ABSTRACT
Availability of accurate high-resolution digital surface models (DSMs) within coastal environments has for many years been the realm of expensive airborne survey techniques. This project aims to utilise the Unmanned Aerial Vehicle (UAV) approach to surveying a coastal dune system that is rapid, can be flown autonomously, collects a large dataset, is cost-effective, and improves spatial accuracy compared to traditional methods. Structure-from-Motion (SfM) software is employed to process the imagery collected from two surveys, creating a three-dimensional (3D) dense cloud totalling 68.6 million points, production of a DSM, a high-quality two-dimensional (2D) orthomosaic image, a 3D mesh model, and an average ground sampling distance (GSD) of 2.66 cm/pixel or 7.02 cm2. With increasing elevation uncertainties within the coastal margin, comparisons made between the UAV-SfM DSM and a 2012 Ordnance Survey DSM, Light Detection and Ranging (LIDAR) DSMs from 2012 and 2014 find on average a 2.36 m discrepancy in fixed elevation points between datasets. Surveying methods include a Real-Time Kinematic Global Navigation Satellite System (RTK GNSS) to validate 35 ground control points (GCPs) used for SfM georectification, producing accuracy tolerances of 0.12 m in X, 0.09 m in Y, and 0.06 m in Z.
