Professional Documents
Culture Documents
Shallow Water Sediments Equations For Fluvial Processes
Shallow Water Sediments Equations For Fluvial Processes
EH14 4AS, U.K. (corresponding author). E-mail: zxcao@whu.edu.cn teractive processes of flow, sediment transport, and morphological
evolution in alluvial rivers constitute a hierarchy of physical prob-
Chunchen Xia lems of significant interest in the fields of fluvial hydraulics and
geomorphology. Enhanced understanding of these processes is es-
Ph.D. Candidate, State Key Laboratory of Water Resources and Hydro-
power Engineering Science, Wuhan Univ., Wuhan 430072, China. E-mail:
sential for not only river engineering practice, but also effective
xcc@whu.edu.cn flood risk management and environmental and ecological well-
being (Gomez 1991; Wu 2007; Frey and Church 2009; ASCE/
EWRI Task Committee 2011).
Gareth Pender
Mathematical modeling has become one of the most proactive
Professor, Institute for Infrastructure and Environment, Heriot-Watt Univ., approaches to enhancing the understanding of fluvial processes
Edinburgh EH14 4AS, U.K. E-mail: g.pender@hw.ac.uk over the last half-century or so (Wu 2007), irrespective of miscon-
ceptions of its capability and uncertainties from a variety of sources
Qingquan Liu (Oreskes et al. 1994). Papanicolaou et al. (2008) present a review of
Professor, Dept. of Mechanics, School of Aerospace Engineering, Beijing current models for fluvial processes, focusing on their main appli-
Institute of Technology, Beijing 100081, China. E-mail: liuqq@bit.edu.cn cations, strengths, and limitations. They also provides insights
into future trends and needs with respect to hydrodynamic and sedi-
Forum papers are thought-provoking opinion pieces or essays ment transport models. While fully three-dimensional modeling
founded in fact, sometimes containing speculation, on a civil en- may facilitate very detailed resolution of the phenomena (Wu et al.
gineering topic of general interest and relevance to the readership 2000; Fang and Wang 2000; Marsooli and Wu 2015), the comput-
of the journal. The views expressed in this Forum article do not ing cost is excessively high in general and thus its application
necessarily reflect the views of ASCE or the Editorial Board of is limited to cases of small spatial scales and short durations.
the journal. Comparatively, SHSM models feature a sensible balance between
theoretical integrity and applicability and therefore have seen wide-
DOI: 10.1061/(ASCE)HY.1943-7900.0001281 spread applications.
Pivotal to shallow water hydrosediment–morphodynamic mod-
els are the governing equations (hereafter SHSM equations), which
Introduction are built on the fundamental mass and momentum conservation
laws of fluid dynamics (Xie 1990; Wu 2007). Unlike traditional
The last several decades have witnessed increasingly widespread shallow water hydrodynamic equations for clear water flows (Toro
applications of shallow water hydrosediment–morphodynamic 2001), the interactions between flow, sediment transport, and bed
(SHSM) models in hydraulic engineering and geomorphological evolution are explicitly accommodated in the complete SHSM
studies. Pivotal to such models are the SHSM equations that have equations. The last 10 years or so have seen an increasing number
been well established from the fundamental mass and momentum of computational studies of a hierarchy of extreme fluvial processes
conservation laws of fluid dynamics. However, there continue to be based on these SHSM equations, i.e., dam-break floods over erod-
various simplified SHSM equations based on the assumptions of ible bed (to name a few, Capart and Young 1998; Cao et al. 2004;
very slow bed evolution, sediment transport capacity, and quasi- Wu and Wang 2007; Xia et al. 2010; ASCE/EWRI Task Committee
steady flow, of which the effects remain insufficiently understood. 2011; Huang et al. 2012, 2014, 2015). More broadly, these equa-
Also, in a recent reformulation, the established SHSM equations tions have been extended for modeling coastal processes (e.g., Zhu
are questioned because it is argued that bed sediment entrainment and Dodd 2015; Kim 2015; Xiao et al. 2010), watershed erosion
effects are incorrectly incorporated in previous models and jump processes (Kim et al. 2013), subaqueous sediment-laden flows and
conditions at the bed surface are needed to rectify the accounting turbidity currents (Hu and Cao 2009; Hu et al. 2012) as well as
for the effects of mass exchange. This paper presents a review of the sharply stratified processes (Li et al. 2013; Spinewine and Capart
traditional SHSM equations along with three simplified versions, 2013; Cao et al. 2015b; Zech et al. 2015), which actually involve
i.e., the decoupled version based on the assumption of very slow double layer-averaged SHSM equations.
bed evolution, the deduced version in line with the assumption of Nevertheless, there continue to be various simplified versions of
sediment transport capacity, and the truncated version due to the the complete SHSM equations used in hydraulic engineering and
assumption of quasi-steady flow. It also addresses the recent refor- geomorphological studies, of which the effects have so far re-
mulation that hinges upon presumed, mesoscopically and macro- mained insufficiently understood. One of the conventional simpli-
scopically controversial jumps of velocity and stress at the bed fied versions concerns the decoupled SHSM equations (referred to
surface. A critical issue is flagged out of the reformulation con- as the decoupled version), in which a few terms characterizing the
cerning an actual momentum exchange with the bed, which must feedback effects of bed deformation and sediment transport on the
vanish physically and otherwise may incur practically unrealistic water–sediment mixture flow are ignored. Another widely evoked
are incorrectly incorporated in previous models and jump condi- grades to a one-way phenomenon. The bed is literally the bottom
tions at the bed surface are needed to rectify the accounting for boundary of the flow and always affects the flow, while the flow
the effects of mass exchange. Their reformulation pertains to not does not directly alter the bed (but indirectly affects bed evolution
only high-speed mass flows (e.g., debris flows), but also fluvial via sediment transport), regardless of the forces that the flow exerts
flows such as dam-break floods over erodible bed. Succinctly, they on the bed surface. Of particular significance to the SHSM equa-
cast doubt on the SHSM equations that have been well established tions in this respect is the issue of physically coupled versus de-
and widely applied in hydraulic engineering practice and geomor- coupled modeling (Fig. 1), as discussed in a subsequent section.
phological studies for a few decades (Xie 1990; Wu 2007), while The bed surface merits elaboration as a key to this paper. It is
leaving a number of issues that are open for debate and invite essentially the interface between the water–sediment mixture flow
more investigations. and the static bed layer. It usually evolves as a result of nonequili-
This paper aims to present a review of the traditional SHSM brium sediment transport due to unbalanced mass exchange,
equations along with three simplified versions, i.e., the decoupled i.e., the net flux between the upward sediment entrainment from
version, the capacity version, and the quasi-steady version. The fo- the bed surface and the downward deposition from the water
cus is placed on their respective applicability and issues in need for column does not vanish. In general, the bed surface may appear
further investigations. It also flags a critical issue related to the indistinct because the sediment particles may move intermittently
recent reformulation by Iverson and Ouyang (2015) in order to and stochastically close to the bed, and also it can be very compli-
spur discussions about the theoretical basis of the SHSM equations. cated when bed forms such as dunes and ripples are present in
Finally, future development of SHSM equations is briefly ad- alluvial rivers. Nevertheless, the position of the mobile bed has
dressed under the two-phase flow framework. to be defined in mathematical models (not just depth-averaged
models). The authors are unaware of an explicit definition of
the bed surface in existing literature in the context of river dynam-
ics. Generally, the time-averaged velocity of the water–sediment
Flow–Sediment-Bed Interactions
mixture flow decreases downward from the free (water) surface
Generally, fluvial flow, sediment transport, and the bed interact along the flow depth [e.g., Figs. 8.4a and 8.15a in Bridge and
with each other. Fig. 1 is a sketch of the interactions that warrant Demicco (2008)]. Therefore, in theory the bed surface can be de-
elaboration in the context of computational river dynamics. Specifi- fined at the level where the time-averaged velocity of the water–
cally, the flow and sediment transport feature a two-way interac- sediment mixture flow just vanishes (Fig. 2). Immediately above
tion. Sediment transport is dictated by the flow, while sediment the bed surface, the flow may feature a substantial velocity because
may modify the flow (e.g., turbulence may be enhanced or sup- the flow usually increases sharply in the immediate vicinity of the
pressed by sediment). In connection with the former way of the bed surface. Naturally, there is no motion below the bed surface
flow-sediment interaction is the issue of noncapacity versus capac- because the seepage flow through the porous medium of the static
ity transport in SHSM models (Fig. 1), which is addressed in a bed layer can be presumed to be negligible.
subsequent section. The latter way of flow-sediment interaction
is not covered in this paper because the effects of sediment on
SHSM Equations
the flow (e.g., turbulence modification) have never been incorpo-
rated in SHSM models as far the authors are aware, despite there The governing equations of shallow water hydrosediment–
morphodynamic models can be derived by directly applying the
Reynolds transport theorem in fluid dynamics (Batchelor 1967),
which has been available for over two decades in the undergraduate
textbook edited by Xie (1990). The textbook was written in
Chinese, and therefore may not be accessible to many. Yet, one
can readily find the SHSM equations presented in English
(Wu 2007), derived by integrating and averaging the full three-
dimensional mass and momentum conservation equations. For
convenience, these are referred to as traditional SHSM equations,
as compared with the recently reformulated equations by Iverson
and Ouyang (2015).
traditional SHSM equations, and their momentum conservation- be able to resolve the fluvial processes properly. Of particular sig-
based Eq. (38) can be rewritten as nificance to this paper is the coupled modeling methodology. As
the bed deforms, it will lead to changes in the flow. This influence
∂ðρ̄1 h1 ū1 Þ ∂ 1 is primarily dependent on how fast the bed deforms. When bed
þ ðβ 1uu ρ̄1 h1 ū21 þ ρ̄1 gh21 Þ
∂t ∂x 2 deformation is significantly slower than that of flow changes by
∂zb orders of magnitude, the feedback effects of bed deformation on
¼ −ρ̄1 gh1 þ τ 1top − τ 1bot þ ρ̄1 u1 ðzb ÞE1bot ð8Þ the flow are negligible, and traditional decoupled mathematical
∂x
river models are approximately applicable, in which the RHS of
where β 1uu = Boussinesq momentum-distribution coefficient, Eq. (5) is neglected along with the third and fourth terms on the
which can be approximately presumed to be equal to unity; RHS of Eq. (6). Otherwise, decoupled models may totally collapse,
E1bot = bed material entrainment rate at the bottom of the flow and fully coupled models using the complete SHSM equations are
layer; τ 1top = top shear stress of the flow layer, usually negligible required in order to properly resolve the strong interactions be-
in open channels; and u1 ðzb Þ = velocity component of the flow tween the flow, sediment transport, and bed. One of the most telling
layer at the bed surface in the x-direction. cases concerns dam-break flow over erodible bed (Cao et al. 2004).
Most notably, Eq. (8) due to Iverson and Ouyang (2015) differs It is necessary to flag out that both physical coupling and
from Eq. (2) (Wu 2007; Cao et al. 2004; Capart and Young 1998; numerical coupling are important for modeling fluvial processes
Xie 1990) in that the former includes an extra term ρ̄1 u1 ðzb ÞE1bot in with active sediment transport and rapidly evolving bed. Physical
its RHS. Fundamentally, this stems from the way the bottom boun- coupling necessitates the use of the complete, rather than simpli-
dary conditions are characterized. In light of the continuum fied, governing equations for the phenomena, i.e., Eqs. (5), (6), (3),
assumption, the velocity components and stresses are continuous and (4). Most notably, the contribution of mass exchange with the
at the bed surface, which is in contrast to jumps evoked by Iverson bed to the mass conservation of the water–sediment mixture flow is
and Ouyang (2015). Specifically, the tangential velocity compo- critical, as characterized by the RHS term in Eq. (5). For extreme
nents vanish at the bed surface because the bed is static, which cases such as dam-break floods over erodible sediment bed, ne-
in the context of fluid dynamics is referred to as the no-slip con- glecting it may lead to collapse of the modeling at the worst
dition of viscous fluids, i.e., u1 ðzb Þ ¼ u2 ðzb Þ ¼ 0. This has been (Cao et al. 2004), while some mild cases may entail considerable
well embedded in models for fluvial processes (Wu 2007) and earth errors. Compared with the RHS term in Eq. (5), the gradient term of
surface flows [Figs. 8.4a and 8.15a in Bridge and Demicco (2008)]. sediment concentration and the apparent momentum exchange term
From a microscopic perspective, the nonslip approximation at a on the RHS of the momentum equation [e.g., Eq. (6)] are much
liquid-solid interface may be affected by a number of factors, as more minor and can be neglected practically. Numerical coupling
revealed by experimental studies (Neto et al. 2005) inspired by requires synchronous solution of the whole set of the governing
the developments in the fields of microfluidic and microelectrome- equations, an issue tentatively ignored in this paper.
chanical devices. Yet, the SHSM equations are certainly developed Arguably, prior to Capart and Young (1998), mathematical river
for applications to processes at the mesoscopic and macroscopic models based on the SHSM equations were rarely coupled as the
scales. More broadly, the continuity condition of the tangential RHS of Eq. (5) and the third and fourth terms on the RHS of Eq. (6)
velocity components at the bed surface holds for subaqueous were exclusively ignored, even though the complete SHSM equa-
granular flows (Armanini 2013; Armanini et al. 2009) and tions had been available since Xie (1990). Studies over the last de-
sediment-laden flows or turbidity currents (Huang et al. 2005; cade have revealed that coupled modeling does not incur any
Kneller and Buckee 2000). appreciable increase in computing cost compared with decoupled
Physically, the mass exchange between the water-sediment mix- modeling, as can be anticipated from the fact that the physical cou-
ture flow and the bed does not involve any momentum exchange pling just additionally necessitates the computation of a few source
because the sediment (and water) entrained from the static bed does or sink terms. At the same time, coupled modeling involves fewer
not have any momentum initially, and therefore does not bring any assumptions than decoupling modeling, thereby minimizing model
momentum into the flow. Likewise, the sediment (and water) de- uncertainty. Indeed, coupled modeling has been more and more
posited into the bed does not take out any momentum from the widely used in computational river dynamics since Capart and
flow. This proposition has been correctly reflected in traditional Young (1998) and Cao et al. (2004), even for large-scale field cases
SHSM equations, e.g., Xie (1990), Capart and Young (1998), such as the Tangjiashan landslide dam failure process and the
Cao et al. (2004), and Wu (2007), as illustrated by Eq. (2). Iverson resulting flood following the Wenchuan Earthquake on May 12,
and Ouyang (2015, p. 13) appreciate that there must not be real 2008, in Sichuan Province, China, and also the historic megaflood
momentum exchange between the flow and the static bed. How- due to glacier dam-break flood in Altai Mountains, Southern
ever, the last term on the RHS of Eq. (8) clearly indicates a real Siberia (Huang et al. 2012, 2014, 2015). Nevertheless, the advan-
momentum gain (when E1bot > 0) or loss (when E1bot < 0) from tages of coupled modeling have not been sufficiently recognized
mass exchange, if u1 ðzb Þ does not vanish. They try to justify this because decoupled modeling continues to be popularly used by
Papanicolaou et al. 2008). Pelosi and Parker (2014)] models are generally justified as opposed
to capacity models. Capacity models could be conditionally appli-
cable if sediment adaptation to capacity regime is fulfilled suffi-
Noncapacity versus Capacity Modeling ciently rapidly and within an adequately short distance.
For fluvial suspended sediment transport, noncapacity models
Sediment transport capacity is a pivotal concept in the context of
have been increasingly widely developed and applied during the
fluvial hydraulics and geomorphology. Succinctly, it is defined as
last several decades. Most likely this relates to the consensus that
the maximum amount of sediment that can be transported by the
it takes long temporal and spatial dimensions for suspended sedi-
flow without causing aggradation or degradation of the riverbed.
ment transport to adapt to capacity in line with local flow regime,
This corresponds to an idealized regime, characterizing equilibrium
simply (and somewhat intuitively) because suspended sediment
sediment transport by steady and uniform flow. Yet fluvial flows
may distribute along the full flow depth, as opposed to bed load
are generally unsteady and nonuniform, and under these circum-
that usually moves in the immediate vicinity of the bed. Indeed,
stances the concept of sediment transport capacity may not be
the analysis of the relative time scale of suspended sediment
strictly applicable. Accordingly, nominal sediment transport capac-
transport concurs with this recognition (Cao et al. 2007), and com-
ity is introduced, determined exclusively by local flow and bed
putational tests (Cao et al. 2012) reveal that a capacity model for
conditions (i.e., at a cross section). For ease of description, the word
suspended sediment transport may entail substantial deviations
nominal is typically dropped. Certainly, the nominal transport
from a noncapacity model and therefore is generally not justified.
capacity reconciles with the actual capacity in steady and uniform
In contrasting, bed-load transport has been widely presumed to
flows over a sufficiently erodible bed.
be at capacity. Likely this is because bed load usually moves in the
For over half a century [arguably since Exner’s analysis of why
immediate vicinity of the bed and therefore is believed to be able to
dunes are asymmetrical in the 1920s (Parker 2004)], fluvial sedi-
adapt to capacity rapidly (van Rijn 1993). This intuition is under-
ment transport has often been quantified based on the assumption
pinned by the theoretical analysis of the relative time scale of uni-
that its transport rate (or concentration) is always equal to capacity
determined exclusively by local flow and sediment conditions. form bed-load transport (Cao et al. 2011). Complementary to the
Based on this capacity assumption and Eqs. (5), (6), (3), and (4), analysis, computational tests (Cao et al. 2012) confirm that a capac-
one can readily yield the governing equations of capacity ity model for uniform bed load leads to practically equivalent
models by eliminating the net flux F of sediment exchange be- results as a noncapacity model. Nevertheless, bed-load sediments
tween the flow and the bed, i.e., the capacity version of the SHSM can differ in their sizes greatly, ranging from sands to gravels,
equations cobbles, and even boulders. The adaptation of nonuniform bed-load
transport to capacity is at least intuitively size-dependent. The
∂h1 ∂h1 ū1 ∂zb analysis of the relative time scale of nonuniform bed-load transport
þ þ ¼0 ð9Þ shows rapid, size-dependent adaptation to capacity regime (Hu
∂t ∂x ∂t
2013; Cao et al. 2016), extending the work for uniform bed-load
transport (Cao et al. 2011). Yet, this finding is essentially governed
∂ðh1 ū1 Þ ∂ 1
þ h1 ū21 þ gh21 by the complete SHSM equations built on the fundamental mass
∂t ∂x 2 and momentum conservation laws. However, the adaptation is con-
∂zb τ 1bot ðρs − ρf Þgh21 ∂ c̄1 ðρ̄2 − ρ̄1 Þū1 ∂zb strained by changes in the water and sediment inputs from either the
¼ −gh1 − − þ upstream or a tributary, in addition to the conservation laws. As a
∂x ρ̄1 2ρ̄1 ∂x ρ̄1 ∂t
result, the capacity assumption for bed load may not be generally
ð10Þ justified. The computational case study of a strong aggradation case
due to sediment feeding (Seal et al. 1997) reveals that the capacity
∂ ∂ðh1 ū1 c1e Þ model entails substantial errors compared to the noncapacity model
½ð1 − pÞzb þ h1 c1e þ ¼0 ð11Þ (Hu 2013; Cao et al. 2016).
∂t ∂x
As Gomez (1991) stated, “Many of the problems that have pre-
where c1e = sediment transport capacity of the flow layer. vented the development of a viable method of predicting bed load
Eqs. (9)–(11) constitute the governing equations for a capacity transport arise directly from the assumption that equilibrium trans-
model [or local model according to Pelosi and Parker (2014)], solv- port conditions prevail.” Here, equilibrium transport condition es-
ing for three unknowns (h1 , ū1 , and zb ). Eq. (11) reduces to the sentially means capacity regime, rather than the real equilibrium as
traditional Exner equation when dðh1 c1e Þ=dt is neglected, which shown in Fig. 1. In essence, Gomez (1991) flagged the significance
provides that the spatial gradient of the sediment transport rate of accounting for the spatial and temporal variability in generally
solely dictates the bed evolution. nonuniform bed-load sediment transport, which explicitly necessi-
From physical perspectives, the temporal and spatial variability tates noncapacity modeling. Charru (2006) demonstrated that
of sediment transport is essentially ignored in the capacity quantifying bed-load transport with a noncapacity model, instead
Fig. 4. (Color) Water surface and bed profiles at (a) t ¼ 20 s; Fig. 5. (Color) Velocity profiles at (a) t ¼ 20 s; (b) t ¼ 60 s;
(b) t ¼ 60 s; (c) t ¼ 150 s (c) t ¼ 150 s
While the pros and cons of two-phase SHSM models remain spells a positive (negative) feedback into the flow momentum when
poorly understood currently as compared with the traditional the bed degrades (aggrades), and inevitably leads to catastrophic
quasi-single-phase SHSM models, in terms of modeling accuracy failure of the model.
and computing cost, two-phase modeling holds great promise be- Succinctly, both the traditional SHSM equations and reformu-
cause more refined physical mechanisms of water-sediment flows lation by Iverson and Ouyang (2015) characterize quasi-single-
are explicitly incorporated. Most notably, fluid-solid two-phase phase models for the water-sediment mixture flow. This class of
flows are primarily characterized by the relative motion and inter- models are preferentially applicable to cases of low sediment con-
actions between the two phases. For fluvial processes, while centrations, in which the water flow reigns over the sediment mo-
suspended sediment may transport at nearly the same velocity as tions, but may not be refined enough to resolve the complicated
the flow, bed-load velocity is considerably lower than the flow physics of hyperconcentrated flows as the interphase and par-
(Einstein 1950; Chien and Wan 1999; Greimann et al. 2008). ticle-particle interactions prevail. The development and applica-
As observed in debris flows, the velocities of the two phases may tions of two-phase SHSM equations have so far remained in
deviate from each other substantially, therefore, affecting the flow their infancy, but hold great promise for modeling shallow water
mechanics (Pudasaini 2012). Accordingly, drag is one of the basic flow–sediment morphological processes, and therefore there is a
and critical mechanisms of two-phase flows, featuring the coupling huge scope for further investigations. In this respect, enhanced
between the two phases. While the relative motion between the understanding of the physics of the interphase and particle-particle
fluid and the solid phases can be resolved by the two-phase ap- interactions is crucial, in addition to the turbulent stresses of the
proach, the associated drag in between the two phases merits care- fluid phase and also the stresses due to solid particle fluctuations.
ful incorporation. Concurrently, at high sediment concentrations,
the particle-particle interactions play an important role in dictating
the flow mechanics. These interactions are normally represented Acknowledgments
by the solid phase stresses (i.e., intergranular stresses) in its mo- The work reported in this paper is funded by the Natural Science
mentum conservation equation. Likewise, the fluid phase stresses Foundation of China (Grant Nos. 51279144 and 11432015) and the
due to viscosity (i.e., viscous stresses) may also be considerable, in Chinese Academy of Sciences (Grant No. KZZD-EW-05-01-03).
addition to the turbulent stresses. The comments of anonymous reviewers and editors helped improve
Among the physical mechanisms mentioned previously, esti- the manuscript, which are very much appreciated.
mating the interphase drag, intergranular stresses, and viscous
stresses might be arguably relatively straightforward. The most
challenging part concerns the turbulent closure model for the fluid Notation
phase and the accounting for the fluctuations of the solid phase
because the current understanding of the physics of water-sediment The following symbols are used in this paper:
flows at high solid concentrations remains very limited, albeit it c̄1 = depth-averaged sediment concentration of the
progresses (Iverson 1997). This is perhaps why turbulent stresses flow layer;
and solid phase fluctuations have been rarely, if ever, incorporated c1e = sediment transport capacity of the flow layer;
in two-phase SHSM models to date, even for debris flows E1bot = bed material entrainment rate at the bottom of the
(e.g., Pitman and Le 2005; Pelanti et al. 2008; Pudasaini 2012). flow layer;
Given the sophisticated nature of the phenomena in question, F = net flux of sediment exchange between the flow
the turbulent closure models will in general necessitate additional column and the bed;
partial differential equation(s) to close the two-phase SHSM equa- fdsf = drag force of fluid on solid particles;
tions, i.e., Eqs. (7) and (3) along with Eqs. (14) and (15) or Eqs. (16)
fsf = forces exerted by the fluid on the solid particles;
and (17), because algebraic closure relations are presumably exces-
g = gravitational acceleration;
sively simplistic and thus may not be generally viable. A similar
observation should hold when accounting for the fluctuations of h1 = flow depth;
the solid particles [e.g., Eq. (35) for granular temperature in Iverson MT = momentum transfer term;
(1997)]. This certainly invites future investigations. MTA = added momentum term due to Iverson and
Ouyang (2015);
MTC = apparent momentum exchange term of Cao et al.
Conclusion (2004);
MTIO = momentum exchange term of Iverson and
The shallow water hydrosediment–morphodynamic equations, es- Ouyang (2015);
tablished based on the fundamental mass and momentum conser- MTOY = momentum exchange term of Ouyang et al.
vation laws of fluid dynamics, are generally applicable to hydraulic (2015);
β 1uu = Boussinesq momentum-distribution coefficient; ment transport and gravel/sand grain size distributions in gravel-bedded
ρf , ρs = densities of water and sediment, respectively; rivers.” Water Resour. Res., 43(10), W10436.
ρ̄1 , ρ̄2 = depth-averaged densities of the flow and bed Cui, Y., Paola, C., and Parker, G. (1996). “Numerical simulation of aggra-
dation and downstream fining.” J Hydraul Res., 34(2), 185–204.
layers;
Cui, Y., and Parker, G. (2005). “Numerical model of sediment pulses and
τ 1bot , τ 1top = bottom and top shear stresses of the flow layer; sediment-supply disturbances in mountain rivers.” J. Hydraul. Eng.,
and 10.1061/(ASCE)0733-9429(2005)131:8(646), 646–656.
τ 1botf , τ 1bots = bottom shear stresses of the fluid and solid Cui, Y., Parker, G., Lisle, T. E., Pizzuto, J. E., and Dodd, A. M. (2005).
phases. “More on the evolution of bed material waves in alluvial rivers.” Earth
Surf. Process. Landforms, 30(1), 107–114.
Cunge, J. A., Holly, F. M., and Verwey, A. (1980). Practical aspects of
computational river hydraulics, Pitman, London.
References
Cunge, J. A., and Perdreau, N. (1973). “Mobile bed fluvial mathematical
Armanini, A. (2013). “Granular flows driven by gravity.” J. Hydraul. Res., models.” La Houille Blanche, 7(7), 561–580.
51(2), 111–120. de Vries, M. (1965). “Considerations about non-steady bed-load transport
Armanini, A., and Di Silvio, G. (1988). “A one-dimensional model for in open channels.” Proc., 11th Congress, International Association for
Hydraulic Research, Delft, Netherlands, 3.8.1–3.8.8.
the transport of a sediment mixture in non-equilibrium conditions.”
J. Hydraul. Res., 26(3), 275–292. Di Cristo, C., Greco, M., Iervolino, M., Leopardi, A., and Vacca, A. (2015).
“Two-dimensional two-phase depth-integrated model for transients over
Armanini, A., Fraccarollo, L., and Rosatti, G. (2009). “Two-dimensional
mobile bed.” J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900
simulation of debris flows in erodible channels.” Comput. Geosci.,
.0001024, 04015043.
35(5), 993–1006.
Einstein, H. A. (1950). “The bed-load function for sediment transportation
ASCE/EWRI Task Committee. (2011). “Earthen embankment breaching.”
in open channel flows.” Technical Bulletins 1026, U.S. Dept. of
J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900.0000498, 1549–1564.
Agriculture, Washington, DC.
Batchelor, G. (1967). An introduction to fluid dynamics, Cambridge
Fang, H., and Wang, G. (2000). “Three-dimensional mathematical model of
University Press, Cambridge, U.K.
suspended-sediment transport.” J. Hydraul. Eng., 10.1061/(ASCE)
Bohorquez, P., and Ancey, C. (2015). “Stochastic-deterministic modeling 0733-9429(2000)126:8(578), 578–592.
of bed load transport in shallow water flow over erodible slope: Linear Frey, P., and Church, M. (2009). “How river beds move.” Science,
stability analysis and numerical simulation.” Adv. Water Resour., 325(5947), 1509–1510.
83, 36–54. Furbish, D. J., Haff, P. K., Roseberry, J. C., and Schmeeckle, M. W. (2012).
Bouchut, F., Fernandez-Nieto, E. D., Mangeney, A., and Narbona-Reina, “A probabilistic description of the bed load sediment flux. 1: Theory.”
G. (2015). “A two-phase shallow debris flow model with energy J. Geophys. Res., 117, F03031.
balance.” ESAIM-Math. Model. Numer. Anal., 49(1), 101–140. Gomez, B. (1991). “Bedload transport.” Earth-Sci. Rev., 31(2), 89–132.
Bridge, J., and Demicco, R. (2008). Earth surface processes, landforms Greco, M., Iervolino, M., Leopardi, A., and Vacca, A. (2012). “A two-phase
and sediment deposits, Cambridge University Press, Cambridge, U.K. model for fast geomorphic shallow flows.” Int. J. Sediment Res., 27(4),
Cao, Z., Egashira, S., and Carling, P. (2003). “Role of suspended-sediment 409–425.
particle size in modifying velocity profiles in open channel flows.” Greimann, B., Lai, Y., and Huang, J. (2008). “Two-dimensional total
Water Resour. Res., 39(2), 1–15. sediment load model equations.” J. Hydraul. Eng., 10.1061/(ASCE)
Cao, Z., and Hu, P. (2008). “Comment on ‘Long waves in erodible channels 0733-9429(2008)134:8(1142), 1142–1146.
and morphodynamic influence’ by Stefano Lanzoni et al.” Water Hirano, M. (1971). “River bed degradation with armouring.” Proc. Jpn.
Resour. Res., 44(6), W06601. Soc. Civ. Eng., 195, 55–65 (in Japanese).
Cao, Z., Hu, P., Hu, K., Pender, G., and Liu, Q. (2015a). “Modelling Hu, P. (2013). “Multiple time scales of fluvial processes and mathematical
roll waves with shallow water equations and turbulent closure.” river modelling.” Ph.D. thesis, Wuhan Univ., Wuhan, China
J. Hydraul. Res., 53(2), 161–177. (in Chinese).
Cao, Z., Hu, P., and Pender, G. (2011). “Multiple time scales of fluvial Hu, P., and Cao, Z. (2009). “Fully coupled mathematical modeling of tur-
processes with bed load sediment and implications for mathematical bidity currents over erodible bed.” Adv. Water Resour., 32(1), 1–15.
modeling.” J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900 Hu, P., Cao, Z., Pender, G., and Liu, H. (2014). “Numerical modelling of
.0000296, 267–276. riverbed grain size stratigraphic evolution.” Int. J. Sediment Res., 29(3),
Cao, Z., Hu, P., Pender, G., and Liu, H. (2016). “Non-capacity transport 329–343.
of non-uniform bed load sediment in alluvial rivers.” J. Mountain Hu, P., Cao, Z., Pender, G., and Tan, G. (2012). “Numerical modelling of
Sci., 13(3), 377–396. turbidity currents in the Xiaolangdi Reservoir, Yellow River, China.”
Cao, Z., Li, J., Pender, G., and Liu, Q. (2015b). “Whole-process modelling J. Hydrol., 464-465, 41–53.
of reservoir turbidity currents by a double layer-averaged model.” J. Huang, H., Imran, J., and Pirmez, C. (2005). “Numerical model of turbidity
Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900.0000951, 04014069. currents with a deforming bottom boundary.” J. Hydraul. Eng., 10.1061
Cao, Z., Li, Y., and Yue, Z. (2007). “Multiple time scales of alluvial rivers /(ASCE)0733-9429(2005)131:4(283), 283–293.
carrying suspended sediment and their implications for mathematical Huang, W., Cao, Z., Carling, P., and Pender, G. (2014). “Coupled 2D
modeling.” Adv. Water Resour., 30(4), 715–729. hydrodynamic and sediment transport modeling of megaflood due to
Iverson, R. M., and Ouyang, C. (2015). “Entrainment of bed material by Pudasaini, S. P. (2012). “A general two-phase debris flow model.”
earth-surface mass flows: Review and reformulation of depth-integrated J. Geophys. Res., 117, F03010.
theory.” Rev. Geophys., 53(1), 27–58. Pudasaini, S. P., Wang, Y., and Hutter, K. (2005). “Modelling debris flows
Kim, D. H. (2015). “H2D morphodynamic model considering wave, down general channels.” Nat. Hazards Earth Syst. Sci., 5(6), 799–819.
current and sediment interaction.” Coast. Eng., 95, 20–34. Qian, H., Cao, Z., Liu, H., and Pender, G. (2016). “Numerical modelling of
Kim, J., Ivanov, V. Y., and Katopodes, N. D. (2013). “Modeling erosion and alternate bar formation, development and sediment sorting in straight
sedimentation coupled with hydrological and overland flow processes channels.” Earth Surf. Processes Landforms, in press.
at the watershed scale.” Water Resour. Res., 49(9), 5134–5154. Qian, H., Cao, Z., Pender, G., Liu, H., and Hu, P. (2015). “Well-balanced
Kneller, B., and Buckee, C. (2000). “The structure and fluid mechanics of numerical modelling of non-uniform sediment transport in alluvial
turbidity currents: A review of some recent studies and their geological rivers.” Int. J. Sediment Res., 30(2), 117–130.
implications.” Sedimentology, 47(s1), 62–94. Rosatti, G., and Begnudelli, L. (2013). “Two-dimensional simulation of
Lanzoni, S., Siviglia, A., Frascati, A., and Seminara, G. (2006). “Long debris flows over mobile bed: Enhancing the TRENT2D model by
waves in erodible channels and morphodynamic influence.” Water using a well-balanced generalized Roe-type solver.” Comput. Fluids,
Resour. Res., 42(6), W06D17. 71, 179–195.
LeVeque, R. J. (2002). Finite volume methods for hyperbolic problems,
Seal, R., Paola, C., Parker, G., Southard, J. B., and Wilcock, P. R. (1997).
Cambridge University Press, Cambridge, U.K.
“Experiments on downstream fining of gravel. I: Narrow-channel
Li, J., Cao, Z., Pender, G., and Liu, Q. (2013). “A double layer-averaged
runs.” J. Hydraul. Eng., 10.1061/(ASCE)0733-9429(1997)123:10(874),
model for dam-break flows over mobile bed.” J. Hydraul. Res., 51(5),
874–884.
518–534.
Spinewine, B., and Capart, H. (2013). “Intense bed-load due to a sudden
Li, Y., and Xie, J. (1986). “Mathematical modelling of two-dimensional
dam-break.” J. Fluid Mech., 731, 579–614.
flow in alluvial rivers.” Chin. J. Hydraul. Eng., 11, 9–15 (in Chinese).
Toro, E. F. (2001). Shock-capturing methods for free-surface shallow flows,
Lyn, D. A. (1987). “Unsteady sediment-transport modeling.” J. Hydraul.
Wiley, Chichester, U.K.
Eng., 10.1061/(ASCE)0733-9429(1987)113:1(1), 1–15.
Marsooli, R., and Wu, W. (2015). “Three-dimensional numerical modeling van Rijn, L. C. (1993). Principles of sediment transport in rivers, estuaries
of dam-break flows with sediment transport over movable beds.” and coastal seas, Aqua Publications, Blokzijl, Netherlands.
J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900.0000947, 04014066. Wallis, G. (1969). One-dimensional two-phase flow, McGraw-Hill,
Nelson, P. A., McDonald, R. R., Nelson, J. M., and Dietrich, W. E. (2015a). New York.
“Coevolution of bed surface patchiness and channel morphology. Wu, W. (2007). Computational river dynamics, Taylor & Francis, London.
1: Mechanisms of forced patch formation.” J. Geophys. Res.: Earth Wu, W., Rodi, W., and Wenka, T. (2000). “3D numerical modeling of flow
Surf., 120(9), 1687–1707. and sediment transport in open channels.” J. Hydraul. Eng., 10.1061
Nelson, P. A., McDonald, R. R., Nelson, J. M., and Dietrich, W. E. (2015b). /(ASCE)0733-9429(2000)126:1(4), 4–15.
“Coevolution of bed surface patchiness and channel morphology. 2: Wu, W., and Wang, S. S. Y. (2007). “One-dimensional modeling of dam-
Numerical experiments.” J. Geophys. Res.: Earth Surf., 120(9), break flow over movable beds.” J. Hydraul. Eng., 10.1061/(ASCE)
1708–1723. 0733-9429(2007)133:1(48), 48–58.
Neto, C., Drew, R. E., Bonaccurso, E., Butt, H. J., and Craig, V. S. J. (2005). Xia, J., Lin, B., Falconer, R. A., and Wang, G. (2010). “Modelling
“Boundary slip in Newtonian liquids: A review of experimental dam-break flows over mobile beds using a 2D coupled approach.”
studies.” Rep. Prog. Phys., 68(12), 2859–2897. Adv. Water Resour., 33(2), 171–183.
Nicholas, A. P. (2010). “Reduced-complexity modeling of free bar morpho- Xiao, H., Young, Y. L., and Prevost, J. H. (2010). “Hydro- and morpho-
dynamics in alluvial channels.” J. Geophys. Res., 115(F4), F04021. dynamic modeling of breaking solitary waves over a fine sand beach.
Oreskes, N., Shrader-Frechette, K., and Belitz, K. (1994). “Verification, Part II: Numerical simulation.” Mar. Geol., 269(3–4), 119–131.
validation and confirmation of numerical models in the earth sciences.” Xie, J., ed. (1990). River modelling, China Water and Power, Beijing
Science, 263(5147), 641–646. (in Chinese).
Ouyang, C., He, S., and Xu, Q. (2015). “MacCormack-TVD finite differ- Zech, Y., Soares-Frazao, S., and Van Emelen, S. (2015). “Modelling of fast
ence solution for dam break hydraulics over erodible sediment beds.” hydraulic transients: Issues, challenges, perspectives.” La Houille
J. Hydraul. Eng., 10.1061/(ASCE)HY.1943-7900.0000986, 06014026. Blanche, (5), 5–15.
Papanicolaou, A. N., Elhakeem, M., Krallis, G., Prakash, S., and Edinger, J. Zhu, F., and Dodd, N. (2015). “The morphodynamics of a swash event on
(2008). “Sediment transport modeling review—Current and future an erodible beach.” J. Fluid Mech., 762, 110–140.