IEEE Geoscience and Remote Sensing Magazine - March 2023 - 82

demonstrating the superiority of UAVs in this field [42],
[61], [83], [84], [85], [86]. Machine learning support vector
regression was used to infer Antarctic moss vigor from
quantitative remote sensing maps of plant canopy chlorophyll
content and leaf density [84]. Additionally, semiautomatic
object-based image analysis was deployed in conjunction
with UAV images to assess moss health and cover
through time to track the changing state of Antarctic mosses
[87]. Accurate spatiotemporal mapping of vegetation in the
ice-free zone of Antarctica provided crucial information for
quantitatively characterizing the ecosystem's development
[88]. A novel methodology based on the integration of multiplatform
data (satellite and UAV) was used to accurately
quantify recent changes in the vegetated areas of the Fildes
Peninsula. Two robust supervised classifiers, a support vector
machine and k-nearest neighbor, were created and calibrated
to be used in situations where field reference data are
scarce or not available [88].
MP
~120 m
~183 m
SR
20170120
(a)
(b)
20170130
(c)
Ice Cave
Ch
SR
20170209
20190114
(d)
~210 m
In recent years, studies on cyanobacterial mats in the McMurdo
Dry Valley have increased. The presence, type, and spatial
interannual variations in cyanobacterial mats can be detected
using UAV remote sensing applications and geographic
information system technology [89], [90]. UAV, Sentinel-2, and
Landsat 8 images have been used to detect vegetation zones
and show that various sensors display varied normalized difference
vegetation index (NDVI) values for the same vegetation
type. The NDVI classification of the region indicates that
vegetation coverage matches the area manually assessed in
UAV images, suggesting that aerial and satellite images may
both be applied to monitor Antarctic vegetation [40].
UNMANNED AERIAL VEHICLES IN FAUNA STUDIES
The most important fauna of Antarctica is penguins, which
are primarily found in the coastal regions. A variety of special
UAVs have been designed considering platform design,
external layout selection, control devices, structural design,
20170120
20170131
Aerial Perspective View
(e)
(f)
UAV Aerial Image
20190114
20170209
Vertical Exaggeration: Two Times
(h)
Aerial Perspective View
(g)
FIGURE 7. The change in the 2017 catastrophic subsidence of the Dålk Glacier, East Antarctica (based on Florinsky et al. [24], Zhang et al. [77],
and research from the 35th CHINARE). (a)-(d) High-resolution orthomosaics on 20 January 2017, 30 January 2017, 9 February 2017, and
14 January 2019, respectively. (e) Aerial perspective view on 20 January 2017. (f) UAV aerial image on 31 January 2017. (g) Aerial perspective
view on 9 February 2017. (h) Three-dimensional visualization on 14 January 2019 (vertical exaggeration in two times). SR: local sledge route;
MP: meltwater pond; Ch: a small drainage channel.
82
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE MARCH 2023
~200 m
~220 m
~220 m
Viewing
Direction in
Figure 7(h)
IEEE Geoscience and Remote Sensing Magazine - March 2023

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - March 2023
