Future Prospects and Challenges for River Scientists and Managers - Fluvial Remote Sensing for Science and Management - page 418

19

Future Prospects and
Challenges for River
Scientists and Managers

Patrice E. Carbonneau 1 and Herv ´ePiegay 2
1 Department of Geography, Durham University, Science site, Durham, UK
2 University of Lyon, CNRS, France

We hope to have demonstrated in this edited volume
that fluvial remote sensing is now capable of delivering
unprecedented data to river scientists and managers.
Starting from basic principles, the volume has illus-
trated how fluvial remote sensing has evolved into a
self-contained discipline. We have discussed hyperspec-
tral imagery, thermal imagery and visible imagery (from
the ground, the air or space) all of which offer new
ways of imaging rivers capable of resolving fine spatial
and spectral details. We have also discussed new LiDAR
approaches, both terrestrial and airborne, which offer
topographic data of unprecedented quality. Finally, we
hope to have shown the value of this data with some
emerging applications which covered the biotic, abiotic
and even social aspects of river sciences. Where is the new
emerging discipline of fluvial remote sensing heading? If
the pace of technical progress is maintained, it would seem
that there are few fundamental limitations impeding the
improvement of spatial and spectral resolutions. Further-
more, the temporal resolution of fluvial remote sensing
datasets, which is currently impeded mostly by cost and
logistic issues, is expected to improve markedly in the
next decade owing to reductions in costs of classic air-
borne platforms and to progress in the areas of terrestrial
fluvial remote sensing and UAV technology. As a result,
our ability to produce vast hyperspatial (see Chapter 8
for a definition) and hyperspectral (see Chapter 4 for a

definition) datasets can only improve. When compared to
the datasets that led to major and significant river sciences
contributions such as Hydraulic Geometry (Leopold and
T., 1953) or the network dynamics hypothesis (Benda
et al., 2004), fluvial remote sensing datasets are bigger by
orders of magnitude. But does this over-abundant nir-
vana of data necessarily lead to improved science, new
knowledge and better management? Our answer here is a
cautious maybe .
This move from a data-sparse to a data-rich situa-
tion is a fundamental change which has not been fully
appreciated by the river sciences community. A key dis-
tinction must be made between data and knowledge. A
vast data set might contain vast amounts of information
but extracting this information and transforming it into
knowledge via the scientific method is a more difficult
task when compared to smaller, sparse datasets. Unfortu-
nately, somewhat less attention has been paid to analysis
methods and conceptual frameworks underpinning these
hyper-large datasets. As authors who have worked exten-
sively with large fluvial remote sensing datasets (e.g.
Carbonneau et al., 2004; Alber et Piegay, 2011; Carbon-
neau et al., 2011; Wawrzyniak et al., 2011), we, the editors
of this volume, are well aware of the difficulties posed by
their analysis. Further progress and reflection focussed on
the meaningful analysis of these datasets is now a crucial
challenge for the future. For example, Carbonneau et al.