Geography Reference
In-Depth Information
with an additional band in the infrared which is generally
intended for vegetation. However, the recently launched
WorldView-2 satellite proposes a marked improvement
in spectral terms with eight bands with widths of 40 to
70 nm in the visible range with two bands in the near-
infrared. This recently available imagery has not yet been
applied to small rivers and holds much potential.
For users interested in studying or managing very small
rivers with metric scale widths, even the best currently
available satellite image may still be insufficient. In such
cases, airborne remote sensing should be considered.
The final two entries in Table 1.2 are meant to give a
broad, preliminary, indication of the potential of airborne
remote sensing (see Chapters 2, 5, 7, 8, 9 and 11 for further
discussions). Airborne remote sensing is obviously a very
wide topical area. Here we present only two broad types of
acquisitionplatforms: air photography fromconventional
aircraft and Unmanned Aerial systems. Traditional air
photography is nowwidely available fromboth the private
sector and government agencies. In addition to colour
imagery, traditional aircraft can be used to mount a range
of instruments whichhave been shown to be useful in river
sciences. For example, Fausch et al. (2002) present high
resolution temperature acquired froma fixedwing aircraft
andMarcus et al. (2003) show how hyperspectral data can
provide a rich database of information which significantly
surpasses the limits of standard RGB imagery. In terms
of spatial resolution, aerial photography generally fills the
niche below satellite imagery. The temporal resolution of
air photos is obviously not as rigid as that of a satellite
which is bound in an elliptical orbit around the earth. In
theory, an aircraft can be mobilised very frequently and
visit a site at least once a day. However, potential users
should be aware that inpractice, this is very rarely possible.
Government agencies only very rarely commission repeat
flights of an area at intervals smaller than one year.
Similarly, private sector companies can sometimes have
the availability for repeat flights within a year although our
experience has been that this is very difficult for a specific
rivers owing to cost and logistic constraints. Unmanned
Aerial Systems (UAS) can free users from these logistic
constraints by giving the opportunity for managers and
scientists to operate their own aircraft. UAS exist in a very
wide range of sizes and purposes. In fact some UAS, for
example the Global Hawk and Ikhana systems operated
by NASA, are in essence full sized, pilotless, aircraft.
However, of particular interest here is the ever growing
range of small, toy-sized, UAS available on the civilian
commercial market. These systems are easy to pilot and
come equipped with small format digital cameras and
onboard navigation hardware which often allows for fully
automated flight and data acquisition. These small aircraft
can fly at very low altitudes and therefore can deliver very
high resolution imagery. Their small sizemakes themvery
easy to deploy at high temporal resolutions. At the time
of writing, publications using UAS data are relatively rare
in river sciences (but see Dunford et al., 2011). However,
this new technology is prompting much excitement in the
river sciences community and the publication record can
be expected to grow in the coming years.
1.2.3 Cost considerations
Most users considering remotely sensed datawill probably
turn to free data sources in the first instance. Classic
Landsat data is freely downloadable from the United
States Geological Service (USGS) via their EarthExplorer
website (earthexplorer.usgs.gov). Whilst the resolution
is low, this data can still provide some initial insights
for medium to large rivers. For smaller rivers, most
users will likely turn to free online mapping services like
Google Earth which displays very good quality imagery,
often with sub-metric resolutions. Google corporation
purchases this imagery from a range of airborne and
satellite sources (some in Table 1.2) and makes them
freely viewable online. However, users cannot download
full, raw, image products from Google Earth. Therefore,
in the majority of cases, the purchase of data will still
be required. The costs of such purchases are obviously
a crucial consideration. Whilst these are quite variable
across the full range of data types, sensors and platforms,
we give here a basic summary which is not specific to
any single company or service provider and which will
hopefully provide the reader with some initial estimates.
In the case of satellite imagery, there are two impor-
tant, broad, distinctions. First, is a new image required?
Satellite image providers maintain full archives of all
previously acquired images. These archived images are
sold at discounted costs which range from 10-20 US$
per km
2
. However, if a new image is required, the pur-
chase of a new acquisition will increase the cost to at
least 20-80 US$/km
2
. The second factor in satellite image
cost is the level of pre-processing. The cost estimates
above are for basic standard imagery. However, image
providers offer pre-processing services which range from
improved image quality in terms of position, geometry
and radiometry to the full production of Digital Terrain
Models (DTMs). These levels of processing will obviously
increase the cost, sometimes in excess of 100 US$/km
2
.
Readers should also note that a minimum area must
Search WWH ::
Custom Search