Quantifying the geomorphology of natural hazards: opportunities provided by image-based 3D modelling and UAVs

High-resolution Digital Elevation Models (DEMs) have proven to be very useful when analyzing geomorphological processes and their relationship with environmental controls. However, multi-temporal DEMs, which are required when volumetric changes need to be quantified, are often not available or expensive to produce. Recently, imagebased 3D modelling has been introduced as a methodology to produce high-resolution DEMs of both small and large landforms of any kind. As it is a rather cheap, flexible and rapid method, it is gaining importance in various geomorphological research areas, such as the study of natural hazards. Input data for the creation of DEMs are multiple overlapping photographs, taken with a normal camera. These can be photographs taken from a ground perspective, or taken from cameras mounted on Unmanned Aerial Vehicles (UAVs). In this study, we present two examples from gully erosion studies: (i) quantification of bank gully erosion from terrestrial photography (Bertem, Central Belgium; Frankl et al., 2014), and (ii) quantification of ephemeral gully erosion from low-altitude aerial photography (Hallines, Northern France). Processing the photographs was done from Structure from Motion – Multi-View Stereo workflows, integrated in various software such as PhotoScan used here. The accuracy of the obtained DEMs is at mm to cm-level. With the erosion features represented in 3D, erosion processes such as undercutting or soil piping can be studied. The feasibility of UAV photographic surveys in geomorphological studies is also discussed