IEEE Geoscience and Remote Sensing Magazine - March 2023 - 85

comparatively cheap and operationally flexible. They can be
equipped with various sensors and rapidly collect data. Owing
to their small size and flexibility, highly specific temporal
data can be obtained for a specified purpose. UAV images
have a greater resolution than satellite remote sensing data
that flexibly balance high spatial resolution and large spatial
coverage, which not only fills the gaps in in situ data but also
compensates for the lack of spatiotemporal resolution in satellite
imagery.
In Antarctic polar research, UAVs have encountered several
issues and challenges. First, strong winds and cold temperatures
in the Antarctic significantly impact UAV missions.
The wind resistance and control skills of UAVs were tested for
strong winds. New algorithms are being developed to adapt
to strong winds in polar settings and obtain precise flight
routes during wind disruption. Before UAV flight operation,
it is critical to collect and evaluate meteorological data (historical
and present) from the study area. UAVs are adversely
affected by low temperatures. An appropriate flight duration
should be considered when designing the flight path, and the
physical insulation of the battery should be improved. Furthermore,
the uneven distribution of solar energy caused by
polar days and nights necessitates a high exposure balance
of the lens throughout the image gathering process. Snowfall
alters the ice surface, which has a major impact on the success
rate of the reconstruction of the data processing algorithm.
The long-term and continual weathering impact of katabatic
wind makes snow more complex than fresh snow, affecting
the task's development and product processing (such as multitime
matching). These factors should be considered during
future UAV missions [8], [14], [17].
Second, the present UAV navigation system in polar regions
relies on more than just the GPS. The reduced cost of
ground station receivers and semiconductor components for
UAV positioning, speed, and timing has led to the joint use of
new GNSSs [such as Beidou, Galileo, and the Global Navigation
Satellite System (GLONASS)] and GPS [115]. The NASA
Global Differential GPS service is extremely useful for polar,
cryospheric, and future UAV research, although it still has to
be validated in the Antarctic. Fixed-wing platforms are challenging
to employ in rugged mountainous regions because
of takeoff and landing constraints [8]. Multirotor UAVs are
simple to operate and can take off and land in hilly locations
with a limited operational range and low altitude. However,
this type of UAV has a restricted operational range (generally
<10 km) when flying at low altitudes. Small fixed-wing UAVs,
which are reasonably straightforward to operate, can increase
this range four times. Large fixed-wing UAVs can fly higher
and farther (>100 km); however, they require specific training
and are difficult to land in mountainous locations. They
are better suited for surveying extensive glacial areas (usually
>10 km2) [116], [117]. Vertical takeoff and landing (VTOL)
UAVs are multirotor and fixed-wing aircraft hybrid platforms
that combine the benefits of both platforms in a single package.
With long-term endurance and high speed, they can
take off and land as smoothly and accurately as multirotor
MARCH 2023 IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE
aircraft. Thus, VTOL UAVs will provide a convenient platform
for future polar environment monitoring, allowing research
in more distant and larger regions [118]. Suitable UAV types
should be chosen for aerial photography missions based on
the topography and climate of the research area. Although
UAVs have better security and logistical support than fixedwing
aircraft and helicopters, there are still hazards. For example,
UAVs may interact with ground operators and other
people in the vicinity of installations. The radio transmission
of a station and other electromagnetic signals may also interfere
with UAV control.
Furthermore, factors such as flight design, camera quality,
camera calibration, SfM algorithms, and the georeferencing
approach can impact the geometrical accuracy of the georeferenced
digital surface models generated by UAV SfM photogrammetry
[119]. The accuracy of the geometric positioning
of UAV remote sensing is critical for highly accurate scientific
applications. The SfM technique is extensively used for UAV
airborne data processing because of its efficiency in challenging
environmental conditions, such as strong winds, cold
temperatures, poor light, and anomalous geomagnetism,
in polar regions, and it provides trustworthy topographic
outputs [120]. However, a deficiency exists in the accuracy
evaluation of the data products produced by this method.
An accurate method for evaluating the correctness of the SfM
algorithm is to employ a well-distributed ground control
point (GCP) [110]. Because of the hazards of crack propagation
and the lack of solid nonmoving components on the
ice sheet surface, the actual layout and data collection of the
GCP in the Antarctic environment can be difficult, hazardous,
time-consuming, and expensive [26]. Photogrammetry
is an essential tool for improving the external positioning and
ground control of UAVs and evaluating the quality of future
SfM data products. Li et al. [110] used onboard GPS and inertial
measurement unit (IMU) data with a direct georeferencing
method to process UAV data without ground control,
demonstrating the clear logistical benefits of UAV surveys in
dangerous locations, such as Antarctic sea ice. However, in the
absence of a dense and well-distributed ground control network,
errors in the onboard GPS record propagate to the final
IEEE Geoscience and Remote Sensing Magazine - March 2023

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