[ESS
ttingham,
n.ac.uk
gineers use
lie studies.
nal method
pportunity
mining the
of LiDAR
rch was to
tion of the
cated from
\utomated’
coefficient
efficient of
ighness or
roughness
creating a
portant step
in reliable
It task often
number of
nining the
yendent on
ters, others
namic scale
Manning's
information
ure (ground
ison, 2001).
efficient of
take a long
niques can
ie. This can
s well as for
otential tool
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B3. Istanbul 2004
for assisting with determining Manning’s ‘n’ values. Airborne
remote sensing techniques normally provide better accuracy in
elevation measurements and more detailed analysis of the
surface compared to satellite remote sensing methods. There
are a number of airborne remote sensing techniques that can be
considered including aerial photography and photogrammetry,
LiDAR (airborne laser scanning), Synthetic Aperture Radar
(SAR) and multi-spectral line scanning.
Digital surface models (DSM's) are standard products from
aerial — photography using analytical and digital
photogrammetric techniques. Analytical photogrammetry is a
well-established mapping technique and can provide reliable
and accurate measurements (Elfick et al. 1994). However, the
measuring technique is a manual process which requires skills
and experience of putting the measuring mark on the stereo
model surface with every new measurement. This is a very
slow operation so obtaining a DSM is usually a time consuming
and very costly process. Digital photogrammetry employs
image-matching techniques to compute elevation measurements
from stereo pairs of digital (softcopy) aerial photographs. Then
creating DSM's can be automatically performed which makes
the process much faster and very cost effective compared to the
analytical technique. ^ Unfortunately, the quality of these
automatically-generated DSMs may not be as high when
compared with a DSM from analytical techniques. LiDAR is a
relatively new technology that can provide accurate DSM's,
with a suggested accuracy of between +10 cm to £20 cm, in a
relatively short time (Baltsavias, 1999).
The accuracy of a DSM or a digital terrain model (DTM) can be
critical in flood risk management in the cases of flat or gentle
sloping floodplains. Fowler, (2000) suggests that contour line
maps of the floodplain at one-foot (—30 cm) interval should be
available for a flooding study.
This paper presents the results from research in to the use of
photogrammetry and LiDAR techniques to provide high quality
DSM's and estimations of Manning's 'n' values. The paper
outlines the aims, an introduction to hydrodynamic studies, and
the results from a test site. Further information can be obtained
from Asal (2003).
1.1 Aims, Objectives and Methodology
Aims:
I. Investigate the use of airborne remote sensing techniques
for creating DSM's.
2. Assess the potential of using airborne remote sensing
techniques including laser scanning systems and aerial
photography in modelling the landscape in particular for
the estimation of Manning's coefficient of roughness.
Objectives:
l. Create and evaluate DSM's from airborne laser scanning
and aerial photography.
Apply comparative analysis of the surface models
generated from aerial photography and LiDAR to assess
the potential of each technique in analysing the landscape.
3. Investigate automatic method(s) for the estimation of the
coefficient of roughness in large areas such as floodplains.
N
Methodology:
l. To undertake the practical trials a test site has been
established at Newark-on-Trent to the east of Nottingham,
UK. LiDAR data was obtained from the Environment
715
Agency of England and Wales and the aerial photographs
were obtained from the National Remote Sensing Centre
(NRSC) at two common scales, 1:10,000 and 1:25,000.
These photo scales were chosen as they are commonly
flown scales in the UK and are readily available from
archives. This ensures this research has the widest
potential use.
The ground control points for the aerial photography and
field ground truth elevations were measured using Global
Positioning System (GPS) techniques.
3. Digital photogrammetry was undertaken using ERDAS
IMAGINE OrthoMAX.
4. Analytical photogrammetry was undertaken using a Leica
SD2000 analytical plotter.
5. Visual and quantitative analysis of the surface models, and
the investigations into automatic techniques for estimating
the coefficient of roughness were investigated using
ERDAS IMAGINE digital image processing system and
ArcView GIS.
n2
2. HYDRODYNAMIC STUDIES
2.1 Manning’s Coefficient of Roughness
Many equations have been developed for the purpose of
studying open channel flow, Manning’s equation being one of
the most widely used in this analysis. It is a semi-empirical
equation and was developed in the 19" century by Manning in
order to simulate open channel water flow. It was first designed
for the purposes of studying uniform steady state flows of
constant discharge, constant velocity and constant channel
dimensions with time. However, practical experience proved
that this equation can be successfully applied on gradually
varied flow, which is the common natural flow. It *is also used?
in defining the water flow over floodplains (LMNO, 2000).
Manning's equation takes the following form (Jain, 2001):
vz(1/n) (R^ s?) (1)
where:
v = the mean velocity through the channel in metres per second.
n ^ Manning's coefficient of roughness.
S 7 the channel bed slope in metres per metre.
R = the channel hydraulic radius calculated from:
R=A/WP (2)
where:
A =the cross sectional area of the channel.
WP = the wetted-perimeter of the channel.
2.2 Factors that Determine Manning's ‘n°
Kay (1998) states that depends on the building material of the
channel and the channel vegetation texture, which impose
difficulty in estimating it with any degree of accuracy.
Furthermore, the value of » is not constant with time in the
same channel due to weed growth and variations of flow
conditions over time. This can be explained as in case of small
flow rate grass and weeds tend to be upright which brings about
bigger resistance to the flow and leads to a bigger value of n.
The situation is different with high discharges in the same
channel due to grass and weeds being unable to continue
standing in high velocity. This leads to their flattening which
results in smaller resistance to the flow and smaller n value.
From this it can be seen that z is a variable quantity wii