3d modelling of dune ecosystems using photogrammetry from remotely piloted air systems surveys

Abstract

Dune ecosystems play a key role in coastal dynamics, where a constant exchange of sediments between the sand dune and the beach exists. Thus, it is essential to measure with high precision its movements and monitor its evolution. Classical topographical surveys – even those based on real-time kinematic GPS surveys – are suitable for obtaining profiles with high accuracy, but their cost, in terms of time and personnel, make them inappropriate for large areas. Moreover, these techniques do not precisely measure volumetric changes in high-variable environments such as sand dunes. LiDAR is an active remote sensing method that reflects laser pulses off the land and records topographic positions, but LiDAR scanners are still an expensive and evolving technology. The use of Remotely Piloted Air Systems (RPAS) aims to solve this problem. RPAS map sand dunes using photogrammetric methods, such as Structure-from-Motion (SfM), to produce three-dimensional models from two-dimensional images. This methodology has the advantage of needing only a digital camera and is, therefore, cost effective. The objective of this study is to demonstrate the use of RPAS to map and model the sand dunes of Guardamar del Segura (Spain). The survey covered an extension of 226,000 m2. The results allow us to obtain a point cloud with a density of 30 points per sq m. A 3D digital elevation model of the dune has been created from the dense point cloud, with a vertical accuracy of 0.2 m Root Mean Square Error. An orthophoto was also created with a spatial resolution of 5 cm. The main conclusion is that the use of RPAS and SfM is an adequate technique to periodically monitor dune ecosystems with high accuracy and cost savings.