Geography Reference
In-Depth Information
8
Hyperspatial Imagery
in Riverine Environments
Patrice E. Carbonneau
1
, Herv ´ePiegay
2
,Jer ome Lejot
3
,
Robert Dunford
4
and Kristell Michel
2
1
Department of Geography, Durham University, Science site, Durham, UK
2
University of Lyon, CNRS, France
3
University of Lyon, CNRS, France
4
Environmental Change Institute, Oxford University Centre for the Environment,
Oxford, UK
clasts and individual trees. Larger in-channel features
such as bedforms, even if smaller than 1 meter, can now
be densely sampled. For example, Figure 8.1 shows a series
of images of a river reach with large exposed gravel bars.
These images were acquired from airborne and, for com-
parison, satellite platforms. Figure 8.1a was acquired from
an Unmanned Aerial Vehicle and has a spatial resolution
of 13 cm. In this image and in the zoomed subset, we
can clearly see small morphological units and individual
shrubs. Figure 8.1b was acquired from a fixed wing, tradi-
tional aircraft (seen in Figure 1.1c of Chapter 1 in this vol-
ume). It has a spatial resolution of 50 cm. Due to the lower
resolution, many small scale features are lost but we can
still appraise the morphological change that has occurred
in the two-year timespan between photos 8.1a and 8.1b.
Figure 8.1c shows a Quickbird, pansharpened, image with
a simulated resolution of 60 cm. Even at such resolutions,
and despite the fact that the red, green and blue bands
in Quickbird do not exactly correspond to those in RGB
photography, we can identify features which are similar
to those in Figure 8.1a which was taken four month
earlier. In contrast, Figures 8.1d to 8.1f all show images
with resolutions in excess of 1m. Clearly in these cases
we can see that the level of feature representation drops
8.1 Introduction: The Hyperspatial
Perspective
The bibliometric analysis in Chapter 1 of the present
volume quite clearly shows that to date, the bulk of pub-
lications in the area of fluvial remote sensing (FRS) are
very strongly grounded in classic remote sensing meth-
ods, approaches and perspectives. We find a large volume
of papers examining large scale features such as river
estuaries (e.g., Hedger et al., 2007; Chen et al., 2009) and
vegetation mapping (in the riparian zone) is also very
common (e.g., Akasheh et al., 2008; Bertoldi et al., 2011).
Published work which uses FRS to examine smaller scale
parameters, such as woody debris and hydraulic variables,
which are usually of interest to fluvial geomorphologists
and lotic ecologists are much less common and generally
quite recent. Marcus and Fonstad (2008) and Marcus
and Fonstad (2010) review and discuss the progress and
fundamental changes occurring in FRS. One of the crucial
elements of this progress is the ability to collect image
data at resolutions often below 10 cm, and as low as
1.5 cm, from airborne platforms. This imagery has the
power to resolve landscape features such as individual
Search WWH ::
Custom Search