Geography Reference
In-Depth Information
to preserve the first-return points that were needed to
map the emergent sandbars in the reach. To ensure that
the spatial densities of the points were equivalent for
modelling, we interpolated the EAARL data to a surface
with kriging and selected points from that surface that
were coincident to the ground survey.
The FaSTMECH model was run with the EAARL
dataset using the same boundary conditions and drag
coefficient used for the topographic dataset collected on
the ground. The predicted water-surface elevation for this
simulation is shown in Figure 7.7b-B. While the simula-
tion also predicts emergent sandbars in the reach, the root
mean square error comparing the predicted water-surface
elevation to the water-surface elevations measured on the
ground was higher than from the ground survey, 0.097 m.
The predicted water-surface elevations were also higher
than the measured elevations by an average of 0.09 m. A
possible reason for this is that the EAARL system mea-
sured the deeper portions of the channel higher than
their actual elevation (Kinzel et al., 2007). This was con-
firmed by a comparison of the ground survey and the
coincident LiDAR which points indicated a mean error
of -0.1 m. The higher elevation LiDAR points had the
effect of reducing the conveyance through the reach and
predicting the water-surface elevation higher than what
was observed.
lines have been used to increase point density and spatial
resolution.
Techniques that improve ranging accuracy to shallow-
water riverine targets are also needed. Extracting infor-
mation from bathymetric waveforms may be improved
by combining theoretical models of waveform shape
with advanced pattern-recognition software routines.
Increased incorporation of passive optical sensors with
LiDAR systems is also beneficial. Synergistic integrated
processing of hyperspectral imagery and LiDAR has been
identified as a growing trend (Crane et al., 2004). This
involves using one data set to facilitate the processing
of another. These possibilities are especially tantalising
with respect to riverine surveys. If used with bathymetric
waveforms, hyperspectral data may be a valuable com-
plementary data set with the potential of refining shallow
depth estimation. Additionally, hyperspectral data could
be leveraged to provide information about the edges and
composition of riverine features. These emerging datasets
will provide managers a continuous representation of the
river channel and floodplain which can inform decisions
on a variety of issues including: connectivity of in-channel
and riparian habitat, flood risk, land use, fate and trans-
port of pollutants, and navigational concerns. These data
would also provide input to models and simulations
which could be used by stakeholders to evaluate or illus-
trate the consequences of proposed management actions.
The future development of remote sensing technologies
and LiDAR for mapping and monitoring river systems
will continue to be stimulated by competing demands for
water and with societal concern for the inventory and
preservation of aquatic resources.
However, the accuracy and spatial resolution of satellite
LiDAR data on continental waters is far better than radar
data (Braun et al., 2004; Baghdadi et al., 2011) and
considering the progress and increasing use of spatial laser
technology, we can speculate that future satellite LiDAR
missions would probably help in accurate monitoring of
river system dynamics.
Any use of trade, firm, or product names is for descriptive
purposes only and does not imply endorsement by the U.S.
Government
.
7.7 Conclusion and perspectives: the
future for airborne LiDAR on rivers
In the future, hardware and software developments will
enhance the resolution, accuracy, and types of data
products derived from riverine airborne LiDAR surveys.
First, the use of polarised LiDAR will probably enhance
the capacities of extremely shallow water bathymetry
(Mitchell et al., 2010). With regard to spatial coverage
and resolution, the pulse repetition frequency of current
bathymetric LiDARs is much smaller than airborne topo-
graphic mapping LiDARs. Using the EAARL sensor as
an example, an increase in the pulse repetition frequency
from 5,000 Hz to 30,000 Hz is planned (C. Wayne Wright,
USGS personal communication, 2010). The laser power
will also be increased from 70 micro joules to 700 micro
joules. To remain eye safe, the laser energy will be divided
over three laser spots, each of the spots will remain at
the current (
References
20 cm) size. The effect will be to triple
the number of laser samples in each raster or swath. This
greater rate of sampling and laser power per sample could
be advantageous in fluvial settings where multiple flight
∼
Allouis, T., Bailly, J., Pastol, Y., and LeRoux, C. (2010). 'Compar-
ison of LiDAR waveform processing methods for very shallow
water bathymetry using Raman, near-infrared and green sign',
Earth Surface Processes and Landforms
35
(6), 640-650.
Search WWH ::
Custom Search