data. Discussions of acquisition and processing of newer forms of data are explored in an effort to determine the potential importance of drone- and LiDAR-based data for urban forestry and ecological restoration planning.