Nat. Hazards Earth Syst. Sci.

, 18, 1055–1071, 2018

https://doi.org/10.5194/nhess-18-1055-2018
© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Combination of UAV and terrestrial photogrammetry to

assess rapid glacier evolution and map glacier hazards
Davide Fugazza1 , Marco Scaioni2 , Manuel Corti2 , Carlo D’Agata3 , Roberto Sergio Azzoni3 , Massimo Cernuschi4 ,
Claudio Smiraglia1 , and Guglielmina Adele Diolaiuti3
1 Department of Earth Sciences “A. Desio”, Università degli Studi di Milano, 20133 Milan, Italy
2 Department of Architecture, Built Environment and Construction Engineering, Politecnico di Milano, 20133 Milan, Italy
3 Department of Environmental Science And Policy, Università degli Studi di Milano, 20133 Milan, Italy
4 Agricola 2000 S.C.P.A., 20067 Tribiano (MI), Italy

Correspondence: Davide Fugazza ([email protected]) and Marco Scaioni ([email protected])

Received: 1 June 2017 – Discussion started: 6 June 2017

Revised: 6 March 2018 – Accepted: 10 March 2018 – Published: 5 April 2018

Abstract. Tourists and hikers visiting glaciers all year round 1 Introduction
face hazards such as sudden terminus collapses, typical of
such a dynamically evolving environment. In this study, we Glacier- and permafrost-related hazards can be a serious
analyzed the potential of different survey techniques to an- threat to humans and infrastructure in high mountain regions
alyze hazards of the Forni Glacier, an important geosite lo- (Carey et al., 2014). The most catastrophic cryospheric haz-
cated in Stelvio Park (Italian Alps). We carried out surveys ards are generally related to water outbursts, either through
in the 2016 ablation season and compared point clouds gen- breaching of moraine- or ice-dammed lakes or from the
erated from an unmanned aerial vehicle (UAV) survey, close- englacial or subglacial system, causing floods and debris
range photogrammetry and terrestrial laser scanning (TLS). flows. Ice avalanches from hanging glaciers (Vincent et al.,
To investigate the evolution of glacier hazards and evalu- 2015) and debris flows caused by the mobilization of accu-
ate the glacier thinning rate, we also used UAV data col- mulated loose sediment on steep slopes (Kaab et al., 2005a)
lected in 2014 and a digital elevation model (DEM) cre- can also have serious consequences for downstream popu-
ated from an aerial photogrammetric survey of 2007. We lations. Less severe hazards, but still particularly threaten-
found that the integration between terrestrial and UAV pho- ing for mountaineers, are the detachment of seracs (Riccardi
togrammetry is ideal for mapping hazards related to the et al., 2010) or the collapse of ice cavities (Gagliardini et
glacier collapse, while TLS is affected by occlusions and al., 2011; Azzoni et al., 2017). While these processes are in
is logistically complex in glacial terrain. Photogrammetric part typical of glacial and periglacial environments, there is
techniques can therefore replace TLS for glacier studies evidence that climate change is increasing the likelihood of
and UAV-based DEMs hold potential for becoming a stan- specific hazards (Kaab et al., 2005a). In the European Alps,
dard tool in the investigation of glacier thickness changes. accelerated formation and growth of proglacial moraine-
Based on our data sets, an increase in the size of collapses dammed lakes has been reported in Switzerland, amongst
was found over the study period, and the glacier thinning concern of possible overtopping of moraine dams provoked
rates went from 4.55 ± 0.24 m a−1 between 2007 and 2014 by ice avalanches (Gobiet et al., 2014). Ice avalanches them-
to 5.20 ± 1.11 m a−1 between 2014 and 2016. selves can be more frequent as basal sliding is enhanced by
the abundance of meltwater in warmer summers (Clague,
2013). Glacier and permafrost retreat, which has been re-
ported in all sectors of the Alps (Smiraglia et al., 2015; Fis-
cher et al., 2014; Gardent et al., 2014; Harris et al., 2009), is a
major cause of slope instabilities, which can result in debris
flows by debuttressing rock and debris flanks and promot-

Published by Copernicus Publications on behalf of the European Geosciences Union.

1056 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

ing the exposure of unconsolidated and ice-cored sediments In this study, we investigated a rapidly downwasting
(Keiler et al., 2010; Chiarle et al., 2007). Glacier downwast- glacier (almost 5 m a−1 water equivalent; Senese et al., 2012)
ing causes changes in water resources, with an initial increase in a protected area and highly touristic sector of the Ital-
in discharge due to enhanced melt followed by a long-term ian Alps, Stelvio National Park. We focused on the glacier
reduction, affecting drinking water supply, irrigation and hy- terminus and the hazards identified there, i.e., the formation
dropower production (Kaab et al., 2005b), along with a rising of normal faults and ring faults. The former occur mainly
occurrence of structural collapses (Azzoni et al., 2017). Fi- on the medial moraines and glacier terminus and are due to
nally, glacier retreat and the increase in glacier hazards both gravitational collapse of debris-laden slopes. The latter de-
negatively influence the tourism sector and the economic velop as a series of circular or semicircular fractures with
prosperity of high mountain regions (Palomo, 2017). stepwise subsidence, caused by englacial or subglacial melt-
The growing threat from cryospheric hazards under cli- water creating voids at the ice–bedrock interface, eventually
mate change calls for the adoption of mitigation strategies. leading to the collapse of the cavity roofs. While often over-
Remote sensing has long been recognized as an important looked, these collapse structures are particularly hazardous
tool for producing supporting data for this purpose, such as for mountaineers and they are likely to increase under a cli-
digital elevation models (DEMs) and multispectral images. mate change scenario (Azzoni et al., 2017). They are more
DEMs are particularly useful for detecting glacier thickness dangerous than crevasses because of their larger size.
and volume variations (Fischer et al., 2015; Berthier et al., We conducted our first UAV survey of the glacier in 2014;
2016) and for identifying steep areas that are most prone to in the summer of 2016, the glacier was surveyed using three
geomorphodynamic changes, such as mass movements (Bla- different techniques for the generation of point clouds, DEMs
sone et al., 2015). Multispectral images at a sufficient spatial and orthophotos. The aims were (1) to compare the different
resolution make it possible to recognize most cryospheric methods and select the most appropriate ones for monitor-
hazards (Quincey et al., 2005; Kaab et al., 2005b). While ing glacier hazards; (2) to identify glacier-related hazards and
satellite images from Landsat and ASTER sensors (15–30 m their evolution between 2014 and 2016; and (3) to investigate
ground sample distance – GSD) are practical for regional- changes in ice thickness between 2014 and 2016 and between
scale mapping (Rounce et al., 2017), the assessment of haz- 2007 and 2016 by comparing the two UAV DEMs and a third
ards at the scale of individual glaciers or basins requires a DEM obtained from stereo-processing of aerial photos cap-
higher spatial resolution, which in the past could only be tured in 2007.
achieved via aerial laser scanner and photogrammetric sur-
veys (Vincent et al., 2010; Janke, 2013) or dedicated field
campaigns with terrestrial laser scanners (TLS) (Kellerer- 2 Study area
Pirklbauer et al., 2005; Riccardi et al., 2010). Recent years
The Forni Glacier (see Fig. 1) has an area of 11.34 km2 based
have seen a resurgence of terrestrial photogrammetric sur-
on the 2007 data from the Italian Glacier Inventory (Smi-
veys for the generation of DEMs (Piermattei et al., 2015,
raglia et al., 2015), an altitudinal range between 2501 and
2016; Kaufmann and Seier, 2016) due to important tech-
3673 m a.s.l. and a north-northwesterly aspect. The glacier
nological advances, including the development of structure-
has retreated markedly since the Little Ice Age, when its
from-motion (SfM) photogrammetry and its implementation
area was 17.80 km2 (Diolaiuti and Smiraglia, 2010), with
in fully automatic processing software, as well as improve-
an acceleration of the shrinkage rate over the past three
ments in the quality of camera sensors (Eltner et al., 2016;
decades, typical of valley glaciers in the Alps (Diolaiuti et al.,
Westoby et al., 2012). In parallel, unmanned aerial vehicles
2012; D’Agata et al., 2014). It has also undergone profound
(UAVs – Colomina and Molina, 2014; O’Connor et al., 2017)
changes in dynamics in recent years, such as the loss of ice
have started to emerge as a viable alternative to TLS for
flow from the eastern accumulation basin towards its tongue
multitemporal monitoring of small areas. UAVs promise to
and the evidence of collapsing areas on the eastern tongue
bridge the gap between field observations, notoriously diffi-
(see Fig. 2d; Azzoni et al., 2017). Continuous monitoring of
cult on glaciers, and coarser-resolution satellite data (Bhard-
these hazards is important, as the site is highly touristic (Gar-
waj et al., 2016). Although the number of studies employing
avaglia et al., 2012). The glacier is in fact frequently visited
these platforms in high mountain environments is on the rise
during both summer and winter months. During the summer,
(see, e.g., Fugazza et al., 2015; Gindraux et al., 2017; Seier et
hikers heading to Mount San Matteo take the trail along the
al., 2017), their full potential for monitoring glaciers and par-
central tongue, accessing the glacier through the left flank of
ticularly glacier hazards has yet to be explored. In particular,
the collapsing glacier terminus (see Fig. 2b and c). During
the advantages of UAV and terrestrial SfM photogrammetry
wintertime, ski mountaineers instead access the glacier from
and the possibility of data fusion and volume change esti-
the eastern side, crossing the medial moraine and potentially
mation to support hazard management strategies in glacial
collapsed areas there (see Figs. 1 and 2a).
environments need to be investigated and assessed.

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1057

3.1 UAV photogrammetry

3.1.1 2014 dataset

The first UAV survey took place on 28 August 2014, us-

ing a SwingletCam fixed wing aircraft (see Fig. 3a). This
commercial platform developed by SenseFly carries a Canon
Ixus 127 HS compact digital camera. The UAV was flown in
autopilot mode with a relative flying height of approximately
380 m above the glacier surface, which resulted in an average
GSD of 12 cm. The flight plan was organized by using the
proprietary software eMotion, by which the aircraft follows
predefined waypoints with a nominal along-strip overlap of
70 %. In our study, side lap was not regular because of the
varying surface topography, but it averaged approximately
60 %. Flight operations started around 07:30 LT and ended
around 08:30 LT. Early-morning operations were preferred
to avoid saturating camera pictures, as during this time of
day the glacier is not yet directly illuminated by the sun, and
to minimize blurring effects due to the UAV motion, since
wind speed is at its lowest on glaciers during morning hours
(Fugazza et al., 2015). Pictures were automatically captured
by the UAV platform, selecting the best combination of sen-
sor aperture (F/2.7), sensitivity (between 100 and 400 ISO)
and shutter speed (between 1/125 and 1/640 s). The survey
covered an area of 2.21 km2 in two flight campaigns, with a
low-altitude take-off (see Fig. 1). Both the terminal parts of
the central and eastern ablation tongue were surveyed.
Since no GCPs were measured during the 2014 campaign,
Figure 1. The tongue of Forni Glacier. The map shows the location the registration of this data set into the mapping reference
of take-off and landing sites for the 2014 and 2016 UAV surveys, system was based on GNSS (Global Navigation Satellite
standpoint of TLS survey, GCPs used in the UAV photogramme- System) navigation data only. Consequently, a global bias
try surveys and trails crossing the glaciers. Letters a–e identify the on the order of 1.5–2 m resulted after georeferencing, and
location of features described in Fig. 2. Base map from 2015 cour-
no control on the intrinsic geometric block stability was pos-
tesy of IIT Regione Lombardia WMS Service. Trails from Kom-
pass online cartography at https://www.kompass-1039italia.it/info/
sible. After the generation of the point cloud, a DEM and
mappa-online/. orthophoto were produced with spatial resolutions of 60 and
15 cm, respectively.

3 Data sources: acquisition and processing 3.1.2 2016 dataset

In this study, we took advantage of a UAV survey performed Two UAV surveys were carried out on 30 August and
in 2014 (Fugazza et al., 2015). Then, through a field cam- 1 September 2016, both around midday with 8/8 of the
paign in 2016, we conducted different surveys using a UAV, sky covered by stratocumulus clouds. The UAV employed
terrestrial photogrammetry and TLS. In the 2014 UAV sur- in these surveys was a customized quadcopter (see Fig. 3b)
vey, no ground control points (GCPs) were collected, while carrying a Canon Powershot 16 Megapixel digital camera.
in 2016 we specifically set up a control network for georef- Two different take-off and landing sites were chosen to gain
erencing purposes. Processing of the UAV and terrestrial im- altitude before take-off and maintain line-of-sight operation
ages was carried out using Agisoft Photoscan version 1.2.4 with a flying altitude of 50 m above ground, which ensured an
(http://www.agisoft.com), implementing a SfM algorithm for average GSD of 6 cm. To reduce motion blur, camera shutter
image orientation followed by a multi-view dense-matching speed was set to the lowest possible setting, 1/2000 s, with
approach for surface 3-D reconstruction (Westoby et al., aperture at F/2.7 and sensitivity at 200 ISO.
2012). In addition, we employed a DEM from an aerial sur- Several flights were conducted to cover a small section
vey of 2007 to calculate glacier thickness changes over a pe- of the proglacial plain and different surface types on the
riod of 7 to 9 years. glacier surface, including the terminus, a collapsed area on
the central tongue, the eastern medial moraine and some

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1058 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

Figure 2. Collapsing areas on the tongue of Forni Glacier: (a) faults cutting across the eastern medial moraine; (b) glacier terminus; (c) near-
circular collapsed area on the central tongue; (d) large ring fault on the eastern tongue at the base of the icefall. (e) Close-up of a vertical ice
cliff at the glacier terminus. The location of features is reported in Fig. 1. Photos courtesy of G. Cola.

Figure 3. The UAVs used in surveys of the Forni Glacier and their characteristics. (a) The SwingletCam fixed-wing aircraft employed
in 2014, at its take-off site by Lake Rosole; (b) the customized quadcopter used in 2016 in the lab.

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1059

debris-covered parts of the eastern tongue. A “zig-zag” fly- GCPs were measured by using a precision theodolite follow-
ing scheme was followed to reduce the flight time. The UAV ing the same procedure adopted for terrestrial photogramme-
was flown in autopilot mode using the open-source software try. Considering the accuracy of registration and the expected
Mission Planner (Oborne, 2013) to ensure 70 % along-strip precision of laser point measurement, the global uncertainty
overlap and side lap. In total, two flights were performed dur- of 3-D points was estimated on the order of ±7.5 cm.
ing the first survey and three during the second, lasting about
20 min each. The surveyed area spanned over 0.59 km2 . 3.4 GNSS ground control points
Eight GCPs (see Fig. 1) were measured for the registra-
tion of the photogrammetric blocks and its byproducts into Prior to the 2016 surveys, eight control targets were placed
the mapping system. The root mean square error (RMSE) of both on the periglacial area and on the glacier tongue (see
the GCP location was 40 cm, which can be used as an in- Fig. 1). Differential GNSS data were acquired at the target lo-
dicator of the internal consistency of the photogrammetric cation for the georeferencing of UAV, terrestrial photogram-
block. The point cloud obtained from the 2016 UAV flight metry and TLS data. GCPs were used (1) to georeference
was interpolated to produce a DEM and orthophoto with UAV data directly, by identifying the targets on the images
the same cell resolution as the 2014 dataset, i.e., 60 and in Photoscan and (2) to register theodolite measurements for
15 cm, respectively. Both products were exported in the georeferencing terrestrial photogrammetry and TLS. The tar-
ITRS2000/UTM 32N mapping reference system. gets consisted in a square piece of white fabric (80 × 80 cm),
with a circular marker in red paint chosen to provide con-
3.2 Terrestrial photogrammetry trast against the background. Except for the one GCP located
at the highest site, such GCPs were positioned on large, flat
The terrestrial photogrammetric survey was carried out on boulders to provide a stable support and reduce the impact of
29 August 2016 to reconstruct the topographic surface of the ice ablation between flights.
glacier terminus, which presented several vertical and sub- GNSS data were acquired by means of a pair of Le-
vertical surfaces (see Fig. 2e) whose measurement was not ica Geosystems 1200 geodetic receivers working in RTK
possible from the UAV platform carrying a camera in nadir (real-time kinematics) mode (see Hoffman-Wellenhof et al.,
configuration. 2008). One of them was set up as master on a precise point
Images were captured from 134 ground-based stations, beside Branca hut, with known coordinates in the mapping
most of them located in front of the glacier and some on reference system ITRS2000/UTM 32N. The second receiver
both flanks of the valley in the downstream area. A single- was used as a rover, communicating via radio link with the
lens-reflex Nikon D700 camera was used, equipped with a master station. The maximum distance between master and
50 mm lens, and a full-frame CMOS sensor (36 × 24 mm) rover was less than 1.5 km, but some points were measured in
with 4256 × 2823 pixels. In this case, since no preliminary static mode with measurement time of approximately 12 min
information about approximate camera position was col- due to the local topography preventing the radio link and the
lected, the SfM procedure was run without any initial infor- lack of mobile phone services (for RTK). The theoretical un-
mation. certainty of GCPs provided by the processing code was on
Seven natural features visible on the glacier front were the order of 2–3 cm.
used as GCPs to be included in the bundle adjustment com-
3.5 2007 DEM
putation. Measurement of GCPs in the field was carried out
by means of a high-precision theodolite. The measurement The 2007 TerraItaly DEM was produced by the BLOM
of points previously recorded with a GNSS geodetic receiver CGR company for the Lombardy region. It is the final prod-
made it possible to register the coordinates of GCPs in the uct of an aerial survey over the entire region, conducted
mapping reference system. The RMSE of 3-D residual vec- with a multispectral push broom Leica ADS40 sensor ac-
tors on GCPs was 34 cm. quiring images from a flying height of 6300 m with an av-
erage GSD of 65 cm. The images were processed to generate
3.3 Terrestrial laser scanning a DEM with a cell resolution of 2 m × 2 m and a ±3 m un-
certainty. We converted the DEM from the “Monte Mario”
On the same days as the first UAV survey of 2016, a long- to the “ITRS2000” datum and the height from ellipsoidal to
range terrestrial laser scanner Riegl LMS-Z420i was used to geodetic using the official software for datum transformation
scan the glacier terminus. One instrumental standpoint lo- in Italy (Verto ver. 3).
cated on the hydrographic left flank of the glacier terminus
(see Fig. 1) was established. The horizontal and vertical scan-
ning resolution were set up to provide a spatial point density
of approx. 5 cm on the ice surface at the terminus. Georefer-
encing was accomplished by placing five GCPs consisting in
cylinders covered by retroreflective paper. The coordinates of

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1060 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

Figure 4. Location of different glacier features or hazard-prone areas on the tongue of Forni Glacier were the point cloud comparison was
performed. The background image is the merged point cloud generated from the 2016 UAV and terrestrial photogrammetry survey.

4 Methods the presence of enough points to completely describe a por-

tion of surface. In this study, the visual inspection of selected
4.1 Analysis of point clouds from the 2016 campaign: sample locations was used to identify occlusions and areas
UAV/terrestrial photogrammetry and TLS with lower point density.
To analyze these properties, five regions were selected (see
The comparison between point clouds generated during the Fig. 4), located on the glacier topographic surface and char-
2016 campaign had the aim of assessing their geometric acterized by different glacier features and the presence of
quality before their application for the analysis of hazards. hazards: (1) a glacial cavity composed of subvertical and
These evaluations were also expected to provide some guide- fractured surfaces over 20 m high and forming a typical semi-
lines for the organization of future investigations in the field circular shape; (2) a glacial cavity over 10 m high with the
at the Forni Glacier and in other Alpine sites. Specifically, same typical semicircular shape as location 1, covered by
we analyzed point density (points m−2 ) and completeness, fine- and medium-sized rock debris; (3) a normal fault over
i.e., % of area in the ray view angle. Point density partly 10 m high; (4) a highly collapsed area covered by fine- and
depends upon the surveying technique used, since it is con- medium-sized rock debris and rock boulders; and (5) a planar
trolled by the distance between sensor and surface, and de- surface with a normal fault covered by fine- and medium-
termines spatial resolution. In SfM photogrammetry, point sized rock debris and rock boulders. The analysis of lo-
density is affected by image texture, sharpness and resolu- cal regions was preferred to the overall analysis of all the
tion, which influence the performance of dense matching al- point clouds due to (1) the incomplete overlap between point
gorithms (Dall’Asta et al., 2015), while in TLS it can be set clouds obtained from different methods and (2) the oppor-
up as a data acquisition input parameter. In this study, the tunity to investigate the performances of the techniques in
number of neighbors N (inside a sphere of radius R = 1 m) diverse geomorphological situations.
divided by the neighborhood surface was used to evaluate Within the same sample locations, we compared the point
the local point density D in CloudCompare (http://www. clouds in a pairwise manner. Since no available benchmark-
cloudcompare.org). To understand the effect of point density ing data set (e.g., accurate static GNSS data) was concur-
dispersion (Teunissen, 2009), the inferior 12.5 percentile of rently collected during the 2016 campaign, the TLS point
the standard deviation σ of point density was also calculated. cloud was used as a reference. When comparing both pho-
The use of these local metrics allowed us to distinguish be- togrammetric data sets, the one obtained from the UAV was
tween point densities in different areas, since this may largely used as reference because of the even distribution of point
change from one portion of surface to another. A further met- density within the sample locations. The presence of resid-
ric in this sense was point cloud completeness, referring to ual, non-homogenous georeferencing errors in the data sets

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1061

required a specific fine registration of each individual sam-

ple location, which was conducted in CloudCompare using
the ICP (iterative closest point) algorithm (Pomerleau et al.,
2013). ICP iteratively matches a source point cloud to a ref-
erence point cloud in Euclidean space and calculates the nec-
essary rotation and translation to align the source point cloud
with the reference based on minimization of a distance met-
ric in a point-to-point fashion. After fine registration, point
clouds in corresponding sample areas were compared using
the M3C2 algorithm implemented in CloudCompare (Lague
et al., 2013). As discussed in Fey and Wichmann (2016), the
distance between a pair of point clouds is often evaluated by
comparing elevations at corresponding nodes of DEMs, after
resampling of the original data. This approach works prop-
erly when both point clouds are approximately aligned along
the same planar direction, but not when there are structures
with different alignments as in the case of the glacier surfaces
under investigation. In fact, the M3C2 algorithm does not al-
ways evaluate the distance between two point clouds along
the same directional axis, but computes a set of local normals
using points within a radius D depending on the local rough-
ness, which is directly estimated from the point cloud data, Figure 5. Maps of point density in sample location 2.
and also considering the accuracy of preliminary local regis-
tration refinement using ICP. In this case, a radius D of 20 cm
and a pre-registration accuracy of 5 cm were considered, the formation, it helped to provide a realistic visualization of the
latter obtained from ICP residuals. This solution allowed us topographic model and therefore to interpret glacier hazards.
to remove registration errors from the analysis and focus on
the capability of the adopted techniques to reconstruct the 4.3 Glacier hazard mapping
local geometric surface of the glacier in an accurate way.
The investigation of glacier hazards was conducted using the
4.2 Point cloud merging point cloud and orthophoto from the 2014 UAV data set as
well as the merged (UAV and terrestrial) point cloud and
To improve coverage of different glacier surfaces, including orthophoto from 2016. In this study, we focused on ring
planar areas and normal faults, photogrammetric point clouds faults and normal faults, which were identified by visually
from the 2016 campaign (UAV and terrestrial surveys) were inspecting their geometric properties in the point clouds and
merged. We chose to avoid TLS and employed the two lower- manually delineated, while color information from orthopho-
cost techniques (Chandler and Buckley, 2016) to assess their tos was used as a cross check. On orthophotos, both types
potential for combined future use. Prior to point cloud merg- of structures generally appear as dark linear features ow-
ing, a preliminary co-registration was performed on the basis ing to shadows projected by fault scarps. As these structures
of the ICP algorithm in CloudCompare. Regions common to may look similar to crevasses, further information concern-
both point clouds were used to minimize the distances be- ing their orientation and location needs to be assessed for
tween them and find the best co-registration. The point cloud discrimination. The orientation of fault structures is not co-
from UAV photogrammetry, which featured the largest ex- herent with glacier flow, with ring faults also appearing in
tension, was used as reference during co-registration, while circular patterns. Their location is limited to the glacier mar-
the other was rigidly transformed to fit with it. After many gins, medial moraines and terminus, whereas crevasses can
iterations, both point clouds were aligned according to the appear anywhere on the glacier surface (Azzoni et al., 2017).
best solution found by the ICP. In order to remove redundant After delineation, we also analyzed the height of vertical fa-
points and to obtain a homogenous point density, the merged cies using information from the point clouds.
point cloud (see Fig. 5) was subsampled keeping a minimum
distance between adjacent points of 20 cm. The final size of 4.4 DEM co-registration for glacier thickness change
this merged data set is approximately 4.4 million points. The estimation
RGB color information associated with each point in the fi-
nal point cloud was derived by averaging the RGB informa- Several studies have found that errors in individual DEMs,
tion of original points in the subsampling volumes. While both in the horizontal and vertical domain, propagate when
this operation resulted in losing part of the original RGB in- calculating their difference, leading to inaccurate estima-

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1062 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

tions of thickness and volume change (Berthier et al., 2007; Table 1. Statistics of the elevation differences between DEM pairs
Nuth and Kaab, 2011). In the present study, different ap- before and after the application of co-registration shifts. DEM 2007
proaches were adopted for georeferencing all the DEMs used is from aerial multispectral survey; DEM 2014 and DEM 2016 are
in the analysis of the volume change of the Forni Glacier from UAV photogrammetry.
tongue (2007, 2014, 2016). To compute the relative dif-
DEM pair Elevation Co-registration shifts Elevation
ferences between the DEMs, a preliminary co-registration
was therefore required. The method proposed by Berthier et differences X (m) Y (m) differences with
without co-registration
al. (2007) for the co-registration of two DEMs was separately
co-registration shifts
applied to each DEM pair (2007–2014; 2007–2016; 2014– shifts (µ1H ± σ1H )
2016). Following this method, in each pair one DEM plays (µ1H ± σ1H ) (m)
as reference (“master”), while the other is used as “slave” (m)
DEM to be iteratively shifted along x and y axes by frac- 2007–2014 1.96 ± 2.60 1.11 −1.11 0.00 ± 1.70
tions of a pixel to minimize the standard deviation of eleva- 2007–2016 −0.43 ± 3.48 2.44 −1.11 0.00 ± 2.60
2014–2016 −2.92 ± 3.21 −0.20 −1.30 0.00 ± 2.22
tion differences with respect to the master DEM. Only ar-
eas assumed to be stable are considered in the calculation
of the co-registration shift. The ice-covered areas were ex-
cluded by overlaying the glacier outlines from D’Agata et 5 Results
al. (2014) for 2007 and Fugazza et al. (2015) for 2014. The
oldest DEM, which is also the widest in each comparison, 5.1 Point cloud analysis
was always set as the master. To co-register the 2014 and
2016 DEMs with the 2007 DEM, both were resampled to The analysis of point density shows significant differences
2 m spatial resolution, whereas the comparison between 2014 between the three techniques for point cloud generation (see
and 2016 was carried out at the original resolution of these Table 2). Values range from 103 to 2297 points m−2 depend-
data sets (60 cm). ing on the surveying method, but the density was gener-
All points resulting in elevation differences greater than ally sufficient for the reconstruction of the different surfaces
15 m were labeled as unreliable, and consequently discarded shown in Fig. 4, except for location 5. Terrestrial photogram-
from the subsequent analysis. Such greater discrepancies metry featured the highest point density, while UAV pho-
may denote errors in one of the DEMs or unstable areas out- togrammetry had the lowest. In relation to UAV photogram-
side the glacier. Values exceeding this threshold, however, metry, similar point densities were found in all sample lo-
were only found in a marginal area with low image over- cations, especially for the standard deviations that were al-
lap in the comparison between the 2014 and 2016 DEMs, ways in the 22–29 point m−2 range. Mean values were 103–
with a maximum elevation difference of 36 m. Once the final 109 points m−2 in locations 2–4, while they were higher in
co-registration shifts were computed (see Table 1), the coef- location 5 (141 points m−2 ). Due to the nadir acquisition
ficients were subtracted from the top left coordinates of the points, the 3-D modeling of vertical/subvertical cliffs in lo-
slave DEM; the residual mean elevation difference was also cation 1 was not possible. In relation to TLS, a mean value of
subtracted from the slave DEM to bring the mean to zero. point density ranging from 141 to 391 points m2 was found,
After DEM co-registration, the resulting shifts reported in with the only exception of location 5, where no sufficient
Table 1 were applied to each slave DEM, including the en- data were recorded due to the position of this region with
tire glacier area. Then the elevations of the slave DEM were respect to the instrumental standpoint. Standard deviations
subtracted from the corresponding elevations of the master ranged between 69 and 217 points m2 , moderately correlated
DEM to obtain the so-called DEM of differences (DoD). with respective mean values. The analysis of the complete-
Over a common glacier area (Fig. 1), we estimated the vol- ness of surface reconstruction also revealed some issues re-
ume change and its uncertainty, which can be expressed as lated to the adopted techniques (see Fig. 5). Specifically, TLS
the combination of (1) uncertainty due to errors in elevation suffered from severe occlusions, which prevented acquisition
and (2) the truncation error caused by the use of a discrete of data in the central part of the sample area, while UAV pho-
sum (sum of DoD at each pixel multiplied by pixel area) togrammetry was able to reconstruct the upper portion of the
in place of the integral in volume calculation (Jokinen and sample area but not the vertical cliff. Only terrestrial pho-
Geist, 2010). We calculated the former following the ap- togrammetry acquired a large number of points in all areas.
proach of Rolstad et al. (2009), taking into account spatial In terms of point cloud distance (see Table 3), the compar-
autocorrelation of elevation change over stable areas, con- ison between TLS and terrestrial photogrammetry resulted
sidering a correlation length of 50 m; for the latter, we used in a high similarity between point clouds, with no great
the method described by Jokinen and Geist (2010). differences between different sample areas. Conversely, the
comparison between TLS and UAV photogrammetry and
terrestrial and UAV photogrammetry provided significantly
worse results, which may be summarized by the RMSEs in

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1063

Table 2. Area and number of points in each sample window on the Forni Glacier terminus, mean and standard deviation of local point
density and number of points above the lower 12.5 % percentile in each window. k stands for thousands of points. UAV refers to UAV
photogrammetry, TP to terrestrial photogrammetry and TLS to terrestrial laser scanning.

Sample Area Number of points above

window (m2 ) Number of points in sample Mean and standard deviation of point the lower 12.5 %
windows density (points m−2 ) percentile
UAV TP TLS UAV TP TLS UAV TP TLS
1 2793 – 1984k 141k – 1654 ± 637 226 ± 100 – 880 26
2 1806 76k 2175k 130k 109 ± 29 2297 ± 708 391 ± 217 61 881 0
3 495 43k 712k 25k 103 ± 27 1978 ± 606 151 ± 60 49 766 31
4 672 62k 557k 33k 108 ± 22 1384 ± 530 141 ± 69 62 324 2
5 3960 406k 810k – 141 ± 22 485 ± 227 – 97 31 –

Table 3. Statistics on distances between point clouds computed on the basis of the M3C2 algorithm, showing mean, standard deviation and
root mean square error (RMSE) of each point cloud pair. UAV refers to UAV photogrammetry, TP to terrestrial photogrammetry and TLS to
terrestrial laser scanning. Ref. stands for reference and “–” means no comparison was performed.

Sample Means and SDs of M3C2 distances RMSE of M3C2

window (cm) distances (cm)
Ref. TLS TLS UAV TLS TLS UAV
Slave TP UAV TP TP UAV TP
1 4.5 ± 7.4 – – 8.7 – –
2 −1.1 ± 10.5 14.8 ± 34.7 −14.5 ± 26.7 10.6 37.7 30.4
3 8.4 ± 4.1 14.7 ± 15.1 −8.5 ± 18.9 9.4 21.1 20.7
4 2.8 ± 5.3 9.4 ± 22.2 −2.3 ± 24.9 6.0 24.0 25.0
5 – – −8.5 ± 25.3 – – 26.7

the ranges of 21.1–37.7 and 20.7–30.4 cm, respectively. The ship with glacier downwasting. While most collapsed areas
greater deviations were in both cases obtained in the analy- on Forni Glacier are normal faults, two large ring fault sys-
sis of location 2, which mostly represents a vertical surface, tems can be identified: the first, located in the eastern sec-
while the best agreement was found within location 3, which tion (see Figs. 2d and 7), covered an area of 25.6 × 103 m2
is less inclined. As the UAV flight was georeferenced on a set and showed surface dips of up to 5 m in 2014. This area was
of GCPs with an RMSE of 40.5 cm, the ICP co-registration not surveyed in 2016, since field observation did not show
may have not totally compensated the existing bias. evidence of further significant subsidence. Conversely, the
In terms of spatial coverage, considering the entire surface ring fault that only emerged as a few semicircular fractures
examined using each technique outside the sample locations, in 2014 grew until cavity collapse, with a vertical displace-
the UAV survey extended over the widest area (0.59 km2 ), ment up to 20 m. Additional fractures, extending southeast-
including part of the proglacial plain, the entire terminus and ward (see Figs. 2c and 7), formed between 2014 and 2016,
the glacier tongue up to the collapsed area on the central suggesting that the extent of the collapse might widen in the
part, but with data gaps on the vertical and subvertical walls near future. In 2014, further ring faults were also identified
(see Fig. 6a). The point cloud obtained from terrestrial pho- at the eastern glacier margin, and the one that was surveyed
togrammetry covered approximately a third of the area sur- in 2016 showed evidence of increasing subsidence (+2 m)
veyed with the UAV (see Fig. 6b), including the full glacier and the formation of additional subparallel fractures.
terminus at very high spatial resolution, with the exception of Normal faults are mostly found on the eastern medial
a few obstructed parts, while the TLS point cloud covered the moraine and at the terminus. Between 2014 and 2016, the
terminus, although with some holes due to the obstructions. first (see Fig. 2a) developed rapidly in the vertical domain
reaching a height of 12 m in 2016. The latter increased in
5.2 Glacier-related hazards and risks number as size, as the terminus underwent collapse, leading
to the formation of three major ice cavities, up to 24 m high
The tongue of Forni Glacier hosts several hazardous struc- (see Fig. 2b and e). In 2016, ice thickness above the cavity
tures, including crevasses, normal faults and ring faults. In vault was between 5 and 10 m (see Fig. 4, location 1). Several
this study, we focused on the latter two due to their relation-

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1064 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

Table 4. Average ice thickness change, thinning rates and vol-

ume loss from DEM differencing over a common reference area
of 0.32 km2 for all DEM pairs. Uncertainty of thickness change
expressed as 1 standard deviation of residual elevation differences
over stable areas after DEM co-registration.

DEM pair Mean thickness Mean thinning Volume change

change (m) rates (m a−1 ) (106 m3 )
2007–2014 −31.91 ± 1.70 −4.55 ± 0.24 −10.00 ± 0.17 (1.74 %)
2007–2016 −42.86 ± 2.60 −4.76 ± 0.29 −13.46 ± 0.20 (1.47 %)
2014–2016 −10.41 ± 2.22 −5.20 ± 1.11 −3.29 ± 0.08 (2.60 %)

gins, potentially compromising access to the glacier. Since

this system of fractures has been developing very rapidly,
and new collapses have already been documented in Septem-
ber 2017, the risk for people walking on the glacier tongue
during summer should be carefully evaluated. Although sur-
face features may be visually detected, the availability of de-
tailed 3-D models that depict the entire outer surface of the
glacier is a great advantage because it allows quickly cap-
turing the glacier topography remotely, helping predict the
possible development of new collapses and understand their
mechanisms of formation.

5.3 Glacier thickness change

The Forni Glacier tongue was affected by substantial thin-

ning throughout the observation period. Between 2007
and 2014, the greatest thinning occurred in the eastern sec-
tion of the glacier tongue, with changes persistently lower
than −30 m (more than 4 m a−1 thinning), whereas the up-
Figure 6. Spatial coverage of UAV and terrestrial photogramme- per part of the central tongue only thinned by 10–18 m (be-
try point clouds and merged point cloud from the two techniques. tween approximately 1 and 2.5 m a−1 ). The greatest ice loss
(a) UAV photogrammetry point cloud; (b) terrestrial photogramme- occurred in correspondence with the normal faults localized
try point cloud; (c) merged point cloud. in small areas at the eastern glacier margin (see Fig. 8a),
with local changes generally lower than −50 m (more than
7 m a−1 thinning) and a minimum of −66.80 m, owing to
fractures identified as normal faults also appear in conjunc- the formation of a lake. Conversely, between 2014 and 2016
tion with the large ring fault located in the central section the central and eastern parts of the tongue had similar thin-
of the glacier, extending the fracture system to the western ning patterns, with average changes of −10 m (5 m a−1 ). The
glacier margin. greatest losses are mainly found in correspondence with nor-
The presence of cavities at the terminus, which is easily mal faults, with a maximum change of −38.71 m at the ter-
reached in a 45 min walk from Branca hut, is particularly minus and local thinning greater than 25 m on the lower me-
hazardous for tourists, because of (1) the danger of cavity dial moraine. The ring fault at the left margin of the central
collapse and (2) the potential fall of large boulders or blocks section of the tongue also shows thinning of 20 to 26 m (10–
of ice from the ice cliffs. Other fracture systems located 13 m a−1 ). In the absence of faults, little thinning occurred
higher up glacier are presumably reached by more experi- instead on the upper part of the medial moraine, where a
enced hikers. Most ring faults, however, show evidence of thick debris cover shielded ice from ablation, with changes
vertical and horizontal expansion, and further cavity collapse of −2 to −5 m (1 to 2.5 m a−1 ; see Fig. 8c). Considering a
would imply a severe risk of injury and death if hikers were common reference area (see Fig. 1, Table 4), an acceleration
involved. of glacier thinning seems to have occurred over recent years
The location of structures in 2016 suggests that the glacier over the lower glacier tongue, from −4.55 ± 0.24 m a−1
terminus will recede through ice cliff backwasting and cavity in 2007–2014 to −5.20 ± 1.11 m a−1 in 2014–2016, also
collapse along the fault system on the eastern medial moraine confirmed by the value of −4.76 ± 0.29 m a−1 obtained from
and along the ring faults at the eastern and western mar- the comparison between 2007 and 2016. Looking at the first

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1065

Figure 7. Location of collapse structures, i.e., normal faults and ring faults and trails crossing the Forni Glacier. (a) Collapse structures
in 2014, with 2014 UAV orthophoto as base map. The red box marks the area surveyed in 2016. (b) Collapse structures in 2016, with
2016 UAV orthophoto as base map. Trails from Kompass online cartography at https://www.kompass-1039italia.it/info/mappa-online/.

Figure 8. Ice thickness change rates from DEM differencing over (a) 2007–2014, (b) 2007–2016 and (c) 2014–2016. Glacier outlines
from 2014 and 2016 are limited to the area surveyed during the UAV campaigns. Base map from hill shading of 2007 DEM.

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1066 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

two DoDs, the trend seems to be caused by the increase in 2014), we found that the TLS point cloud was affected by
collapsing areas (Fig. 8a and b). In all DoDs, the uncertainty occlusions (see, e.g., location 2 in Figs. 4 and 5), which can
in ice thickness change affects less than 3 % of the respective only be compensated by increasing the number of stations.
volume change (see Table 4). Data acquisition with this platform was in general difficult in
regions subparallel to the laser beams and in the presence of
wet surfaces.
6 Discussion: comparison of techniques for point cloud
generation 6.2 Point cloud and DEM uncertainty

In this study, the distance between the UAV and TLS point
The choice of a technique to monitor glacier hazards and the
clouds (21.1–37.7 cm RMSE), assumed as a measure of the
glacier thickness changes depends on several factors, includ-
uncertainty of the 2016 UAV data set, was slightly higher
ing the size of the area, the desired spatial resolution and
than previously reported in high mountain glacial environ-
accuracy, logistics and cost. In this study, we focused on spa-
ments (e.g., Immerzeel et al., 2014; Gindraux et al., 2017;
tial metrics, i.e., point density, completeness and distance
Seier et al., 2017), although in these studies the comparison
between point clouds, to evaluate the performance of UAV,
was between DEMs and GNSS control points. Contributing
close-range photogrammetry and TLS in a variety of condi-
factors might include the suboptimal distribution and den-
tions.
sity of GCPs (Gindraux et al., 2017), the delay between the
6.1 Point density and completeness UAV surveys as well as between the UAV and TLS, and the
lack of coincidence between GCP placement and the UAV
Considering point density, terrestrial photogrammetry re- flights. This means the UAV photogrammetric reconstruc-
sulted in a denser data set than the other techniques. This tion was affected by ice ablation and glacier flow, which on
is mostly motivated by the possibility of acquiring data from Forni Glacier range between 3 and 5 cm day−1 (Senese et al.,
several stations using this methodology, depending only on 2012) and between 1 and 4 cm day−1 , respectively (Urbini
the terrain accessibility, reducing the effect of occlusions et al., 2017). We thus expect a combined 3-day uncertainty
with a consequently more complete 3-D modeling. However, on the 2016 UAV data set between 10 and 20 cm and lower
the mean point density achieved when using terrestrial pho- on GCPs, considering reduced ablation due to their place-
togrammetry is highly variable, both between different sam- ment on boulders. A further contribution to the GCP error
ple locations and within each location as shown by the stan- budget might stem from the intrinsic precision of GNSS and
dard deviations of D. Point densities related to UAV pho- theodolite measurements and image resolution. The compar-
togrammetry and TLS are more regular and constant. In the ison between close-range photogrammetry and TLS was less
case of UAV photogrammetry, the homogeneity of point den- affected by glacier change because data were collected 1 day
sity might be due to the regular structure of the airborne apart and the RMSE of 6–10.6 cm is in line with previous
photogrammetric block. In the case of TLS, the regularity is findings by Kaufmann and Landstaedter (2008). To reduce
motivated by the constant angular resolution adopted during the uncertainty of UAV photogrammetric blocks, a better dis-
scanning. Since any technique may perform better when the tribution of GCPs or switching to an RTK system should be
surface to survey is approximately orthogonal to the sensor’s considered, while close-range photogrammetry could benefit
point of view, terrestrial photogrammetry is more efficient from measuring a part of the photo-stations as proposed in
for reconstructing vertical and subvertical cliffs (sample ar- Forlani et al. (2014) instead of placing GCPs on the glacier
eas 1 and 2) and high-sloped surfaces (sample areas 3 and 4). surface.
In contrast, airborne UAV photogrammetry provided the best The uncertainty in UAV photogrammetric reconstruction
results in location 5 which is less inclined and consequently also factored in the standard deviation still present after the
could be well depicted in vertical photos. In general, point co-registration between DEMs in areas outside the glacier
clouds from terrestrial photogrammetry provide a better de- (2.22 m between 2014 and 2016). Another important factor
scription of the vertical and subvertical parts (see, e.g., Win- here is the morphology of the co-registration area, i.e., the
kler et al., 2012), while point clouds obtained from UAV pho- outwash plain, still subject to changes due to the inflow
togrammetry are more suitable to describe the horizontal or of glacier meltwater and sediment reworking. UAV pho-
subhorizontal surfaces on the glacier tongue and periglacial togrammetric products permitted us to investigate ice volume
area (Seier et al., 2017), unless the camera is tilted to an off- changes over 2 years with an uncertainty of 2.60 %, while the
nadir viewpoint (Dewez et al., 2016; Aicardi et al., 2016). integration with close-range photogrammetry was required to
Results obtained from photogrammetry based on terrestrial investigate hazards related to the collapse of the glacier ter-
and UAV platforms can thus be considered quite complemen- minus.
tary and they support the concept of merging the point clouds
from these two techniques, as seen in Fig. 6c. In agreement
with other studies of vertical rock slopes (e.g., Abellán et al.,

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1067

6.3 Logistics and costs identify safer paths where mountaineers and skiers can visit
the glacier and reach the most important summits. However,
In our surveys, it became evident that the main disadvantage the increase in collapse structures owing to climate change
of TLS compared to photogrammetry is the complexity of in- requires multitemporal monitoring. A comprehensive risk as-
strument transport and setup. In terms of logistics and work- sessment should also cover the entire glacier, to investigate
load, up to five people were involved in the transportation of the probability of serac detachment and provide an estimate
the TLS instruments (laser scanner, theodolite, at least two of the glacier mass balance with the geodetic method. While
topographic tripods and poles, electric generator and ancil- our integrated approach using a multicopter and terrestrial
lary accessories) while two people were required for UAV photogrammetry should be preferred to TLS for the investi-
and close-range photogrammetric surveys, which were also gation of small individual ice bodies, fixed-wing UAVs, ide-
considerably faster. Meteorological conditions and the lim- ally equipped with an RTK system and the ability to tilt the
ited access to unstable areas close to the glacier terminus also camera off-nadir, might be the platform of choice to cover
prevented the acquisition of TLS data from other viewpoints large distances (see, e.g., Ryan et al., 2017), potentially re-
as done with photogrammetry. Concerning UAV surveys, we ducing the number of flights and solving issues with GCP
conducted them under different meteorological scenarios and placement. Such platforms could help collect sufficient data
obtained adequate results in early-morning operations with for hazard management strategies up to the basin scale in
0/8 cloud cover and midday flights with 8/8 cloud cover. Stelvio National Park and other sectors of the Italian Alps,
Both scenarios can provide diffuse light conditions allow- eventually replacing higher-altitude aerial surveys. Cost anal-
ing collection of pictures suitable for photogrammetric pro- yses (Matese et al., 2015) should also be performed to eval-
cessing, but camera settings need to be carefully adjusted be- uate the benefits of improved spatial resolution and lower
forehand (O’Connor et al., 2017). If early-morning flights are DEM uncertainty of UAVs compared to aerial and satellite
not feasible in the study area for logistical reasons or when surveys and choose the best approach for individual cases.
surveying glaciers with eastern exposures, the latter scenario
should be considered.
In terms of costs, UAV and terrestrial photogrammet- 7 Conclusions
ric surveys are also advantageous, since TLS instruments
In our study, we compared point clouds generated from UAV
are much more expensive at EUR 70 000–100 000 compared
photogrammetry, close-range photogrammetry and TLS to
to UAVs (EUR 3500 for our platform) and DSLR (digital
assess their quality and evaluate their potential in mapping
single-lens reflex) cameras used in photogrammetry, in the
and describing glacier hazards such as ring faults and normal
EUR 500–3500 range.
faults, in a specific campaign carried out in summer 2016. In
addition, we employed orthophotos and point clouds from a
6.4 Additional remarks
UAV survey conducted in 2014 to analyze the evolution of
glacier hazards, as well as a DEM from an aerial photogram-
In summary, although TLS point clouds are regarded as the
metric survey conducted in 2007, to investigate glacier thick-
most accurate (Naumann et al., 2013), they suffer from in-
ness changes between 2007 and 2016. The main findings of
homogeneous point density and cumbersome logistics, and
our study are as follows:
their potential in glacial environments is limited, unless a
maximum uncertainty of 5–10 cm can be tolerated. Laser – UAVs and terrestrial photogrammetric surveys provide
scanners are also employed on aerial platforms, including reliable performances in glacial environments and out-
UAVs, where they can reconstruct terrain morphology with perform TLS in terms of logistics and costs.
only slightly higher uncertainty than the terrestrial counter-
parts with a much greater coverage (Rayburg et al., 2009), – UAV and terrestrial photogrammetric blocks can be eas-
but the high operational cost has limited the diffusion of ily integrated providing more information than individ-
this technique. Lastly, photogrammetry from higher-altitude ual techniques to help identify glacier hazards.
aerial platforms (mostly planes, but also helicopters and
– UAV-based DEMs can be employed to estimate thick-
satellites) can similarly achieve low uncertainty (3 m; An-
ness and volume changes but improvements are neces-
dreassen et al., 2002) and extensive coverage at the price of
sary in terms of area covered and to reduce uncertainty.
a lower spatial resolution compared to UAVs (e.g., 2 m in
our case), and due to its popularity in the past it is often the – The Forni Glacier is rapidly collapsing with an increase
only means to acquire good quality archive data to investi- in ring fault sizes, providing evidence of climate change
gate glacier changes over broad timescales (Andreassen et in the region.
al., 2002; Moelg and Bloch, 2017).
In our pilot study, we covered part of the Forni Glacier – The glacier thinning rate increased due to collapses to
tongue, and investigated different techniques to map/monitor 5.20 ± 1.11 m a−1 between 2014 and 2016.
hazards related to the glacier collapse. Our maps can help

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1068 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

The maps produced from the combined analysis of UAV and Edited by: Paolo Tarolli
terrestrial photogrammetric point clouds and orthophotos can Reviewed by: three anonymous referees
be made available through GIS web portals of the Stelvio Na-
tional Park or the Lombardy region (http://www.geoportale.
regione.lombardia.it/). A permanent monitoring programme
should be set up to help manage risk in the area, issuing References
warnings and assisting mountain guides in changing hiking
and ski routes as needed. The analysis of glacier thickness Abellán, A., Oppikofer, T., Jaboyedoff, M., Rosser, N. J.,
changes suggests a feedback mechanism which should be Lim, M., and Lato, M. J.: Terrestrial laser scanning of
further analyzed, with higher thinning rates leading to in- rock slope instabilities, Earth Surf. Proc. Land., 39, 80–97,
creased occurrence of collapses. Glacier downwasting is also https://doi.org/10.1002/esp.3493, 2014.
of relevance for risk management in the protected area, pro- Aicardi, I., Chiabrando, F.,Grasso, N., Lingua, A. M., Noardo, F.,
viding valuable data to assess the increased chance of rock- and Spanò, A.: UAV photogrammetry with oblique images: first
falls and to improve forecasts of glacier meltwater produc- analysis on data acquisition and processing, in: International
tion. Archives of the Photogrammetry, Remote Sensing and Spatial In-
formation Sciences, 12–19 July 2016, Prague, Czech Republic,
While our test was conducted on one of the largest glaciers
41-B1, 835–842, https://doi.org/10.5194/isprs-archives-XLI-B1-
in the Italian Alps, the integrated photogrammetric approach 835-2016, 2016.
is easily transferable to similarly sized and much smaller Andreassen, L. M., Hallgeir, E., and Kjollmoen, B.: Using aerial
glaciers, where it would be able to provide a comprehen- photography to study glacier changes in Norway, Ann. Glaciol.,
sive assessment of hazards and thickness changes and be- 34, 343–348, https://doi.org/10.3189/172756402781817626,
come useful in decision support systems for natural hazard 2010.
management. In larger regions, UAVs hold the potential to Azzoni, R. S., Fugazza, D., Zennaro, M., Zucali, M., D’Agata, C.,
become the platform of choice, but their performances and Maragno, D., Cernuschi, M., Smiraglia, C., and Diolaiuti, G.
cost-effectiveness compared to aerial and satellite surveys A.: Recent structural evolution of Forni Glacier tongue (Ortles-
must be further evaluated. Cevedale Group, Central Italian Alps), J. Maps, 13, 870–878,
https://doi.org/10.1080/17445647.2017.1394227, 2017.
Berthier, E., Arnaud, Y., Kumar, R., Ahmad, S., Wagnon,
P., and Chevallier, P.: Remote sensing estimates of glacier
Data availability. All data except for the 2007 photogrammet-
mass balances in the Himachal Pradesh (Western Hi-
ric DEM were acquired by the authors. Data can be re-
malaya, India), Remote Sens. Environ., 108, 327–338,
quested by email from the authors [email protected] or
https://doi.org/10.1016/j.rse.2006.11.017, 2007.
[email protected].
Berthier, E., Cabot, V., Vincent, C., and Six, D.: Decadal
Region-Wide and Glacier-Wide Mass Balances Derived from
Multi-Temporal ASTER Satellite Digital Elevation Mod-
Competing interests. The authors declare that they have no conflict els.Validation over the Mont-Blanc Area, Front. Earth Sci., 4, 63,
of interest. https://doi.org/10.3389/feart.2016.00063, 2016.
Bhardwaj, A., Sam, L., Akanksha, Martin-Torres, F. J., and Ku-
mar, R.: UAVs as remote sensing platform in glaciology: Present
Special issue statement. This article is part of the special issue applications and future prospects, Remote Sens. Environ., 175,
“The use of remotely piloted aircraft systems (RPAS) in monitoring 196–204, https://doi.org/10.1016/j.rse.2015.12.029, 2016.
applications and management of natural hazards”. It is a result of the Blasone, G., Cavalli, M., and Cazorzi, F.: Debris-Flow Monitor-
EGU General Assembly 2016, Vienna, Austria, 17–22 April 2016. ing and Geomorphic Change Detection Combining Laser Scan-
ning and Fast Photogrammetric Surveys in the Moscardo Catch-
ment (Eastern Italian Alps), in: Engineering Geology for So-
Acknowledgements. This study was funded by DARA, the De- ciety and Territory, Vol. 3, edited by: Lollino, G., Arattano,
partment for Autonomies and Regional Affairs of the Italian gov- M., Rinaldi, M., Giustolisi, O., Marechal, J. C., and Grant,
ernment’s Presidency of the Council of Ministers. The authors G., Springer, Cham, 51–54, https://doi.org/10.1007/978-3-319-
are grateful to the central scientific committee of CAI (Club 09054-2_10, 2015.
Alpino Italiano – Italian Alpine Club) and Levissima Sanpellegrino Carey, M., McDowell, G., Huggel, C., Jackson, M., Portocar-
S.P.A. for funding the UAV quadcopter. The authors also thank rero, C., Reynolds, J. M., and Vicuña, L.: Integrated ap-
Stelvio Park Authority for the logistic support and for permitting the proaches to adaptation and disaster risk reduction in dy-
UAV surveys and IIT Regione Lombardia for the provision of the namic sociocryospheric systems, in: Snow and Ice-related
2007 DEM. Our gratitude also goes to the GICARUS lab of Politec- Hazards, Risks and Disasters, edited by: Haeberli, W. and
nico Milano at Lecco Campus for providing the survey equipment. Whiteman, C., Elsevier, Amsterdam, the Netherlands, 219–261,
Finally, the authors would also like to thank Tullio Feifer, Livio Pi- https://doi.org/10.1016/B978-0-12-394849-6.00008-1, 2014.
atta, and Andrea Grossoni for their help during field operations and Chandler, J. H. and Buckley, S.: Structure from motion (SFM) pho-
the reviewers for their helpful comments. togrammetry vs terrestrial laser scanning, in: Geoscience Hand-
book 2016, AGI Data Sheets, 5th Edn., Section 20.1, edited by:

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1069

Carpenter, M. B. and Keane, C. M., American Geosciences In- Fugazza, D., Senese, A., Azzoni, R. S., Smiraglia, C., Cernuschi,
stitute, Alexandria, USA, 2016. M., Severi, D., and Diolaiuti, G. A.: High-resolution mapping
Chiarle, M., Iannotti, S., Mortara, G., and Deline, P.: Re- of glacier surface features. The UAV survey of the Forni glacier
cent debris flow occurrences associated with glaciers (Stelvio national park, Italy), Geografia Fisica e Dinamica Qua-
in the Alps, Global Planet. Change, 56, 123–136, ternaria, 38, 25–33, https://doi.org/10.4461/GFDQ.2015.38.03,
https://doi.org/10.1016/j.gloplacha.2006.07.003, 2007. 2015.
Clague, J.: Glacier Hazards, in: Encyclopedia of Natural Gagliardini, O., Gillet-Chaulet, F., Durand, G., Vincent, C., and Du-
Hazards, edited by: Bobrowski, P., Springer, Dordrecht, val, P.: Estimating the risk of glacier cavity collapse during arti-
the Netherlands, 400–405, https://doi.org/10.1007/978-1-4020- ficial drainage: The case of Tête Rousse Glacier, Geophys. Res.
4399-4_156, 2013. Lett., 38, L10505, https://doi.org/10.1029/2011GL047536, 2011.
Colomina, I. and Molina, P.: Unmanned aerial sys- Garavaglia, V., Diolaiuti, G. A., Smiraglia, C., Pasquale, V., and
tems for photogrammetry and remote sensing: A re- Pelfini, M.: Evaluating Tourist Perception of Environmental
view, ISPRS J. Photogram. Remote Sens., 92, 79–97, Changes as a Contribution to Managing Natural Resources in
https://doi.org/10.1016/j.isprsjprs.2014.02.013, 2014. Glacierized areas: A Case Study of the Forni Glacier (Stelvio
D’Agata, C., Bocchiola, D., Maragno, D., Smiraglia, C., and Dio- National Park, Italian Alps), Environ. Manage., 50, 1125–1138,
laiuti, G. A.: Glacier shrinkage driven by climate change during https://doi.org/10.1007/s00267-012-9948-9, 2012.
half a century (1954–2007) in the Ortles-Cevedale group (Stelvio Gardent, M., Rabatel, A., Dedieu, J.-P., and Deline, P.: Mul-
National Park, Lombardy, Italian Alps), Theor. Appl. Cima- titemporal glacier inventory of the French Alps from the
tol., 116, 169–190, https://doi.org/10.1007/s00704-013-0938-5, late 1960s to the late 2000s, Global Planet. Change, 120, 24–37,
2014. https://doi.org/10.1016/j.gloplacha.2014.05.004, 2014.
Dall’Asta, E., Thoeni, K., Santise, M., Forlani, G., Giacomini, A., Gindraux, S., Boesch, R., and Farinotti, D.: Accuracy As-
and Roncella, R.: Network design and quality checks in auto- sessment of Digital Surface Models from Unmanned Aerial
matic orientation of close-range photogrammetric blocks, Sen- Vehicles’ Imagery on Glaciers, Remote Sensing, 9, 2–15,
sors, 15, 7985–8008, https://doi.org/10.3390/s150407985, 2015. https://doi.org/10.3390/rs9020186, 2017.
Dewez, T. J. B., Leroux, J., and Morelli, S.: Cliff collapse haz- Gobiet, A., Kotlarski, S., Beniston, M., Heinrich, G., Rajczak,
ard from repeated multicopter UAV acquisitions: return on ex- J., Stoffel, M.: 21st century climate change in the Euro-
perience, in: The International Archives of the Photogrammetry, pean Alps – A review, Sci. Total Environ., 493, 1138–1151,
Remote Sensing and Spatial Information Sciences, XXIII IS- https://doi.org/10.1016/j.scitotenv.2013.07.050, 2014.
PRS Congress, 12–19 July 2016, Prague, Czech Republic, Harris, C., Arenson, L. U., Christiansen, H. H., Etzelmueller,
41-B5, 805–811, https://doi.org/10.5194/isprs-archives-XLI-B5- B., Frauenfelder, R., Gruber, S., Haeberli, W., Hauck, C.,
805-2016, 2016. Hoelzle, M., Humlum, O., Isaksen, K., Kaab, A., Kern-
Diolaiuti, G. A. and Smiraglia, C.: Changing glaciers Luetschg, M., Lehning, M., Matsuoka, N., Murton, J. B.,
in a changing climate: how vanishing geomorphosites Noetzli, J., Phillips, M., Ross, N., Seppaelae, M., Spring-
have been driving deep changes in mountain land- man, S. M., and Vonder Muehll, D.: Permafrost and climate
scapes and environments, Géomorphologie, 2, 131–152, in Europe: Monitoring and modelling thermal, geomorpholog-
https://doi.org/10.4000/geomorphologie.7882, 2010. ical and geotechnical responses, Earth-Sci. Rev., 92, 117–171,
Diolaiuti, G. A., Bocchiola, D., D’Agata, C., and Smiraglia, C.: https://doi.org/10.1016/j.earscirev.2008.12.002, 2009.
Evidence of climate change impact upon glaciers’ recession Hoffmann-Wellenhof, B., Lichtenegger, H., and Wasle, E.: GNSS
within the Italian Alps, Theor. Appl. Climatol., 109, 429–445, – GPS, GLONASS, Galileo & more, Springer, Vienna, Austria,
https://doi.org/10.1007/s00704-012-0589-y, 2012. https://doi.org/10.1007/978-3-211-73017-1, 2008.
Eltner, A., Kaiser, A., Castillo, C., Rock, G., Neugirg, F., and Abel- Immerzeel, W. W., Kraaijenbrink, P. D. A., Shea, J. M., Shrestha,
lán, A.: Image-based surface reconstruction in geomorphometry A. B., Pellicciotti, F., Bierkens, M. F. P., and de Jong, S. M.:
– merits, limits and developments, Earth Surf. Dynam., 4, 359– High-resolution monitoring of Himalayan glacier dynamics us-
389, https://doi.org/10.5194/esurf-4-359-2016, 2016. ing unmanned aerial vehicles, Remote Sens. Environ., 150, 93–
Fey, C. and Wichmann, V.: Long-range Terrestrial laser scan- 103, https://doi.org/10.1016/j.rse.2014.04.025, 2014.
ning for geomorphological change detection in alpine terrain – Janke, J. R.: Using airborne LiDAR and USGS DEM data for as-
handling uncertainties, Earth Surf. Proc. Land., 42, 789–802, sessing rock glaciers and glaciers, Geomorphology, 195, 118–
https://doi.org/10.1002/esp.4022, 2016. 130, https://doi.org/10.1016/j.geomorph.2013.04.036, 2013.
Fischer, M., Huss, M., Barboux, C., and Hoelzle, M.: The Jokinen, O. and Geist, T.: Accuracy aspects in topographical
new Swiss Glacier Inventory SGI2010: relevance of us- change detection of glacier surface, in: Remote sensing of
ing high-resolution source data in areas dominated by glaciers, CRC Press/Balkema, Leiden, the Netherlands, 269–
very small glaciers, Arct. Antarct. Alp. Res., 46, 933–945, 283, https://doi.org/10.1201/b10155-15, 2010.
https://doi.org/10.1657/1938-4246-46.4.933, 2014. Kaab, A., Huggel., C., Fischer, L., Guex, S. Paul, F., Roer., I., Salz-
Fischer, M., Huss, M., and Hoelzle, M.: Surface elevation and mass mann, N., Schlaefli, S., Schmutz, K., Schneider, D., Strozzi, T.,
changes of all Swiss glaciers 1980–2010, The Cryosphere, 9, and Weidmann, Y.: Remote sensing of glacier- and permafrost-
525–540, https://doi.org/10.5194/tc-9-525-2015, 2015. related hazards in high mountains: an overview, Nat. Hazards
Forlani, G., Pinto, L., Roncella, R., and Pagliari, D.: Terrestrial pho- Earth Syst. Sci., 5, 527–554, https://doi.org/10.5194/nhess-5-
togrammetry without ground control points, Earth Sci. Inform., 527-2005, 2005a.
7, 71–81, https://doi.org/10.1007/s12145-013-0127-1, 2014.

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

1070 D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution

Kaab, A., Reynolds, J. M., and Haeberli, W.: Glacier and Per- Palomo, I.: Climate Change Impacts on Ecosystem Services
mafrost hazards in high mountains, in: Global Change and in High Mountain Areas: A Literature Review, Mount. Res.
Mountain Regions. Advances in Global Change Research, edited Dev., 37, 179–187, https://doi.org/10.1659/MRD-JOURNAL-D-
by: Huber U. M., Bugmann H. K. M., and Reasoner, M. A., 16-00110.1, 2017.
Springer, Dordrecht, 225–234, https://doi.org/10.1007/1-4020- Piermattei, L., Carturan, L., and Guarnieri, A.: Use of terrestrial
3508-X_23, 2005b. photogrammetry based on structure from motion for mass bal-
Kaufmann, V. and Ladstädter, R.: Application of terrestrial pho- ance estimation of a small glacier in the Italian Alps, Earth Surf.
togrammetry for glacier monitoring in Alpine environments, in: Proc. Land., 40, 1791–1802, https://doi.org/10.1002/esp.3756,
International Archives of the Photogrammetry, Remote Sensing 2015.
and Spatial Information Sciences, Beijing, China, 37-B8, 813– Piermattei, L., Carturan, L., de Blasi, F., Tarolli, P., Dalla Fontana,
818, 2008. G., Vettore, A., and Pfeifer, N.: Suitability of ground-based SfM–
Kaufmann, V. and Seier, G.: Long-term monitoring of glacier MVS for monitoring glacial and periglacial processes, Earth
change at Gössnitzkees (Austria) using terrestrial photogram- Surf. Dynam., 4, 325–443, https://doi.org/10.5194/esurf-4-425-
metry, in: The International Archives of the Photogrammetry, 2016, 2016.
Remote Sensing and Spatial Information Sciences, XXIII IS- Pomerleau, F., Colas, F., Siegwart, R., and Magnenat, S.: Compar-
PRS Congress, 12–19 July 2016, Prague, Czech Republic, ing ICP variants on real world data sets, Autonomous Robots, 34,
41-B8, 495–502, https://doi.org/10.5194/isprs-archives-XLI-B8- 133–148, https://doi.org/10.1007/s10514-013-9327-2, 2013.
495-2016, 2016. Quincey, D. J., Lucas, R. M., Richardson, S. D., Glasser, N.
Keiler, M., Knight, J., and Harrison, S.: Climate change F., Hambrey, N. J., and Reynolds, J. M.: Optical remote
and geomorphological hazards in the eastern Euro- sensing techniques in high-mountain environments: applica-
pean Alps, Philos. T. Roy. Soc. A, 368, 2461–2479, tion to glacial hazards, Prog. Phys. Geogr., 29, 475–505,
https://doi.org/10.1098/rsta.2010.0047, 2010. https://doi.org/10.1191/0309133305pp456ra, 2005.
Kellerer-Pirklbauer, A., Bauer, A., and Proske, H.: Terrestrial Rayburg, S., Thoms, M., and Neave, M.: A compari-
laser scanning for glacier monitoring: Glaciation changes of son of digital elevation models generated from dif-
the Gößnitzkees glacier (Schober group, Austria) between 2000 ferent data sources, Geomorphology, 106, 261–270,
and 2004, in: 3rd Symposion of the Hohe Tauern National Park https://doi.org/10.1016/j.geomorph.2008.11.007, 2009.
for research in protected areas, 15–17 September 2005, castle of Riccardi, A., Vassena, G., Scotti, R., and Sgrenzaroli, M.: Recent
Kaprun, Austria, 97–106, 2005. evolution of the punta S. Matteo serac (Ortles-Cevedale Group,
Lague, D., Brodu, N., and Leroux, J.: Accurate 3D comparison of Italian Alps), Geografia Fisica e Dinamica Quaternaria, 33, 215–
complex topography with terrestrial laser scanner: application to 219, 2010.
the Rangitikei canyon (N-Z), J. Photogram. Remote Sens., 82, Rolstad, C., Haug, T., and Denby, B.: Spatially integrated
10–26, https://doi.org/10.1016/j.isprsjprs.2013.04.009, 2013. geodetic glacier mass balance and its uncertainty based
Matese, A., Toscano, P., Di Gennaro, S. F., Genesio, L., Vaccari, F. on geostatistical analysis: application to the western
P., Primicerio, J., Belli, C., Zaldei, A., Bianconi, R., and Gioli, B.: Svartisen ice cap, Norway, J. Glaciol., 55, 666–680,
Intercomparison of UAV, Aircraft and Satellite Remote Sensing https://doi.org/10.3189/172756409787769528, 2009.
Platforms for Precision Viticulture, Remote Sensing, 7, 2971– Rounce, D. R., Watson, C. S., and McKinney, D. C.: Identification
2990, https://doi.org/10.3390/rs70302971, 2015. of Hazard and Risk for Glacial Lakes in the Nepal Himalaya Us-
Moelg, N. and Bolch, T.: Structure-from-Motion Using Historical ing Satellite Imagery from 2000–2015, Remote Sensing, 9, 654,
Aerial Images to Analyse Changes in Glacier Surface Elevation, https://doi.org/10.3390/rs9070654, 2017.
Remote Sensing, 9, 1021, https://doi.org/10.3390/rs9101021, Ryan, J. C., Hubbard, A., Box, J. E., Brough, S., Cameron, K.,
2017. Cook, J. M., Cooper, M., Doyle, S. H., Edwards, A., Holt, T.,
Naumann, M., Geist, M., Bill, R., Niemeyer, F., and Grenzdo- Irvine-Fynn, T., Jones, C., Pitcher, L. H., Rennermalm, A. K.,
erffer, G.: Accuracy comparison of digital surface models cre- Smith, L. C., Stibal, M., and Snooke, N.: Derivation of High
ated by Unmanned Aerial Systems imagery and Terrestrial Laser Spatial Resolution Albedo from UAV Digital Imagery: Appli-
Scanner, in: International Archives of the Photogrammetry, Re- cation over the Greenland Ice Sheet, Front. Earth Sci., 5, 1–18,
mote Sensing and Spatial Information Sciences, UAV-g2013, https://doi.org/10.3389/feart.2017.00040, 2017.
4–6 September 2013, Rostock, Germany, 61-W2, 281–286, Seier, G., Kellerer-Pirklbauer, A., Wecht, M., Hirschmann, S.,
https://doi.org/10.5194/isprsarchives-XL-1-W2-281-2013, 2013. Kaufmann, V., Lieb, G. K., and Sulzer, W.: UAS-Based
Nuth, C. and Kaab, A.: Co-registration and bias corrections of satel- Change Detection of the Glacial and Proglacial Transition
lite elevation data sets for quantifying glacier thickness change, Zone at Pasterze Glacier, Austria, Remote Sensing, 9, 549,
The Cryosphere, 5, 271–290, https://doi.org/10.5194/tc-5-271- https://doi.org/10.3390/rs9060549, 2017.
2011, 2011. Senese, A., Diolaiuti, G. A., Mihalcea, C., and Smiraglia, C.:
Oborne, M.: Mission planner software, available at: http://ardupilot. Energy and Mass Balance of Forni Glacier (Stelvio Na-
org/planner/ (last access: 18 May 2017), 2013. tional Park, Italian Alps) from a Four-Year Meteorologi-
O’Connor, J., Smith, M. J., and James, M. R.: Cameras cal Data Record, Arct. Antarct. Alp. Res., 44, 122–134,
and settings for aerial surveys in the geosciences: op- https://doi.org/10.1657/1938-4246-44.1.122, 2012.
timising image data, Prog. Phys. Geogr., 41, 1–20, Smiraglia, C., Azzoni, R. S., D’Agata, C., Maragno, D., Fugazza,
https://doi.org/10.1177/0309133317703092, 2017. D., and Diolaiuti, G. A.: The evolution of the Italian glaciers from
the previous data base to the new Italian inventory. Preliminary

Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018 www.nat-hazards-earth-syst-sci.net/18/1055/2018/

D. Fugazza et al.: Combination of UAV and terrestrial photogrammetry to assess rapid glacier evolution 1071

considerations and results, Geogr. Fis. Dinam. Quat., 38, 79–87, Vincent, C., Thibert, E., Harter, M., Soruco, A., and Gilbert,
https://doi.org/10.4461/GFDQ.2015.38.08, 2015. A.: Volume and frequency of ice avalanches from Tacon-
Teunissen, P. J. G.: Testing theory. An introduction, in: Series on naz hanging glacier, French Alps, Ann. Glaciol., 56, 17–25,
Mathematical Geodesy and Positioning, VSSD Delft, Delft, the https://doi.org/10.3189/2015AoG70A017, 2015.
Netherlands, 2009. Westoby, M. J., Brasington, J., Glasser, N. F., Hambrey,
Urbini, S., Zirizzotti, A., Baskaradas, J. A., Tabacco, I. E., Cafarella, M. J., and Reynolds, J. M.: Structure-from-Motion’
L., Senese, A., Smiraglia, C., and Diolaiuti, G.: Airborne radio photogrammetry: A low-cost, effective tool for geo-
echo sounding (RES) measures on alpine glaciers to evaluate ice science applications, Geomorphology, 179, 300–314,
thickness and bedrock geometry: Preliminary results from pilot https://doi.org/10.1016/j.geomorph.2012.08.021, 2012.
tests performed in the ortles-cevedale group (Italian alps), Ann. Winkler, M., Pfeffer, W. T., and Hanke, K.: Kilimanjaro ice cliff
Geophys., 60, G0226, https://doi.org/10.4401/ag-7122, 2017. monitoring with close range photogrammetry, in: International
Vincent, C., Auclair, S., and Le Meur, E.: Outburst flood hazard Archives of the Photogrammetry, Remote Sensing and Spa-
for glacier-dammed Lac de Rochemelon, France, J. Glaciol., 56, tial Information Sciences, XXII ISPRS Congress, 25 August–
91–100, https://doi.org/10.3189/002214310791190857, 2010. 1 September 2012, Melbourne, Australia, 39-B5, 441–446, 2012.

www.nat-hazards-earth-syst-sci.net/18/1055/2018/ Nat. Hazards Earth Syst. Sci., 18, 1055–1071, 2018

Copyright of Natural Hazards & Earth System Sciences is the property of Copernicus
Gesellschaft mbH and its content may not be copied or emailed to multiple sites or posted to a
listserv without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.