Coincident beach surveys using UAS, vehicle mounted and airborne laser scanner: Point cloud inter-comparison and effects of surface type heterogeneity on elevation accuracies

REMOTE SENSING OF ENVIRONMENT

卷:208

Article

Elsner, Paul¹  Dornbusch, Uwe²  Thomas, Ian³  Amos, Dan⁴  Bovington, James⁴  Horn, Diane¹ 

[1] Birkbeck Univ London, Dept Geog Environm & Dev Studies, Motet St, London WC1E 7HX, England

[2] Environm Agcy, Worthing BN11 1LD, England

[3] Pevensey Coastal Def Ltd, Environm Agcy Dept, Coast Rd, Pevensey Bay BN24 6ND, E Sussex, England

[4] Adur & Worthing Councils, Strateg Reg Coastal Monitoring Programme, Worthing Town Hall,Chapel Rd, Worthing BN11 1HA, W Sussex, England

关键词: UAS;    UAV;    Laser scanning;    Beach monitoring;    Accuracy analysis;    LiDAR;   

DOI  :  10.1016/j.rse.2018.02.008

来源: Elsevier

【 摘 要 】

Reliable and consistent topographic data is key to a multitude of environmental manangement and research applications. Unmanned Aerial Systems WAS) are fast establishing themselves as a promising additional remote sensing platform that provides high spatial resolution not only of topography but also surface types and coastal features together with comparatively low costs and high deployment flexibility. However, comprehensive information on the accuracy of UAS-based elevation models in comparison to other available surveying methodology is regulary limited to be referenced to individual methods. This paper addresses this shortcoming by comparing coincident beach surveys of three different point cloud generating methods: ATV mounted mobile laser scan (MLS), airborne LiDAR (ALS), and UAS. This was complemented by two RTK-GPS surveys on a pole with wheel attachment and mounted on an ATV. We present results in relation to elevation accuracies on a concrete control surface, the entire beach and for six different beach surface types together with how differences between point clouds propagate during the construction of gridded elevation models. Overall, UAS point cloud elevations were consistently higher than those of ALS (+ 0.063 m) and MLS (+ 0.087 m). However, these results for the entire beach mask larger and smaller differences related to the individual surface characteristics. For all surface types, UAS records higher (from 0.006 m for wet sand to 0.118 m for cobbles, average of 0.063 m) elevations than ALS. The MLS on the other hand, records predominantly lower elevation than ALS (- 0.005 m for beach gravel to - 0.089 m for soft mud, average of - 0.025 m) except for cobbles, where elevations are 0.056 m higher. The comparison shows that all point cloud methods produce elevations that are suitable for monitoring changes in beach topography in the context of operational coastal management applications. However, due to the systematic differences between respective monitoring approaches, care needs to be taken when analysing beach topographies for the same area based on different methods. The eventual choice of monitoring method is therefore guided by a range of practical factors, including capital cost of the system and operating costs per survey area, conditions under which the system can operate, data processing time, and legal restrictions in the use of the system such as air safety regulations or limitation of ground access to areas with environmental protection.

【 授权许可】

Free