Kinematic analysis of sea cliff stability using UAV
photogrammetry
Article (Accepted Version)
http://sro.sussex.ac.uk
Barlow, John, Gilham, Jamie Mark and Ibarra Cofrã, Ignacio (2017) Kinematic analysis of sea cliff
stability using UAV photogrammetry. International Journal of Remote Sensing, 38 (8-10). pp.
2464-2479. ISSN 0143-1161
This version is available from Sussex Research Online: http://sro.sussex.ac.uk/id/eprint/66174/
This document is made available in accordance with publisher policies and may differ from the
published version or from the version of record. If you wish to cite this item you are advised to
consult the publisher’s version. Please see the URL above for details on accessing the published
version.
Copyright and reuse:
Sussex Research Online is a digital repository of the research output of the University.
Copyright and all moral rights to the version of the paper presented here belong to the individual
author(s) and/or other copyright owners. To the extent reasonable and practicable, the material
made available in SRO has been checked for eligibility before being made available.
Copies of full text items generally can be reproduced, displayed or performed and given to third
parties in any format or medium for personal research or study, educational, or not-for-profit
purposes without prior permission or charge, provided that the authors, title and full bibliographic
details are credited, a hyperlink and/or URL is given for the original metadata page and the
content is not changed in any way.
Kinematic analysis of sea cliff stability using UAV photogrammetry
1
John Barlow
2
Department of Geography, University of Sussex, Brighton UK, BN1 9QJ
3
Corresponding author: (E) john.barlow@sussex.ac.uk (T) +44 01273873622
4
Jamie Gilham
5
Department of Geography, University of Sussex, Brighton UK, BN1 9QJ
6
(E) J.Gilham@sussex.ac.uk
7
Ignacio Ibarra Cofrã
8
Department of Geography, University of Sussex, Brighton UK, BN1 9QJ
9
(E) ii43@sussex.ac.uk
10
11
Kinematic analysis of sea cliff stability using UAV photogrammetry
12
Abstract
13
Erosion and slope instability poses a significant hazard to communities and
14
infrastructure located is coastal areas. We use point cloud and spectral data derived from close
15
range digital photogrammetry to perform a kinematic analysis of chalk sea cliffs located at
16
Telscombe, UK. Our data were captured from an unmanned aerial vehicle (UAV) and cover a
17
cliff face that is about 750 m long and ranges from 20 to 49 m in height. The resulting point
18
clouds had an average density of 354 points m
-2
. The models fitted our ground control network
19
within a standard error of 0.03 m. Structural features such as joints, bedding planes, and faults
20
were manually mapped and are consistent with results from other studies that have been
21
conducted using direct measurement in the field. These data were then used to assess differing
22
modes of failure at the site. Our results indicate that wedge failure is by far the most likely
23
mode of slope instability. A large wedge failure occurred at the site during the period of study
24
supporting our analysis. Volumetric analysis of this failure through a comparison of sequential
25
models indicates a failure volume of about 160 m
3
. Our results show that data capture through
26
UAV photogrammetry can provide a useful basis for slope stability analysis over long sections
27
of coast. This technology offers significant benefits in equipment costs and field time over
28
existing methods.
29
Keywords: UAV, photogrammetry, landslide, sea cliffs
30
1. Introduction
31
In recent years, digital surface model acquisition had been dominated by the use of
32
airborne and terrestrial light detection and ranging (LiDAR) (Haala and Rothermel, 2012;
33
Gonçalves and Henriques, 2015). However, the emergence of unmanned aerial vehicles (UAVs)
34
or small Unmanned Aircraft Systems (sUAS) alongside the proliferation of inexpensive digital
35
cameras and various software platforms for processing of this data (Hugenholtz et al., 2013)
36
have provided a method of data capture which can achieve results of comparable accuracy
37
whilst significantly reducing costs and data capture time (Slatton et al., 2007; Remondino et
38
al., 2011; Hugenholtz et al., 2013). The relative affordability of these various systems, when
39
compared to airborne LiDAR and terrestrial laser scanning (TLS) surveys, has led to a
40
significant rise in procurement for a diverse range of research applications (Dunford et al., 2009;
41
Rango et al., 2009; Jaakkola et al., 2010; Lin et al., 2011; Stefanik et al., 2011; Hugenholtz et
42
al., 2012; Hugenholtz et al., 2013; Colomina and Molina, 2014).
43
With regard to sea cliffs, high precision monitoring of erosion has typically been
44
undertaken using TLS or a combination of TLS and terrestrial digital photogrammetry (e.g.
45
Rosser et al. 2005; Lim et al. 2010; James and Robson 2012; Barlow et al. 2012; Martino and
46
Mazzanti, 2014). Although these studies have provided improved control over the rate and
47
processes of coastal cliff recession compared to those based on historical maps and aerial
48
photographs (e.g. Dornbusch et al. 2008), the spatial extent of high precision monitoring has
49
typically been limited by the logistics of terrestrial data collection from shore platforms (e.g.
50
Rosser et al. 2005; Martino and Mazzanti, 2014). This research demonstrates the use of UAV
51
photogrammetry to produce data of similar characteristics to that derived from TLS for sea cliff
52
research. The rapid nature of data capture using UAVs means that much longer sections of cliff
53
can be surveyed in much shorter periods of time than with previous studies.
54
The vast majority of digital photogrammetric applications using UAVs have relied on
55
pure aerial triangulation or indirect sensor orientation (Colomina and Molina, 2014), whereby
56
the navigation system and software determine an approximation of the image location and
57
generate tie points with associated measurements in space which are later compared to the
58
known GCPs. Rehak et al. (2013) report that the application of this method can produce models
59
which are sub-decimetre in accuracy as reported by component or overall three dimensional
60
standard error (3DSE) values. Many of these studies rely on the structure from motion (SfM)
61
method in which the bundle adjustment is based on features automatically extracted from
62
multiple overlapping images (Westoby et al. 2012). SfM requires convergent imagery taken
63
from multiple ranges in order to produce the best possible accuracy (James and Robinson,
64
2014). The method is therefore not well suited to coastal cliff research in that it requires
65
multiple passes in order to photograph each section of cliff from multiple angles and ranges.
66
Given the limited endurance of micro UAVs, the use of more traditional photogrammetry
67
involving a calibrated camera and strip photography maximises the length of cliff that can be
68
surveyed and this is the method that has been adopted for the current research.
69
The structural geology of cliffs often provides primary control over the type of slope
70
failure that may occur. This is because the uniaxial compressive and shear strength of
71
penetrative discontinuities are generally lower than those of the intact rock (Terzaghi, 1962).
72
Kinematic analysis involves mapping the orientations of penetrative discontinuities within a
73
rock slope in order to identify those that are oriented unfavourably for slope stability given the
74
shear strength along discontinuity surfaces (Richards et al. 1978; Hoek and Bray, 1981; Wyllie
75
and Mah, 2004). The analysis is based solely on the geometric conditions of rock slopes such
76
that it does not locate discontinuities in space, gives no reference of their size, and does not
77
consider the influence of hydrogeological or seismic boundary conditions on slope stability
78
(Hoek and Bray, 1981). The assessment of the orientation of structural geology data requires
79
plotting poles on a stereonet which indicates the dip and dip direction of discontinuities. This
80
is executed with the aim of identifying clusters or sets of discontinuities, for which average dip
81
and dip direction can be determined. The use of dense point clouds derived from close-range
82
digital photogrammetry and terrestrial laser scanning in the characterization of rock slope
83
morphology is well established within the research literature and has the advantage of allowing
84
measurements to be taken from the entire rock surface rather than only those sections accessible
85
to manual measurement (e.g. Lato et al. 2009; Sturzenegger and Stead, 2009a; Lim et al. 2010;
86
