Geography Reference
In-Depth Information
10.2.2 Mappingspeciesand individuals
et al., 2010). The majority of studies using diachronic
analysis are carried out on land cover or vegetation types.
Additional maps can be used in the reconstruction of
the changes of vegetation if we look back to the 1930s
and earlier (Muller et al., 2002; Gurnell et al., 2003). For
instance in France, the Napoleon land registers, dating
from the first half of the nineteenth century, are a good
source of information because all parcels are drawn and
detailed by land uses (Piegay and Salvador, 1997). Old
documents can have a very fine scale, such as the 1/2372
land register drawn between 1728 and 1738 in the King-
dom of Sardinia and used by Girel and Manneville (1998).
Old maps are usually easier to find for large rivers, which
were the first to have flow control structures and canals
constructed (see for example, in California, the California
Debris Commission maps dating from the end of nine-
teenth century). Thus, for a 10 km reach of the Danube
River in Austria, Hohensinner et al. (2004) found more
than 100 maps drawn between 1714 and 1991, 41 of which
were useful for their study (i.e., with scales ranging from
1/6900 to 1/25 000). The main limitations of old maps
concern the characterisation of information, as land cover
typology that can be vague and unevenly represented. For
example, there are frequently few details for vegetation
units in undeveloped areas, whereas there are more details
in cultivated areas. Furthermore, older maps have greater
issues with the accuracy of the image, and greater difficulty
in rectification for comparison with other images.
Until the 1990s infrared colour photographs were the
main medium used to identify some species, such as non
native species (Neale, 1997; Lonard et al., 2000; Everitt
et al., 2007), and the diachronic analysis of the vegetation
type was used as a surrogate to monitor species dynamic
(Peinetti et al., 2002). Indeed the colour of the foliage of
several species (e.g.
Salix alba
) makes it easy to delineate
units dominated by such species. Improvement in spatial
and spectral resolution now allows us to target the species
with improved accuracy (Table 10.2 and Figure 10.1d).
This has mainly been used to monitor the installation
and distribution of exotic or invasive species (Rowlinson
et al., 1999; Dipietro et al., 2002; Underwood et al.,
2006; Hamada et al., 2007). Development of very high
resolution images provides data that are well adapted
to this scale. For example, Dunford et al. (2009), used
colour photographs along the Dr ome River (France)
provided by unmanned aerial vehicles (UAV) with a
spatial resolution finer than 25 cm. They were thus able
to distinguish between species (mainly
Salix
sp.,
Populus
nigra
and
Pinus nigra
) and also able to separate deciduous
and coniferous dead crowns. Hyperspectral images can
also achieve such detail (Pengra et al., 2007).
10.2.3 Mappingchangesandhistorical
trajectories
Because the presence of the observer is not necessary
in order to access information provided by images, they
have been extensively used to reconstruct landscape
trajectories and thus define past conditions of the river. In
the European context a new emphasis on historical stud-
ies is being promoted by the European Water Framework
Directive in order to identify reference conditions for
water body functional assessment. The nature of imagery
employed by researchers relates directly to the latest tech-
nology available at the start of their study periods and the
timeframe studied: satellite images since the 1970s, aerial
photographs since the 1930s, oblique pictures since the
nineteenth century (Charlton, 2000; Grams and Schmidt,
2002), or old maps mainly since the seventeenth or eigh-
teenth century. Aerial photographs have been widely used
to map landscape configuration changes, turn-over rates
for geomorphic features, and the recovery process after
flooding (Miller et al., 1995; Marston et al., 1995; Men-
donca et al., 2001; Sloan et al., 2001; Freidman and Lee,
2002; Peinetti et al., 2002; Greco and Plant, 2003; Ferreira
et al., 2005; Geerling et al., 2006; Petit, 2006; Gonzalez
10.2.4 Mappingotherfloodplaincharacteristics
Besides mapping vegetation, remote sensing is also used
to study abiotic parameters relevant to riparian vegetation
studies. For instance, morphological features within the
floodplain can be identified and delineated from LiDAR
data (Jones et al., 2007). Maximum stand age can be
inferred from the age of the landforms obtained by a
historical analysis of channel mobility in the corridor
(Greco et al., 2007; Stella et al., 2012). This superposition
of historical channel boundaries also allows researchers to
estimate erosion rates and life spans of different landforms
(Van der Nat et al., 2003; Latterel et al., 2006; Nakamura
et al., 2007).
Inundation extent and some substrate characteristics
can also be obtained from images. Muller and James
(1994) reconstructed substrate characteristics along a
reach of the Garonne River (France) from multidate
Landsat Thematic Mapper (TM) images, even if they
could not infer fluvial landforms from this information.
Radar images can be used to very accurately map flood
Search WWH ::
Custom Search