Unmanned Aerial Survey for Modelling Glacier Topography in

Antarctica: First Results

Dmitrii Bliakharskii

1,2

and Igor Florinsky

Institute of Earth Sciences, St. Petersburg University, 199034, St. Petersburg, Russian Federation

Institute of Mathematical Problems of Biology,

Keldysh Institute of Applied Mathematics, Russian Academy of Sciences,

142290, Pushchino, Moscow Region, Russian Federation

Keywords: Unmanned Aerial System, Unmanned Aerial Vehicle, Photogrammetry, Digital Elevation Model, Crevasse.

Abstract: For Antarctic research, one of the most important support tasks is a rapid and safe monitoring of sledge

routes, snow / ice airfields, and other visited areas for detection of open crevasses, revealing of hidden,

snow-covered ones, as well as studying of their dynamics. We present the first results from a study of

applying unmanned aerial systems (UASs) and UAS-derived data to model glacier topography in contexts

of detecting crevasses and monitoring changes in glacier surfaces. The study was conducted in East

Antarctica in the austral summer 2016/2017. The surveyed areas included an eastern part of the Larsemann

Hills, an airfield of the Progress Station, an initial section of a sledge route from the Progress to Vostok

Stations, and a north-western portion of the Dålk Glacier before and after its collaps. The surveying was

performed by Geoscan 201, a flying-wing UAS. For the photogrammetric processing of imagery, we

applied software Agisoft PhotoScan Professional. High-resolution digital elevation models (DEMs) for

surveyed areas were produced. For the Dålk Glacier, we derived two DEMs related to the pre- and post-

collapsed glacier surface. A further analysis will be performed by methods of geomorphometry. The focus

will be on the revealing of crevasses.

1 INTRODUCTION

Unmanned aerial systems (UASs) are increasingly

used in sciences and industry (Aber et al., 2010;

Colomina and Molina, 2014; Shahbazi et al., 2014;

Pajares, 2015; Toth and Jóźków, 2016). In

particular, UAS-derived imagery is utilized for

producing high-resolution digital elevation models

(DEMs), which are then applied, e.g., to study and

model dynamics of slope, coastal, and fluvial

processes, to reveal fine geological features, etc.

(Smith et al., 2009; D’Oleire-Oltmanns et al., 2012;

Mancini et al., 2013; Bemis et al., 2014; Johnson et

al., 2014; Brunier et al., 2016; Florinsky et al.,

2018).

The use of UASs may significantly facilitate the

work of researchers in severe conditions of high

mountain glaciers and ice sheets (see a review of

UAS application in glaciology — Bhardwaj et al.,

2016). In Antarctica, UAS-derived data including

DEMs are used for geomorphological and vegetation

studies and surface evolution analysis (Lucieer et al.,

2014; Westoby et al., 2015; Dąbski et al., 2017).

Extreme temperature and meteorological conditions

of Antarctica make special demands on the

characteristics of UASs, installed equipment, and

aerial surveying techniques (Funaki and Hirasawa,

2008; Crocker et al., 2011; Goetzendorf-Grabowski

and Rodzewicz, 2017).

The glaciers of Antarctica are in continuous

motion. As a particular result, crevasses constantly

form on the surface and in the near-surface layer of

glaciers (Van der Veen, 1999; Colgan et al., 2016).

Crevasses pose a significant danger to polar

explorers. Thus, for Antarctic research, one of the

most important support tasks is a rapid and safe

monitoring of sledge routes, airfields (snow / ice

runways and skiways), and other visited areas for

detection of open crevasses, revealing of hidden

(snow-covered) ones, as well as studying of their

dynamics. For real-time detection and revealing of

crevasses, ground penetration radars are usually

applied, with antennas mounted in front of the

moving vehicle (for a detailed review of approaches

for crevasse detection, see Colgan et al., 2016, pp.

15–20). Although aerial imagery may be utilized to

Bliakharskii, D. and Florinsky, I.

Unmanned Aerial Survey for Modelling Glacier Topography in Antarctica: First Results.

DOI: 10.5220/0006812603190326

In Proceedings of the 4th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2018), pages 319-326

ISBN: 978-989-758-294-3

319

monitor crevasses and their dynamics, a rapid aerial

survey from manned aircrafts is not always possible

due to its high cost as well as weather and lighting

conditions. Moreover, aerial photographs of glacier

surfaces cannot always be interpreted because of

specific conditions for image formation, such as low

contrast of scenes, glare from snow and ice surfaces.

UASs introduce a low-cost alternative to manned

aerial survey. The processes of crevasse formation

and propagation are reflected in the geometry of a

glacier surface. So, to reveal crevasses and monitor

glacier surfaces, it might be possible to use

secondary products of unmanned aerial survey, that

is, DEMs derived from stereo images. In this paper,

we present the first results from a study of applying

unmanned aerial survey and UAS-derived data to

model glacier topography in Antarctica in contexts

of detecting crevasses and monitoring changes in

glacier surfaces under austral summer conditions.

2 STUDY AREAS

The study was conducted in and near the Larsemann

Hills, East Antarctica (Figure 1). The Larsemann

Hills are ice-free, low rounded hills with an area of

about 40 sq km located on the south-eastern coast of

the Prydz Bay, Princess Elizabeth Land (Stüwe et

al., 1989). This coastal oasis includes Stornes and

Broknes peninsulas, four small peninsulas, and

about 130 islands. The Larsemann Hills is one of the

Antarctic Specially Managed Area (ATCM, 2014).

There are three year-round operated polar stations in

this area: Progress (Russia), Zhongshan (China), and

Bharati (India).

There were four surveyed areas (Figure 1):

 An eastern part of the Larsemann Hills.

 An airfield (skiway) of the Progress Station

located on the ice plateau south of the

Larsemann Hills.

 An initial section (from 69°27'24" S, 76°19'04"

E to 69°40'40" S, 76°32'46" E) of a sledge

route from the Progress to Vostok Stations.

 A north-western portion of the Dålk Glacier.

The Dålk Glacier is a 15-km long, polar outlet

glacier draining into the south-eastern part of the

Prydz Bay east of the Larsemann Hills. On January

30, 2017, a rare natural catastrophic phenomenon

occurred: a wide collapsed depression formed in the

north-western margin of this glacier, at coordinates

69°23'58" S, 76°24'49" E (Popov et al., 2017).

3 UAS CHARACTERISTICS

The surveying was performed by Geoscan 201

(Figure 2), a flying-wing UAS, with a wingspan of

2.3 m and a takeoff weight of 8 kg, launching by a

catapult and landing with a parachute (Geoscan,

2016). Geoscan 201 has an electric motor; the flight

duration is up to 2.5 hrs with a cruising speed of 110

km/h. On board there are:

 A modem for the telemetry communication

with a laptop ground control station (GCS).

 A visible-band digital camera Sony DSC-RX1

equipped with a Carl Zeiss Vario Sonnar T

lens (central shutter; 35 mm focal length) and

Figure 1: Geographical location of the surveyed areas o

the background of the Google Earth image mosaic: the

white frame LH — the eastern part of the Larsemann Hills

(see Figure 3); the black frame A — the airfield area (see

Figure 4); the white frame SR — the area of the initial

section of the sledge route (the red line) from the Progress

to Vostok Stations (see Figure 5); the red frame marks the

area of a collapsed portion of the Dålk Glacier (see Figure

7); DG — Dålk Glacier, PB — Prydz Bay.

GISTAM 2018 - 4th International Conference on Geographical Information Systems Theory, Applications and Management

320

a 35.8 × 23.9 mm sensor (a matrix of 6000 ×

4000 pixels with pixel sizes of 6 × 6 μm).

 A GNSS receiver Topcon for high-precision

determination of image projection centres.

A ground GNSS base station includes a receiver

Topcon HiPer V.

To consider the lens distortion, we previously

performed a laboratory photogrammetric calibration

of the camera, as well as a field camera calibration.

The following camera settings are used during an

aerial survey: (1) the lens is locked to focus to

infinity; (2) the shutter priority are 1/1000 and 1/800

seconds for sunny and cloudy weather,

correspondingly; (3) an aperture and ISO sensitivity

values are selected automatically, one time per a

flight stripe; (4) images are recorded as JPEG files.

The combination of the mentioned camera and

GNSS receiver allows determining image projection

centres with the planimetric and vertical accuracies

of 2 cm and 3–4 cm, correspondingly.

The Geoscan 201 flight is performed in an

automatic mode using an autopilot. One should

specify a flight mission in the GCS by selecting a

surveyed area and setting flight parameters. The key

parameter is an overlap between images. It is

recommended to set 70 % forward and 50 % side

overlaps. Then, one of the three interrelated

parameters should be set: ground sample distance,

flight altitude, and strip width. Flights can be

conducted either at a constant altitude above ground

level (AGL) or a constant altitude above sea level.

We used the first option. Finally, a GCS program

calculates the flight stripes and coordinates of image

capture positions.

Before sending the Geoscan 201 to Antarctica, it

was modernized as follows: (1) a filament was built

into the camera lens to evaporate the possible

condensate when passing through 0° С; (2) a self-

heating function was embedded in the rechargeable

batteries; and (3) the rubber catapult was

supplemented by a spring.

4 AERIAL SURVEYING

The aerial survey was conducted in January and

February 2017 under various meteorological

conditions including clear sunny weather, variable

cloudiness, and overcast. The air temperature ranged

from –1 °C to –12 °C. The wind speed varied from 0

m/s to 28 m/s at the flight altitude. On and near the

coast, the wind was northeast. On the ice sheet, the

wind was south katabatic in morning and north

afternoon. General flight characteristics are

summarized in Table 1.

The eastern part of the Larsemann Hills with an

area of about 67 sq km (Figure 3a) was surveyed in

four flights (Table 1). The weather was sunny; the

air temperature ranged from –1 °С to –5 °С; the

wind speed was up to 5 m/s at the flight altitude.

Totally, the aerial survey consisted of 99 flight

stripes including 5,451 images.

Figure 2: A UAS Geoscan 201 (Geoscan, 2016).

Table 1: Flight characteristics.

Date Flight

Flight

altitude (m

AGL)

Number of

flight

stripes

Number of

images

Larsemann Hills

14.01 2 370–410 19 1,089

–"– 3 –"– 26 1,230

20.01 14 310–350 20 1,594

–"– 15 –"– 34 1,538

Total 99 5,451

Airfield

14.01 1 520 16 856

14.02 30 390 17 1,271

Sledge route

17.01 5 400–430 26 1,280

–"– 6 –"– 32 1,484

–"– 7 300–320 23 1,428

–"– 8 –"– 25 1,260

18.01 9 –"– 16 766

–"– 10 –"– 11 535

–"– 11 –"– 24 1,025

–"– 12 –"– 17 800

–"– 13 400–430 21 803

Total 195 9,381

03.02 20 400–430 21 813

–"– 21 –"– 21 1,053

04.02 23 300–320 29 1,400

–"– 24 –"– 26 1,614

–"– 25 –"– 29 1,316

Total 126 6,196

Dålk Glacier, collapsed area

20.01 14 310–350 20 1,594

09.02 29 320–330 30 1,011

Unmanned Aerial Survey for Modelling Glacier Topography in Antarctica: First Results

321

The airfield area of about 10 sq km (Figure 4a)

was surveyed twice — in mid-January and mid-

February — to estimate the snow melting (Table 1).

During the first flight, the weather was sunny; the air

temperature ranged from –3 °С to –5 °С; the wind

speed was 3–5 m/s at the flight altitude. The first

survey consisted of 16 flight stripes including 856

images. During the second flight, it was variable,

mid-level cloudiness; the air temperature was –8 °С;

the wind speed was about 10 m/s at the flight

altitude. The second survey consisted of 17 flight

stripes including 1,271 images.

The area of the initial section of the sledge route

from the Progress to Vostok Stations with a length

of about 30 km and a width of around 3 km (Figure

5a) was surveyed twice — in mid-January and early

February — to estimate the ice movement. As

ground control points, we used 17 plastic stakes and

10 barrels. During the first and second surveys, we

performed nine and five flights, correspondingly

(Table 1). During the first survey, the weather was

sunny; the air temperature ranged m/s gusting up to

18–20 m/s at the flight altitude. The surveyed area

was about 102 sq km. The first survey consisted of

195 flight stripes including 9,381 images. During the

second survey, weather conditions deteriorated: it

Figure 3: The eastern part of the Larsemann Hills: (a) orthomosaic; (b) elevations. BP — Broknes Peninsula, PS

—

Progress Station, ZS — Zhongshan Station, BS — Bharati Station, DG — Dålk Glacier, PB — Prydz Bay, QB — Quilty

Bay, TF — Thala Fjord, CF — Clemence Fjord, NF — Nella Fjord, DI — Dalkøy Island; the white frame marks the are

of a collapsed portion of the Dålk Glacier (see Figure 7).

GISTAM 2018 - 4th International Conference on Geographical Information Systems Theory, Applications and Management

322

was overcast with a cloud base of about 1000 m; the

air temperature ranged from –7 °С to –12 °С; the

wind speed was 15–20 m/s gusting up to 25–28 m/s

at the flight altitude. In this connection, it was

impossible to perform aerial surveying of the entire

study area (the surveyed area was about 73 sq km).

The second survey consisted of 126 flight stripes

with 6,196 images (Figure 6).

The collapsed portion of the Dålk Glacier and an

adjacent territory with an area of about 7.5 sq km

(Figure 7a) was surveyed in two flights, before and

after the glacier’s collapse (Table 1). The weather

was sunny; the air temperature ranged from –1 °С to

–5 °С; the wind speed was 0–5 m/s at the flight

altitude. The aerial surveying consisted of 20 and 30

flight stripes including 1,594 and 1,011 images for

the first and second flights, correspondingly.

Each flight took about 2 hrs. For all flights,

forward and side overlaps were about 70 % and

50 %, correspondingly. Aerial images have the

average resolution of 6 cm.

We described only those aerial surveys, which

images were then successfully processed (Section 5).

Testing flights were also conducted. It has been

empirically found that in Antarctica, it is advisable

to perform unmanned aerial survey in sunny

weather, at variable and high cloudiness (5–6 km

AGL). On contrast images captured in such a good

lighting conditions, microtopographic ice features

created by the wind are clearly visible on the ice

surface. A correlation algorithm uses these ice

microfeatures for searching conjugate points in the

photogrammetric processing of images (Section 5).

Under low cloudiness and hence diffuse lighting,

one obtains low-contrast, almost white aerial

photographs, which are difficult or impossible to

process because ice microfeatures are not visible.

It is absolutely impossible to perform unmanned

aerial survey under low clouds (100–200 m AGL).

In this case, whiteout occurs when there is no

contrast between ice- and snow-covered terrain and

sky, that is, ice, sky, and horizon are

indistinguishable.

Sometimes, it was difficult to determine the

cloud height. As a result, the UAS went into the

clouds. In these cases, the UAS had ice build-up on

the leading edge of wing, Pitot tube, and screw

Figure 6: An example of aerial photographs of the sledge

route area. Arrows mark crevasses.

Figure 4: The airfield area (mid-January): (a) orthomosaic;

(b) elevations.

Figure 5: The area of the initial section of the sledge route

from the Progress to Vostok Stations (mid-January):

(a) elevations and thalweg lines; (b) ice movement

(meters). Arrows show the direction of movement.

Unmanned Aerial Survey for Modelling Glacier Topography in Antarctica: First Results

323

blades that led to an increase in UAS’s weight and

its emergency landing.

5 IMAGE PROCESSING

The image processing consisted of two stages: (1) a

post-processing of the on-board GNSS receiver

measurements relative to the base station; and (2) a

photogrammetric processing of the aerial images.

For the first stage, we used Pinnacle software

(Topcon Positioning Systems, Inc.). We positioned

images with corresponding coordinates of image

projection centres. The coordinates were utilized in

the next stage as reference data.

For the second stage, we used Agisoft PhotoScan

Professional 1.3.2 software (Agisoft, 2017). This

choice is determined by the fact that Geoscan 201

and Agisoft PhotoScan represent a unified software–

hardware system (Inozemtsev, 2014). This software

implements a set of traditional photogrammetric

Figure 7: The north-western margin of the Dålk Glacier before (left) and after (right) the surface collapse (for the location

of this area, see Figures 1 and 3): (a) orthomosaics; (b) elevations.

GISTAM 2018 - 4th International Conference on Geographical Information Systems Theory, Applications and Management

324

Table 2: Accuracy of the aerial triangulation in terms of

standard deviation.

Area

Planimetric (m)

Vertical

(m)

X Y

Larsemann Hills 0.09 0.12 0.08

Airfield 0.05 0.04 0.03

Sledge route 1.9 0.5 0.8

Dålk Glacier,

collapsed zone

0.05 0.05 0.06

Table 3: DEM characteristics.

Area Matrix size Resolution (m)

Larsemann Hills

68,913 × 41,879

0.28

Airfield

30,991 × 26,771

0.27

Sledge route

12,817 × 19,082

1.9

Dålk Glacier,

collapsed zone

8,400 × 7,655

0.22

methods (Kraus, 2007; Mikhailov and Chibunichev,

2016) and structure-from-motion approaches

(Hirschmüller, 2008; Smith et al., 2016).

The image processing is based on an algorithm

developed in Agisoft (2017). The brightness of

images is analysed, so that a large number of

conjugate points are defined on overlapping images.

Using conjugate points, the parameters of image

orientation and lens distortion are determined. Using

reference points, a photogrammetric model, derived

from all images of the block, becomes a metric one.

Thus, the block aerial triangulation is performed. For

accuracy of the aerial triangulation, see Table 2.

Densifying a cloud of conjugate points to a dense

point cloud using all obtained elements, one can

obtain an irregular digital surface model (DSM).

Then, one can derive a regular DSM from the

irregular one by linear interpolation. In a general

case, when the dense point cloud is obtained, one

can produce a DEM by applying a classification

algorithm developed by Agisoft (2017). Recall that a

DEM describes the topographic surface while a

DSM represents a surface enveloping or consisting

of portions of the topographic surface, vegetation

canopy, top of buildings, and other objects located

on the topographic surface. However, in this study, a

DSM can be considered a DEM for almost all areas

(except for very small zones including polar

station’s buildings).

Using obtained aerial imagery, we generated

high-resolution DEMs (Figures 3b, 4b, 5a, and 7b).

For the airfield area, we derived two DEMs related

to mid-January and mid-February dates of survey.

For the initial section of the sledge route, two DEMs

related to mid-January and early February dates of

survey were also produced. For the north-western

margin of the Dålk Glacier, we derived two DEMs

related to the pre- and post-collapsed glacier surface

(Figure 7b). General characteristics of the DEMs can

be found in Table 3. Notice that the DEMs of the

Larsemann Hills (Figure 3b) and the Dålk Glacier

(Figure 7b) describe mixed surfaces of the bedrock

and ice. The DEMs of the airfield area and the

sledge route describe the ice surface only.

Accuracy estimation of the obtained DEMs is a

non-trivial task. Usually, it can be solved using

GCPs with known planimetric coordinates and

elevation values. However, in this study it was

impossible to set up GCPs at all glacier areas due to

safety issues. At the sledge route, we set up 27

GCPs, but these cannot be used for the DEM

accuracy estimation because of a high speed of the

ice movement and a time gap between geodetic and

aerial works (GCPs moved together with the ice).

On the initial section of the sledge route, the ice

movement ranged from 0.28 m to 5.31 m in 21 days

between the first and second surveys (Figure 5b),

that is, up to 0.25 m per day. For a rough estimation

of the DEM accuracy, one may use data for the

accuracy of the aerial triangulation (Table 2).

6 CONCLUSIONS

We studied possibilities of applying a small UAS

and UAS-derived images to model glacier

topography in Antarctica. Optimal meteorological

conditions were empirically determined for

conducting unmanned aerial survey to obtain images

suitable for subsequent photogrammetric processing

and DEM generation. For four areas, high-resolution

DEMs of bedrock/glacier and glacier surfaces were

produced. Further analysis of the DEMs will be

performed by methods of geomorphometry

(Florinsky, 2016). The focus will be on a detailed

modelling of the collapsed area of the Dålk Glacier,

as well as the revealing of glacier crevasses.

ACKNOWLEDGEMENTS

The aerial surveying was conducted during the 62nd

Russian Antarctic Expedition in cooperation with

Geoscan Ltd. Data processing is supported by the

Russian Foundation for Basic Research, grant 17-37-

50011.

Unmanned Aerial Survey for Modelling Glacier Topography in Antarctica: First Results

325

REFERENCES

Aber, J.S., Marzolff, I., Ries, J., 2010. Small-format aerial

photography: Principles, techniques and geoscience

applications. Elsevier, Amsterdam.

Agisoft, 2017. Agisoft PhotoScan user manual:

Professional edition, version 1.3. St. Petersburg.

ATCM, 2014. Larsemann Hills, East Antarctica: Antarctic

Specially Managed Area Management Plan. Measure

15. In: Antarctic Treaty Consultative Meeting (ATCM)

XXXVII Final Report. ATCM.

Bemis, S.P., Micklethwaite, S., Turner, D., James, M.R.,

Akciz, S., Thiele, S.T., Bangash, H.A., 2014. Ground-

based and UAV-based photogrammetry: A multi-

scale, high-resolution mapping tool for structural

geology and paleoseismology. J. Struct. Geol., 69:

163-178.

Bhardwaj, A., Sam, L., Akanksha, Martín-Torres, F.J.,

Kumar, R., 2016. UAVs as remote sensing platform in

glaciology: Present applications and future prospects.

Remote Sens. Envir., 175: 196-204.

Brunier, G., Fleury, J., Anthony, E.J., Gardel, A.,

Dussouillez, P., 2016. Close-range airborne Structure-

from-Motion Photogrammetry for high-resolution

beach morphometric surveys: Examples from an

embayed rotating beach. Geomorph., 261: 76-88.

Colgan, W., Rajaram, H., Abdalati, W., McCutchan, C.,

Mottram, R., Moussavi, M.S., Grigsby, S., 2016.

Glacier crevasses: Observations, models, and mass

balance implications. Rev. Geophys., 54: 119-161.

Colomina, I., Molina, P., 2014. Unmanned aerial systems

for photogrammetry and remote sensing: A review.

ISPRS J. Photogramm. Remote Sens., 92: 79-97.

Crocker, R.I., Maslanik, J.A., Adler, J.J., Palo, S.E.,

Herzfeld, U.C., Emery, W.J., 2011. A sensor package

for ice surface observations using small unmanned

aircraft systems. IEEE Trans. Geosci. Remote Sens.,

50: 1033-1047.

Dąbski, M., Zmarz, A., Pabjanek, P., Korczak-Abshire,

M., Karsznia, I., Chwedorzewska, K.J., 2017. UAV-

based detection and spatial analyses of periglacial

landforms on Demay Point (King George Island,

South Shetland Islands, Antarctica). Geomorph., 290:

29-38.

D’Oleire-Oltmanns, S., Marzolff, I., Peter, K.D., Ries,

J.B., 2012. Unmanned aerial vehicle (UAV) for

monitoring soil erosion in Morocco. Remote Sens., 4:

3390-3416.

Florinsky, I.V., 2016. Digital terrain analysis in soil

science and geology. 2

ed. Academic Press,

Amsterdam.

Florinsky, I.V., Kurkov, V.M., Bliakharskii, D.P., 2018.

Geomorphometry from unmanned aerial surveys.

Trans. GIS, 22, doi:10.1111/tgis.12296.

Funaki, M., Hirasawa, N., 2008. Outline of a small

unmanned aerial vehicle (Ant-Plane) designed for

Antarctic research. Polar Sci., 2: 129-142.

Geoscan, 2016. Geoscan 201. Geoscan Ltd, St.Petersburg,

https://www.geoscan.aero/en/products/geoscan201/bas

Goetzendorf-Grabowski, T., Rodzewicz, M., 2017. Design

of UAV for photogrammetric mission in Antarctic

area. J. Aerospace Eng., 231: 1660-1675.

Hirschmüller, H., 2008. Stereo processing by semiglobal

matching and mutual information. IEEE Trans. Patt.

Anal. Mach. Intell., 30: 328-341.

Inozemtsev, D.P., 2014. Automated aerial survey using a

software–hardware system ‘Geoscan–Photoscan’.

CAD & GIS for Roads, 1(2), 46-51 (in Russian).

Johnson, K., Nissen, E., Saripalli, S., Arrowsmith, J.R.,

McGarey, P., Scharer, K., Williams, P., Blisniuk, K.,

2014. Rapid mapping of ultrafine fault zone

topography with structure from motion. Geosphere,

10: 969-986.

Kraus, K., 2007. Photogrammetry: Geometry from images

and laser scans, 2

ed. de Gruyter, Berlin.

Lucieer, A., Turner, D., King, D.H., Robinson, S.A., 2014.

Using an Unmanned Aerial Vehicle (UAV) to capture

micro-topography of Antarctic moss beds. Int. J. Appl.

Earth Observ. Geoinf., 27: 53-62.

Mancini, F., Dubbini, M., Gattelli, M., Stecchi, F., Fabbri,

S., Gabbianelli, G., 2013. Using unmanned aerial

vehicles (UAV) for high-resolution reconstruction of

topography: The structure from motion approach on

coastal environments. Remote Sens., 5: 6880-6898.

Mikhailov, A.P., Chibunichev, A.G. (2016).

Photogrammetry. MIIGAiK, Moscow (in Russian).

Pajares, G., 2015. Overview and current status of remote

sensing applications based on unmanned aerial

vehicles (UAVs). Photogramm. Eng. Remote Sens.,

81: 281-329.

Popov, S.V., Pryakhin, S.S., Bliakharskii, D.P., Pryakhina,

G.V., Tyurin, S.V., 2017. Vast ice depression in Dålk

Glacier, East Antarctica. Ice Snow, 57: 427-432.

Shahbazi, M., Théau, J., Ménard, P., 2014. Recent

applications of unmanned aerial imagery in natural

resource management. GISci. Remote Sens., 51: 339-

365.

Smith, M.J., Chandler, J., Rose, J., 2009. High spatial

resolution data acquisition for the geosciences: Kite

aerial photography. Earth Surf. Process. Landforms,

34: 155-161.

Smith, M.W., Carrivick, J.L., Quincey, D.J., 2016.

Structure from motion photogrammetry in physical

geography. Progr. Phys. Geogr., 40: 247-275.

Stüwe, K., Braun, H.-M., Peer, H., 1989. Geology and

structure of the Larsemann Hills area, Prydz Bay, East

Antarctica. Austral. J. Earth Sci., 36: 219-241.

Toth, C., Jóźków, G., 2016. Remote sensing platforms and

sensors: A survey. ISPRS J. Photogramm. Remote

Sens., 115: 22-36.

Van der Veen, C.J., 1999. Crevasses on glaciers. Polar

Geogr., 23: 213-245.

Westoby, M.J., Dunning, S.A., Woodward, J., Hein, A.S.,

Marrero, S.M., Winter, K., Sugden, D.E., 2015.

Sedimentological characterization of Antarctic

moraines using UAVs and Structure-from-Motion

photogrammetry. J. Glaciol., 61: 1088-1102.

GISTAM 2018 - 4th International Conference on Geographical Information Systems Theory, Applications and Management

326