Image Utilisation for the Study and Management of Riparian Vegetation: Overview and Applications - Fluvial Remote Sensing for Science and Management - page 222

Figure 10.6 Example of a stratified approach for implementing sampling strategy. Map d and map e give the five types of vegetation
to sample, and were obtained from the overlay of the map a (age of surface), map b (vegetation types in 2000) and map c ('non
natural units'). For the signification of the five types, see the text.

a progressive shift in stand composition from younger
stands characterised by Salix spp .and Alnus glutinosa
to older stands dominated by Fraxinus excelsior, Salix
alba and Alnus incana and individualised by Acer spp .and
Juglans regia (Figure 10.7). They also provide information
about which kind of vegetation should be expected at a
given age. The stands that colonise formerly cultivated or
urbanised surfaces are close to the stands that are between
20 and 43 years old.

and tools are now easily accessible outside of the scientific
community. However, such an approach is not error-free.
Indeed, it is based on the hypothesis that two units that
look the same on two different photos at different dates
are the same in the field. This is a fundamental limit of the
physiognomic approach; behind the same physiognomy
we can have very different communities and environ-
mental conditions. Moreover, some aspects cannot be
described or quantified solely using remotely sensed data.
A field survey is complementary and necessary, for species
composition analysis, for instance.

10.5.1.3 Conclusions and limitations
To conclude, the approach developed along the Arve
River is a very easy way to implement a historical analy-
sis, a stratified strategy for the description of vegetation
and a network of plots for a monitoring program (sur-
veyed every 10 years). The feasibility at the reach scale
for managers is good, because required data, software

10.5.2 Flowresistanceandvegetationroughness
parametrisation: LiDARandmultispectral
imagery

The objective of this case study is to produce a spatially
explicit map of vegetation structure with relevant input