Geography Reference
In-Depth Information
Digital surface model
(a)
-
Digital elevation model
=
Canopy height model
(b)
(c)
Figure 7.5b Computation of the Canopy Height Model (CHM, a) as the difference between the Digital Surface Model (DSM, b) and
the Digital Elevation Model (DEM, c). Courtesy of Airborne Hydrography AB.
Similarly, the canopy heights or positions of isolated
trees can be extracted from LiDAR data and used in
hydraulic models (Straatsma and Baptist, 2008).
has only been applied in coastal areas (Billard et al., 1986;
Churnside, 2008; Shamanaev, 2007). For example, Billard
et al. (1986) used the LADS system in a deep area near
the Australian coast. The signal used for analysis ranged
from approximately 1 m up to 33 m in depth to remove
the water surface and bottom peaks. By regression and
by using the Kalman filter, they estimated the B and k
parameters. Then, the turbidity profiling across the sur-
vey was considered homogeneous in the vertical column
with an estimated 15% relative error. However, in all
of these studies, estimations are only possible for water
depths greater than about five meters, which make them
generally unsuitable for riverine environments.
In the case of inelastic LiDAR, the relation between
water-column characteristics to Raman waveforms
using the ranging SHOALS system has been successfully
explored by Peeri and Philpot (2007). We can also
mention in passing the use of more specific fluorosensors
on oceanic waters for phytoplankton pigment mapping
(Hoge et al., 1983; Babichenko et al., 1993).
7.5.2 Prospectiveestimations
7.5.2.1 Water column properties
As a photon of light is transmitted through water, it
may interact with water molecules, dissolved substances,
or suspended particles. As a result, the photon may be
absorbed, backscattered (elastic LiDAR), or re-emitted as
a photon at a slightly altered wavelength (inelastic LiDAR)
(Peeri and Philpot 2007).
In the case of elastic LiDAR, as explained in Sections
7.2.2 and 7.4.1, waveforms of green laser pulses pene-
trating the water surface show decreasing power between
the water surface and bottom peaks due to the expo-
nential decay of transmission within the water column.
This decreasing waveform power (P) due to water vol-
ume backscattering can be simplified as the following
equation:
7.5.2.2 River bottom features
B exp 2kt
P ( d )
=
(7.4)
The interaction of LiDAR signals with river bottoms
has the future potential to provide information on the
nature of the substratum and the presence of immersed
objects or vegetation. To date, the authors are not aware
of any published studies (Irish and White, 1998) con-
cerning river bottom feature extraction from bathymetric
LiDAR data. The most recent developments of this tech-
nique have primarily concerned coastal areas (Wang and
Philpot, 2007, Collin et al., 2011a, Collin et al., 2011b), but
these developments foreshadow what could be possible in
riverine environments. First, the enhancement and inter-
pretation of a LiDAR-derived DEM allows the extraction
where t denotes time, and it can be transformed in depth
(d) by:
=
d
c 2 t
(7.5)
where k denotes the attenuation coefficient (approxima-
tion of the absorption coefficient), c 2 denotes the light
celerity in water and B is proportional to the volume
backscatter coefficient (Billard et al., 1986, p.2081). As a
consequence, the slope of this decreasing part of the wave-
form is theoretically a measure of k , and a relationship
to turbidity can be established. Recently, this approach
Search WWH ::




Custom Search