Geography Reference
In-Depth Information
(a)
(c)
(b)
UAV IMAGE
June 23rd 2008
Ground resolution: 0.13 m
AERIAL IGN PHOTOGRAPHY
June 23rd 2006
Ground resolution: 0.50 m
QUICKBIRD IMAGE (PAN-SHARPENED)
September 21st 2008
Ground resolution: 0.60 m
(d)
(e)
(f)
QUICKBIRD IMAGE (MULTISPECTRAL)
September 21st 2008
Ground resolution: 2.40 m
LANDSAT ETM IMAGE (PAN-SHARPENED)
June 22nd 2002
Ground resolution: 14.25 m
LANDSAT ETM IMAGE (MULTISPECTRAL)
June 22nd 2002
Ground resolution: 28.50 m
Figure 8.1
Effect of spatial resolution on feature identification.
rapidly as spatial resolution increases from 2.4m to 28.5m.
Figure 8.1 clearly demonstrates that as spatial resolutions
drop below 1m and approach the 10 cm range, individual
landscape features which are relevant to river sciences
become visible. As a further example, Figure 8.2 shows
another series of images with resolutions ranging from
1.5 cm to 16 cm. In these images, individual trees, aquatic
plants, boulders, cobbles, sedimentary plumes and other
morphological features all become clearly resolved to the
point where quantitative measurements on their size and
number can reliably be made directly from the imagery.
Clearly, such imagery with centimetric scale resolutions
has a strong potential as a data acquisition method in the
broad field of river sciences and management.
In other fields of remote sensing, the term 'hyperspa-
tial' is currently emerging as a descriptor of such imagery
capable of resolving small, individual, spatial landscape
features. For example, Greenberg et al. (2009), working
in the area of forestry remote sensing, define hyperspatial
imagery as having pixels smaller than single trees. Further-
more, Rango et al. (2009) give a quantitative definition
of hyperspatial imagery as imagery with sub-decimetric
spatial resolutions. While the selection of fixed bound-
aries applicable to the term 'hyperspatial resolutions' is
somewhat subjective, we argue that the sub-decimetric
threshold identified by Rango et al. (2009) is sensible
in a fluvial context. Therefore we propose that the FRS
community adopt the term 'hyperspatial imagery' when
referring to sub-decimetric imagery. This new generation
of image products offers a powerful new perspective to
river scientists and managers. In particular the ability to
sample small features such as illustrated in Figure 8.2 has
the potential to change the way traditional reach scale
field studies are conducted. Existing knowledge and the
published literature in the fields of fluvial geomorphology
and lotic ecology are dominated by small to medium scale
studies with the main data acquisition methods based on
labour intensive fieldwork (Kondolf and Piegay, 2003).
Search WWH ::
Custom Search