3D mapping efficacy of a drone and terrestrial laser scanner over a temperate beach-dune zone

摘要

Understanding dynamic earth surface processes requires various spatial and temporal information to help produce patterns of landform change. Recent developments in sensor technology such as Structure from Motion (SfM), camera-mounted airborne Unmanned Aerial Vehicles (UAVs) and Terrestrial Laser Scanning (us) have provided a means of acquiring high-resolution spatial data on land surface topography. Through repeat surveys, these techniques enable much better understanding of what is termed 'geomorphometry', where we can examine a geomorphic surface for change over space and time. In coastal environments, change can involve significant alteration and generation of landforms over relatively short periods and, therefore, we require a means of measuring surface morphology quickly and over large areas. Here, we examine a section of a beach-dune system in NW Ireland using SfM-UAV and TLS plus baseline dGPS data points to assess the value of these techniques and to understand their effectiveness (and limitations). Issues such as accuracy, resolution and differences of Digital Elevation Models (DEMs) are assessed for their efficiency, associated challenges and relative performance over variations in terrain types and analytical approaches. We also examine the implications for differences in areal and volume calculations of the coastal landforms using both approaches. We find that sensor performance is highly dependent on the terrain being measured, with undulations, slope, vegetation cover, acquisition resolution (point density) and interpolation issues all having pronounced impacts on effectiveness and data quality. In general, the TLS performed better over flatter, low-angled topography containing sparseon-vegetated areas than in areas with complex landforms where survey shadows appear to compromise final DEMs. The SfM-UAV shows good performance over different terrains with the exception of relatively flat, featureless areas such as sandy beaches and densely-vegetated surfaces where differences between techniques are 1 m. Data acquisition however is much (x30) faster using a SfM-UAV with more extensive survey areas covered than using a TLS. (C) 2018 Elsevier B.V. All rights reserved.