Environmental Engineering Reference
In-Depth Information
the processes and mechanics of sediment trans-
port and transfer in the fluvial system in ques-
tion. While this is understood by a small corps of
experienced sediment modellers, it does not al-
ways seem to be fully comprehended by the
wider community of flood risk modellers and
stakeholders.
Recognizing this, some of the major issues that
complicate 1-D modelling of sediment dynamics
are discussed here for the benefit of those new to,
or on the fringes of, sediment modelling and the
users of sediment model outputs in the context of
flood risk management. As iSIS is currently the
hydrodynamicmodel most widely used in the UK,
it and its sediment module provide a suitable
vehicle for considering these issues. However,
the points raised herein are equally applicable to
most of the available sediment transport programs
and modules.
iSIS Sediment and Issues inOne-Dimensional
Sediment Modelling
Background
One-dimensional (1-D), computational models of
river flow have been used since the 1960s (e.g.
SOGREAH 1963) and are firmly established as the
standard technique for routine analysis of river
hydraulics and hydrodynamics. In the UK, a sig-
nificant proportion of all main rivers have had at
least part of their length modelled using commer-
cial 1-D software packages such as iSIS, HEC-RAS
or Mike11, and the current Environment Agency
flood mapping budget of around £10m per year
contributes further to the stock of available mod-
els. This represents a significant investment in
survey, calibration and assembly of one-dimen-
sional models and a valuable resource not only of
information but also of people experienced in use
of these models.
The hydraulic parameters needed for a sedi-
ment transport calculation are common to those
used in the hydraulic model calculations and thus
the existing model packages are suited for adapta-
tion to incorporate a sediment transport analysis.
Commonly used European codes such as iSIS,
Mike 11 and SOBEK have modules to compute
sediment fluxes and bed level adjustment at cross-
sections, and the same is true of programs devel-
oped in the USA, including HEC-6, GSTARS,
Fluvial 12 and, most recently, HEC-RAS 4.0. It
can, thus, be argued that the one-dimensional
approach has utility because it builds on existing
information and models and so is a logical first
option when a fully quantitative analysis of sed-
iment dynamics is required.
There are, however, practical and theoretical
difficulties in the application of 1-D models that
are not addressed in the relevant user manuals.
These mainly relate to data input and the steps
involved in running the model. As highlighted by
Hayter (2002), experience shows that while it is
relatively easy to create a 1-D sediment model
using any of the available programs, accurate
interpretation of sediment modelling results re-
quires substantial knowledge and insight into
Development of iSIS Sediment
iSIS Sediment originated in 1994 as part of the
iSIS collaborative venture between Halcrow
and HR Wallingford. The earlier code developed
at HR Wallingford is described by Bettess and
White (1981). iSIS Sediment was initially devel-
oped for simulation of siltation in large irrigation
canals such as those in Pakistan, but with the
addition during the 1990s of routines for graded
sediments and improved accounting for layering
effects, its capabilities and applications widened
rapidly. In 2001, further extensionswere developed
and tested at Herriot Watt and Glasgow Univer-
sities (Schvidchenko et al. 2001), although the
research version has not, to date, been completely
incorporated into the standard code.
A licence for the iSIS hydrodynamic model is
required to run the model though there is current-
ly no additional charge for access to the sediment
module.
The model can be used in a number of ways,
including long-term simulations of bed evolution
or to explore the effectiveness of engineering in-
terventions such dredging or modification of the
channel to improve flood conveyance or deal with
a sediment-related problem.
