Identifying rockfall hazards in the Fraser Canyon, British Columbia: a semi-automated approach to the classification and assessment of topographic information from airborne LiDAR and orthoimagery

Thumbnail Image
Date
Authors
Carter, Richard
Keyword
LiDAR , Rockfall , Geohazards , Remote Sensing , GIS , Landslide
Abstract
Rockfall hazards on railway corridors create risk of derailment which can result in damage to property or the environment and cause injury or loss of life. There is interest in understanding the location and severity of such hazards so that management strategies can be implemented. Ongoing collection of high resolution 3D data from terrestrial and airborne platforms has proven to be an effective medium for achieving a better understanding of the spatial-temporal, geological and geomechanical properties of rockfall hazards. The limiting factor is that such data is time consuming to collect and process and needs to be specifically directed. This study aims to provide new officelevel screening tools which can be used to identify areas where more refined data collection may offer value. Airborne LiDAR and orthoimagery for the CN Thompson Fraser Corridor is used to develop new techniques for the classification and interpretation of topographic data for the study of rockfall hazards in the area. In Chapter 3, spectral reflectance in the visible range, extracted from orthoimagery, is used to classify colourized point cloud data in the interest of identifying areas of exposed rock. In Chapter 4, this technique is paired with existing techniques for the geomorphic classification of 3D information using slope angle and use as the main input to a probabilistic rockfall hazard assessment. Analysis of colour values, paired with conventional morphometric analysis using slope angle, shows promise as a means of classifying topographic data into geomorphic domains. The study successfully derives all inputs for the probabilistic assessment of rockfall hazard from the above mentioned airborne LiDAR and orthoimagery dataset.
External DOI