Geography Reference
In-Depth Information
cover larger areas up to maximum areas approaching a
square kilometre. The first aerial photography from a
UAV was taken in 1955 for a military application (New-
come, 2004). However, UAVs are currently undergoing a
tremendous expansion. These systems generally employ
airframe technology developed in the field of model-
making and available for several decades. However, much
of their new-found popularity can be attributed to recent
technological developments enabling automated flight
and controlled image acquisition. Thanks to the miniatur-
isation of GPS technology which can measure the aircraft
position in 3D space and to the development of small
electronic Inertial Measurement Units (IMU) capable of
measuring aircraft orientation in 3D space, these systems
can now generally assist the user in manual piloting and
even fly autonomously. As a result, many systems are now
capable of pre-programmed flight paths with automated,
controlled, image acquisition. When combined to recent
developments in small format, cheap, digital cameras
intended for the wider consumer market, UAV platforms
now offer scientists and managers an opportunity for 'do-
it-yourself' remote sensing hyperspatial data acquisition.
For example, the entries in Table 8.1 relating to UAV
work illustrate the typical performance of these systems.
After proper training and adequate experience with the
system, pilots can expect to cover areas which exceed
100 ha (
=
0.1 km
2
). Flying altitudes are low which results
in image resolutions well within the hyperspatial range.
An important caveat which is not mentioned in Table 8.1
and which will be encountered by new UAV users is the
need to obey air traffic regulations. UAVs are increasingly
recognised in air traffic regulations and potential users
should examine these regulations with care as they place
constraints on UAV operations. In Europe the general
principle is that UAV operations are possible in unpopu-
lated areas if the pilot keeps the UAV within line of sight
(approx. 500 m) and flies at low altitudes. For example,
in the UK, small UAVs with weights below 7 kg can oper-
ate freely outside of urban areas provided that the pilot
maintains visual contact at all times and that the flight
altitude is below 123 metres (400 ft) (Defourny et al.,
2006). Consequently, UAV surveys are best targeted to
smaller study areas at the reach or small river scale. This
new, exciting, technology is not yet suited to catchment
scale data acquisition.
In a similar manner, piloted, Ultra-Light Aerial Vehi-
cles (ULAVs) are an excellent means to provide data
over slightly larger study areas. They do this, however,
at the cost of some of the flexibility and rapid-response
capability of their unmanned counterparts. ULAVs are
larger, more expensive, and require significantly more
skill, experience (as well as a fearless pilot!) than is
required for UAV work. As a result, it is common for
ULAV work to be tasked from existing private compa-
nies rather than river managers or research teams having
access to their own ULAV. Nevertheless, ULAV services
are offered by more specialised companies, at more afford-
able costs and with more flexible re-flight opportunities
than traditional aerial photographic surveys. There is
again a great variation within ULAV platforms includ-
ing a variety of light paragliders, small planes and small
helicopters. The difference in size between these and the
generally smaller UAVs provides additional advantages
in terms of greater carrying capacity for larger and more
complex sensors. Peer-reviewed publications making use
of ULAV-acquired data are still quite rare. However, in
Table 8.1, we see that Hervouet et al. (2011) report spa-
tial resolutions ranging from 3-15 cm and study areas
approaching 0.5 km
2
.
8.2.2.2 Traditional aerial photographic surveys
In addition to these new technologies, traditional aerial
photography also offers hyperspatial imagery and remains
an important data acquisition platform. Conventional
planes and helicopters are able to cover very large areas
and traditional aircraft still remain the only sensible
option for study areas which significantly exceed 1 km
2
.
Furthermore, full sized aircraft can lift heavy loads thus
allowing for a variety of image capture devices to be
used including optical, multispectral RADAR and LiDAR
devices. However, their significant disadvantage is cost
and logistics, especially in cases where repeat imagery is
required. Whilst the cost of airborne surveys is dropping
significantly (Carbonneau et al., 2011), repeat imagery
remains costly and logistically complex. The evidence of
this is that, to our knowledge, there is no published work
in the peer-reviewed literature which relies on repeated
hyperspatial imagery acquired at catchment-scales. Fur-
thermore, since scientists and managers very rarely have
access to aircraft, this platform is rarely scrambled within
hours or days of a trigger event such as a flood. The
examples in Table 8.1 illustrate the typical performances
of this traditional platform. Typically, full sized aircraft
can deliver hyperspatial imagery when flying below 500 m.
Even at such low altitudes, it is quite feasible to sample
long river reaches or several 10s of kilometres. Further-
more, unpublished data now exists which covers several
hundreds of kilometres at hyperspatial resolutions (N.E.
Bergeron 2012,
personal communication
).
Search WWH ::
Custom Search