C OURLIS

UserManual

Version v8p4
December 1, 2022
 Contents

Introduction 4

1 Theoretical aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1 Hydrodynamics 7
1.1.1 Shallow water equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.1.2 Friction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.1.3 Vertical discretization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Sediment transport 8
1.2.1 C OURLIS - SUSPENSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2.2 C OURLIS - BEDLOAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.3 Slope stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.4 Coupling between C OURLIS and M ASCARET . . . . . . . . . . . . . . . . . . . . . . . . . 14

2 Pre-treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1 PreCourlis 16
2.2 Pretel 16

3 The inputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1 The cas and xcas files 17
3.1.1 Current set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1.2 Coming set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.2 Geometry and sediment layers 19
3.2.1 The .geo file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2.2 The .geoC file . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3 Optional inputs 20

4 General setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.1 Sediment transport 21
4.1.1 Transport mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.1.2 Sediment properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 Geometry set up 21
4.3 Initial conditions 22
4.4 Boundary conditions 22

5 Physical parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1 Sediment layers 24
5.1.1 C OURLIS - BEDLOAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.1.2 C OURLIS - SUSPENSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.2 Friction parameters 25
5.3 Slope stability model 25

6 Numerical parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6.1 Vertical discretization and clipping parameters 26
6.2 Local slope option 26
6.3 Coupling parameters 27
6.4 Uncentered scheme for SARAP kernel 27
6.5 Printouts 27

7 Calculation run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
7.1 First run 29
7.2 Calculation continuation 31

8 The outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
8.1 Mascaret outputs 32
8.2 Dictionnary 32
8.3 The listingcourlis file 32
8.4 The plong file 32
8.5 The ptravers file 32
8.6 The damoc files 32

9 Post-treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
9.1 Fudaa-mascaret 33
9.2 Pretel 33

A Transport formulae for C OURLIS-bedload . . . . . . . . . . . . . . . . . . . . . 34

A.1 Meyer-Peter & Müller (1948) 34
A.2 Lefort (2015) 34
A.3 Recking (2013) 36
A.4 Recking (2015) 36
4

B Examples of cas and xcas files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

B.1 The xcas file 38
B.2 The cas file 41
B.2.1 C OURLIS - BEDLOAD example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
B.2.2 C OURLIS - SUSPENSION example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
 Introduction

The one-dimensionnal hydro-sedimentary code C OURLIS is designed to simulate sediment

transport in rivers and reservoirs where the flows can be considered as 1D. It is weakly coupled
with the one-dimensionnal hydrodynamic software M ASCARET. Both are part of the open-
source system T ELEMAC -M ASCARET SYSTEM (www.opentelemac.org).

Figure 1: C OURLIS within the T ELEMAC -M ASCARET SYSTEM

The software supports steady and unsteady flow computations thanks to the three computational
kernels of M ASCARET:
• SARAP : the steady flow kernel for subcritical, supercritical or mixed flow regimes (finite
differences)
• REZO : the unsteady subcritical flow kernel (finite differences)
• MASCARET : the unsteady trans-critical flow kernel (finite volumes)
Complete description of those kernels can be found in the M ASCARET guides. M ASCARET
computes surface elevations, discharge and velocities along the bief. These variables are then
passed to the module C OURLIS to calculate sediment transport capacity and bed evolutions.
6

Two options are available in C OURLIS to model sediment transport :

• Modelling suspended load : the sediment particles are transported by the flow and main-
tained (possibly temporary) in suspension above the bottom by the action of shear stress
at bottom.

• Modelling bedload : the sediment particles are transported in direct contact with the
bottom or next to the bed without being affected by the fluid turbulence.

In physical systems, both mechanisms are generally observed but their mathematical represen-
tation are quite different. In C OURLIS suspended load is the solution of an advection-diffusion
equation with erosion and deposition fluxes while bedload is estimated thanks to closure relan-
tionships for sediment transport capacity and a continuity equation for the bed (Exner equation).

Figure 2: Sketch summarizing sediment transport mechanisms - saltation can be modeled with
both bedload and suspended load modules (extracted from [1])

C OURLIS - SUSPENSION model both cohesive (fine particles such as silt or clay) or non-cohesive
sediment (sand). C OURLIS - BEDLOAD is usually used to model gravel or coarse sand transport.

Figure 3: Canadian system soil classification also called SCCS classification (1987)
 1. Theoretical aspects

1.1 Hydrodynamics
M ASCARET solves the shallow water equations for incompressible flow, hydrostatic pressure
and uniform distribution of velocities along the vertical axis. The flow slope and the horizontal
curvature radius are supposed low. Wind effects on the free surface are also neglected. Flow is
described on each section through mean velocity and mean free surface elevation.
Limit
To use C OURLIS, only one reach can be considered. Besides, only the main chan-
nel can be taken into account and no singularities can be modeled.

1.1.1 Shallow water equations

The shallow water equations with variable section are given below :

∂S ∂Q
∂t + ∂ x = ql

∂Q ∂ β QS
2   (1.1)

∂t + ∂x + gZ ∂Z
∂x + J = QS ql

with :

• S the wetted area ;

• ql the lateral inflows (confluence) in m2 .s−1 ;

• Q the mean discharge across the section ;

• Z the free surface elevation ;

• g the acceleration due to gravity ;

• J the mean energy dissipation rate ;

• β accounts for the variations of the real flow velocity accross the section [5].
 8 Chapter 1. Theoretical aspects

1.1.2 Friction
Bed and shores shear stress is taken into account through the Strickler friction law :

Q2
J= 4 (1.2)
S2 K 2 Rh3

where :

• K is the Strickler coefficient ;

• Rh the hydraulic radius.

The local shear stress τtot is computed by :

ρw gU 2
τtot = ρw gRh J = 1 (1.3)
K 2 Rh3
For real cases, the hydraulic radius is usually approximated by the mean water depth across the
section : Rh ∼ H. This option is activated by default in C OURLIS.

1.1.3 Vertical discretization

Vertical discretization is used to transform cross-sections (2D) to unidimensional variables.
Curves are generated for each section to link hydraulic radius, discharges, wetted areas, wetted
perimeters and widths at the free surface to the elevation (starting at the lowest point of the
section with a spatial step chosen by the user as shown on the Figure 1.1 below).

Figure 1.1: Vertical discretization of a section : for each elevation, the width at the free surface
and the middle point abscissa of the free surface are computed

For long-term simulations with a lot of bed evolutions, vertical discretization becomes time-
consuming et can represent on its own up to more than 90% of the total calculation time [16].
Clipping parameters can reduce the calls to the vertical discretization process.

1.2 Sediment transport

C OURLIS development was initiated in 1990 by EDF to model fine particles transport for the
emptying of the Grangent reservoir in 1995. The bedload module of C OURLIS has been main-
tained and developed since 2012 and the Hydraulic Engineering Centre of EDF is officialy in
charge of this module inside EDF since 2015.
1.2 Sediment transport 9

The two modules can not be used simultaneously so no sediment mixing can be modeled with
C OURLIS.
Up to 6 sediment layers (i.e. 7 interfaces) can be modeled in C OURLIS. Sediment interfaces are
described in the geoC file (cf §3.2.2) and define homogeneous layers of sediment such as a 100%
sand layer, a 100% silt layer and a mixed silt-sand sediment layer with the suspension module
or a gravel layer and a bottom layer with the bedload module for example. Their elevation, and
thus the width of the sediment layers, can vary transversaly as shown on Figure 1.3 below.

Figure 1.2: Longitudinal profile of a reach

Figure 1.3: Cross-section view of sediment
layers in C OURLIS

The local shear stress can be decomposed into a stress associated to the skin friction, also called
efficient shear stress, τe f f and a stress due to bed forms τ f orms : τtot = τe f f + τ f orms The skin
friction coefficient is often computed from grain sizes, for example with the Strickler formula
(Equation 1.4) or the Meyer-Peter and Müller formula (Equation 1.5).
1
21 6
Kp = (1.4)
d50
1
26 6
Kp = (1.5)
d90
d50 is the median grain diameter and d90 is the diameter for which 90% of the grains are smaller.
Warning
For really fine sediment (< 200 µm and cohesive sediment), these formulae be-
come irrelevant and would give surfaces too smooth for natural beds. In practice,
skin friction should be limited to a maximal value of 85. This is the case when
looking for deposition and erosion conditions for cohesive sediment ; a critical
shear stress value of 85 Pa is systematically used.

The efficient shear stress is estimated as

 γ
K
τe f f = τtot (1.6)
Kp
with γ a coefficient, by default equal to 2.
The Shields number θ is defined as :
τtot
θ= (1.7)
(ρs − ρw )gd50
Using Equation 1.3, this expression becomes :
U2
θ= 1 (1.8)
RK 2 Rh3 d50
ρs
R is the relative reduced density R = ρw − 1 = s − 1.
 10 Chapter 1. Theoretical aspects

1.2.1 C OURLIS - SUSPENSION

Sediment in C OURLIS - SUSPENSION are considered as a passive tracer (no influence on the
hydrodynamics), hence the fluid is always considered as a Newtonian fluid. The conservative
form of the advection-diffusion equation for suspended sediment transport transport links the
averaged concentration C of the sediment with the erosion and deposition source terms E and
D.
 
∂ SC ∂ QC ∂ ∂C
+ = kx S + E − D + qcon f luence (1.9)
∂t ∂x ∂x ∂x
• kx is the longitudinal dispersion coefficient ;
• qcon f luence the local inflows at a confluence for example.
The longitudinal dispersion coefficient can be estimated from [7] as done in [6].
Cohesive and non-cohesive sediment transport are supposed independant.
Cohesive sediment fluxes
Cohesive sediment transport is considered unsaturated, i.e. concentration in the fluid is not lim-
ited (no saturation effect). Deposition flux for cohesive sediment is computed with the empirical
Krone law [8]. (  
τ
wsC 1 − τc,deposition if τ < τc,deposition ,
D= (1.10)
0 otherwise
where τ is the bed shear stress, τcr,deposition the critical bed shear stress for deposition and ws the
settling velocity for silts.
Similarly, erosion flux is computed with the empirical Partheniades law [12] :
(  
τ
M τc,erosion − 1 if τ > τc,erosion ,
E= (1.11)
0 otherwise
where τcr,erosion is the critical bed shear stress for erosion and M the Partheniades constant.
Note
For cohesive sediment, the critical bed shear stress τcr may be different for depo-
sition and erosion, i.e. there can be a range of bed shear stresses for which there
is no deposition or erosion (suspended load). This is not the case for non-cohesive
sediment.

Non-cohesive sediment fluxes

Sand transport is supposed saturated and is estimated via the Engelund-Hansen formula [4].
This formula is valid for grain sizes between 0.15 mm and 0.9 mm (d50 ) and low slopes.
s
3
Rd50 1 5
qs = 0.05 K 2 Rh3 θ 2 (1.12)
g
with θ the Shields number defined in Equation A.3.
An equilibrium concentration is then defined as Ceq = qsQρs . When the solid discharge becomes
higher than the transport capacity C > Ceq , the sediment flux becomes a deposition flux. Sim-
ilarly, when the solid discharge is lower than the transport capacity C < Ceq , the flux becomes
an erosion flux as described in the following equation :

αws (Ceq −C) if C > Ceq ,
qs = (1.13)
αws (C −Ceq ) otherwise
 1.2 Sediment transport 11

where α is an adjustment coefficient to the new equilibrium for suspension [3]. It represents the
velocity with which the sediment system tends to its new equilibrium state. The sand settling
velocity ws can be estimated thanks to the Stokes law (1851) or the Camenen formula [2].
Bed evolution
The bed evolution of the layer i during ∆t is computed with the solid discharges of deposition
and erosion of the different layers :
qdeposition ∆t
∆hi,deposition (∆t) = (1.14)
Ci
 
qerosion ∆t1
∆hi,erosion (∆t) = − ∑ hk,(∆t−∆t1 ) + (1.15)
Ci
where the variation in height of the layer i by deposition or erosion is noted ∆hi end the height
of the layer k eroded between ∆t1 and ∆t is noted hk,(∆t−∆t1 ) .

1.2.2 C OURLIS - BEDLOAD

For now, only one grain size distribution can be given to C OURLIS - BEDLOAD for all layers and
all sections. C OURLIS - BEDLOAD is therefore commonly used with only one sediment layer i.e.
two interfaces (one interface between the movable gravel bed and the water and a second one
to separate the gravel bed from immovable non-erodible materials).
Bed evolution
This module is based on the Exner continuity equation (1.16) to model gravel transport and bed
evolutions in rivers or reservoirs. It does not take into account armoring and paving phenomena.

∂ Zb ∂ Qs
ρS (1 − ρ)
+ =0 (1.16)
∂t ∂x
Resolution is done thanks to a finite volume scheme. The numerical flux term between two cells
is computed with an uncentered scheme depending on the flow regime [16] :

λi−1/2 λi+1/2 λi−1/2 λi+1/2

Sub. Sub. Sup. Sup.

x x
xi−1/2 xi xi+1/2 xi−1/2 xi xi+1/2
(a) Subcritical input – Subcritical output (b) Supercritical input – Supercritical output

λi−1/2 λi+1/2 λi−1/2 λi+1/2

Sub. Sup. Sup. Sub.

x x
xi−1/2 xi xi+1/2 xi−1/2 xi xi+1/2
(c) Subcritical input – Supercritical output (d) Supercritical input – Subcritical output

Figure 1.4: Different situations depending on the nature of the flow at the input and output of a
cell from [16] - λ is the wave velocity associated to the bed evolution.
12 Chapter 1. Theoretical aspects

Depending on the flow motion at the input and the output of the cell i, the variation of volume
in the cell i between two consecutive timsteps is computed from the following finite volume
scheme :

(V )n+1
i − (V )ni =
(Qs )ni+1 − (Qs )ni
 δt

− if (Fr )i−1/2 < 1 and (Fr )i+1/2 < 1 (case 1.4a),
 δδtx

(1.17)

(Qs )ni − (Qs )ni−1


−δx if (Fr )i−1/2 > 1 and (Fr )i+1/2 > 1 (case 1.4b),
n n


 − 2δδtx
 (Qs )i+1 − (Qs )i−1 if (Fr )i−1/2 < 1 and (Fr )i+1/2 > 1 (case 1.4c),

0 if (Fr )i−1/2 > 1 and (Fr )i+1/2 < 1 (case 1.4d).
where

• (V )ni is the volume of sediment in the cell i at the timestep n ;

• δt is the timestep ;

• δ x is the spacestep ;

• (Qs )ni is the solid discharge in the cell i at the timestep n given by the transport law ;

• Fr is the Froude number defined such as

u
Fr = √
ghw
with hw the water depth.

The bottom elevation is then updated accordingly to the variation of volume computed above.
The solid discharge (Qs )ni in cell i at the timstep n is computed with different transport formulae
described below.
Evolution of the river bottom is not applied in the same way for erosion and for deposition :

• Evolution due to erosion is applied uniformely on all points of the river bottom located
under the free surface ;

• Evolution due to deposition is applied horizontally (at constant elevation). For each hori-
zontal deposition surface, the corresponding elevation is computed by linear interpolation
of the curves generated by vertical discretization (cf Section 6.1).

Figure 1.5: Deposition and erosion mechanisms in C OURLIS - BEDLOAD

 1.2 Sediment transport 13

Transport formulae for the solid discharge

Several transport formulae (cf Appendix A for detailed expressions of these formulae) are avail-
able in C OURLIS - BEDLOAD:

• Meyer-Peter and Müller (1948) [10] : most commom transport formula, it is based on
grain mouvement threshold concept and is valid for sediment sizes between 0.4 and
29 mm and slopes between 0.4 and 2.4 % [14]. This formula is not adapted to sand
transport and is well known to underestimate solid transport at low flow rates.

• Lefort (2015) : Lefort [9] law was recently developed and validated over around 1 000
fields data and 3 400 laboratoy data. It is especially fitted for alpine rivers. This formula
is valid for a wide range of slopes I f < 20% and grain sizes 0.1 mm < d50 < 55 mm. It
was shown that this law remained relevant even with very high flow up to 28 000 m3 /s
[14]. This formula uses discharge which is easier to measure than bottom shear stress and
therefore often offers a better estimation of transport rate. This formula is particularly
suitable for gravel rivers but remains questionable for low flow.

• Recking (2013) [13] : This threshold formula was validated over 15 different reaches.
This law is valid for grain sizes between 0.4 mm and 220 mm and a wide range of slopes
0.1% < I f < 7%[14]. It is especially suitable for gravel beds.

• Recking (2015) [15]: In 2015, changes were proposed from the Recking formula to better
take into account the effect of river morphologies on solid transport. The new formula
was developped studying bedload measured in flumes (more than 12 data sets) and in the
field (more than 133 data sets) [15].

1.2.3 Slope stability

When emptying reservoirs, non-negligeable amount of sediment comes from sediment slide.
Consequently, a simple sediment slide model was proposed in C OURLIS. It is available with
both transport modules.
Slope stability is defined by two equilibrium slopes, one for underwater sediment Istab,UN and a
second one for emerged sediment Istab,EM .

Limit
Equilibrium slopes are given by the user but C OURLIS allows for only one value
of (Istab,UN , Istab,EM ) along the reach and for all sediment layers.

Equilibrium slopes can be estimated thanks to soil measurements and empirical formulae pro-
posed by Migniot [11] :
tan Istab,UN = kτy (1.18)

tan Istab,EM = k0 τy (1.19)

where

• τy is the intial stiffness of deposition;

• k = 0.01 ;

• k0 = 0.003.
 14 Chapter 1. Theoretical aspects

In all cross-section points, sediment slope is compared to the corresponding stability slope. If
the sediment slope is larger than the stability slope, sediment crumbles to match the stability
slope on this point. Emerged sediment slide model is particularly relevant when emptying reser-
voirs as described above. On the other side, underwater sediment slide is essential to stabilize
bed evolutions due to erosion. Indeed, C OURLIS can not reproduce channel enlargement, dur-
ing flood for example, as free surface elevation and energy slope are supposed constant over the
cross-section. Local shear stress is therefore maximal at the lower point where water depth is
maximal and cause excessive erosion at the lower point and high lateral slopes as shown below
on Figure 1.6.

Figure 1.6: Channel erosion with C OURLIS without slope stability model

This model, based on a single threshold, remains simple and can still be improved.
Warning
The user should also be aware that initial states can be unstable (initial geometry
slopes higher than the stability slope) and therefore generate high, non-physical,
sediment transport rates at the beginning of the simulation (mainly with C OURLIS -
SUSPENSION ).

1.2.4 Coupling between C OURLIS and M ASCARET

The coupling between C OURLIS and M ASCARET is weak. Shallow water equations are solved
independently by the M ASCARET kernel during n timesteps. Hydraulics variables are used
by C OURLIS for solid transport calculations and bed evolutions are given back to M ASCARET
after p timesteps. Each code uses its own timestep δtH and δtS . The hydraulics timestep δtH is
chosen by the user while the solid transport timestep δtS is set such as : ∆t = pδtS = nδtH The
coupling parameters p and n are chosen by the user according to bed evolutions magnitude and
speed.
When the transcritical kernel of M ASCARET (also called MASCARET) is used, the hydraulics
timestep δtH can vary to match a Courant-Friedrich-Levy number set by the user (traditionally
CFL = 0.8). This Courant-Friedrich-Levy number can reduce computational time as it adapts
the hydraulics timestep to spacestep and hydraulics variables values at each iteration:
 p  δtH
CFL = U + gH (1.20)
δx
In this case, the sediment transport timestep δtS also varies during simulation accordingly.
1.2 Sediment transport 15

Figure 1.7: Coupling between C OURLIS and M ASCARET over a time ∆t = pδtS = nδtH
 2. Pre-treatment

2.1 PreCourlis
wip

2.2 Pretel
wip
 3. The inputs

To run a simulation, at least 4 files are needed :

• The cas file for C OURLIS [Section 3.1]
• The xcas file for M ASCARET [Section 3.1]
• The geometry file for M ASCARET (geometry.geo) [Section 3.2.1]
• The sediment layers file for C OURLIS (geometry.geoC) [Section 3.2.2]
Additional inputs can be used to specify liquid and solid boundary conditions or initial states.

3.1 The cas and xcas files

3.1.1 Current set-up
The xcas file containing M ASCARET set up information can be adapted from M ASCARET calcu-
lations or generated thanks to F UDAA -M ASCARET (available at www.opentelemac.org).
The reader can refer to the M ASCARET guide for more information.
From a M ASCARET file, the C OURLIS module is activated by modifying xml lines in the xcas
file as described below :

Figure 3.1: Changes in the xcas file to activate C OURLIS

First, the C OURLIS dictionary replaces the M ASCARET one. Two lines are added to specify the
activation of C OURLIS (<optionCourlis>true</optionCourlis>) and the name of
the C OURLIS cas file (in this example called courlis.cas). These changes will call the C OURLIS
routines with the parameters given in the corresponding cas file. C OURLIS can be shut off at any
time by setting <optionCourlis>false</optionCourlis> resulting in a traditionnal
M ASCARET simulation.
 18 Chapter 3. The inputs

An example of an xcas file is given in Appendix B.1.

Warning
F UDAA -M ASCARET provides an interface to generate xcas files but since F UDAA -
M ASCARET v7p2, the software is no longer supported. Consequently, 4 tags xml
are missing from xcas files generated with F UDAA -M ASCARET:

• The C OURLIS dictionnary tag (Figure 3.1)

• The C OURLIS option tag (Figure 3.1)

• The C OURLIS key file tag (Figure 3.1)

• The uncentered scheme option tag for the permanent kernel of M ASCARET
(Figure 3.2)

If a Fortran error (At line 531 of file .../pretrait.f90) appears while

running M ASCARET with SARAP it may be that the recently developed option of
uncentered scheme for this kernel has not been added to the xcas file.

Figure 3.2: Changes in the xcas file to add the uncentered option (set to TRUE to activate it)

The cas file put in the xcas file controls C OURLIS parameters. C OURLIS keywords, gathered in
this file, are presented in the next sections. An example is given in Appendix B.2.

3.1.2 Coming set-up

Soon the code will only parse xcas files, one for M ASCARET and one for C OURLIS, and no
longer one xcas file for M ASCARET and one cas file for C OURLIS. Nevertheless, thanks to
Python script, the user will have the choice to launch the calculation with two cas files or two
xcas files.
This improvement will make M ASCARET parameters more readable and intuitive to facilitate
its use. xcas files can be converted to cas file via the command
manip_cas.py xcas2cas_1d <my_xcas_file.xcas> <output_cas_file.cas>.

Warning
The keywords are not yet all translated in the dictionnary. So, in v8p3, this con-
version will make a mixed english/french cas file!
 3.2 Geometry and sediment layers 19

3.2 Geometry and sediment layers

3.2.1 The .geo file

The geometry file contains the initial bottom elevations of the reach, i.e. the initial elevation
of the interface between the bottom and the water. Simple geometries can be generated thanks
to F UDAA -M ASCARET. Extraction of profiles from bathymetric measurements can be done
thanks to P RE C OURLIS1 . While using C OURLIS, no mesh can be generated during simulation.
That is why, unlike M ASCARET, the cross-sections must be interpolated in pre-treatment. If the
space step of the geometry is not sufficient, interpolation can also be done with P RE C OURLIS.

Warning
It is recommended to avoid as much as possible vertical walls in geometry when
using C OURLIS. Indeed, some approximations are done by M ASCARET when lat-
eral walls are vertical and hydraulics errors will be amplified by C OURLIS calcula-
tions.

3.2.2 The .geoC file

The C OURLIS geometry file contains the initial elevations of the different sediment interfaces.
For each cross-section, the first column corresponds to the bottom layer as specified in the
geo file. The next columns correspond to the different interfaces between sediment layers (as
shown below). The last column corresponds to a non-erodable substratum (hard bottom). With
the suspension module, two virtual layers are commonly set to represent sand and silt transport
i.e. the initial width of these two layers is null. A third layer can be added to represent sediment
already present in the initial step. This configuration corresponds to 4 interfaces in the geoC
file as represented below When modeling bedload, only one layer is useful and therefore 2
interfaces are given in the geoC file as represented below.

Figure 3.3: Examples of one cross-section in the geo and geoC files for (a) bedload and (b)
suspension simulations

1 In QGIS : https://plugins.qgis.org/plugins/PreCourlis/ or GitHub : https://

github.com/msecher/PreCourlis
 20 Chapter 3. The inputs

3.3 Optional inputs

Hydraulics boundary conditions are often written in an external file, for example hydrograph.loi
at the upstream boundary and elevation.loi at the downstream boundary. Initial free surface el-
evations can also be given in an external file for example called init.lig. All external files names
are specified in the xcas file. The reader can refer to the M ASCARET guide for more informa-
tion.

Similarly, sediment concentrations at boundaries can be given to C OURLIS from external files,
for example qs.loi. The corresponding LOI CONC X MODE D”ENTREE keyword, with X
the law number, should be set to 1 in the cas file (and conversely to 2 if the information is given
directly in the cas file). The name of the files are then given by LOI CONC X FICHIER.

Initial concentrations for C OURLIS - SUSPENSION can be set into an external file :

• MODE D’ENTREE DES CONCENTRATIONS INITIALES POUR COURLIS = 1 (by

default)

• FICHIER DES CONCENTRATIONS INITIALES POUR COURLIS : the external file

name

In addition, sediment characteristics can also be given in an external file :

• MODE D”ENTREE DES CARACTERISTIQUES SEDIMENTAIRES = 1 (by default)

• FICHIER DES CARACTERISTIQUES SEDIMENTAIRES : the external file name

French-English keywords
Keywords were first developed in French. Translation of keywords in English has
been started (sources/mascaret/mascaret.dico) but is not yet available in the code.
Consequently, the following sections will present the French keywords and the
corresponding English keywords when they are available in C OURLIS v8p3 (listed
in the dico_Courlis.txt generated when launching a C OURLIS simulation).
 4. General setup

4.1 Sediment transport

4.1.1 Transport mechanism
The modelling of either bedload or suspended load is done thanks to :
• OPTION CHARRIAGE or BEDLOAD OPTION : Set to TRUE or VRAI for C OURLIS -
BEDLOAD (default=FALSE)

• OPTION SUSPENSION or SUSPENSION OPTION : Set to TRUE or VRAI for C OURLIS -

SUSPENSION (default=FALSE)

If C OURLIS - BEDLOAD is chosen, the transport law is defined by LOI DE TRANSPORT :

• 1 : Meyer-Peter and Müller (1948), value by default (§A.1)
• 2 : Lefort (2015) (§A.2)
• 3 : Recking (2013) (§A.3)
• 4 : Recking (2015) (§A.4)
If C OURLIS - SUSPENSION is chosen, sand calculations can be disabled by setting CALCUL
AVEC SABLE to FALSE (by default sand calculations are allowed).

4.1.2 Sediment properties

Sediment properties can be given in an external file (by default, MODE D”ENTREE DES CARACTERISTIQUES
SEDIMENTAIRES = 1) or directly in the cas file: MODE D”ENTREE DES CARACTERISTIQUES
SEDIMENTAIRES = 2 (cf §3.3).

Using a cas file is recommanded.

4.2 Geometry set up

The geoC file name is given by FICHIER DE GEOMETRIE COURLIS. The corresponding
number of sediment layers should be specified in NOMBRE DE COUCHES.

The geo file name is given in the xcas file file to M ASCARET. It is also indicated at the beginning
of the cas file file :
 22 Chapter 4. General setup

• FICHIER DES MOT-CLES = your xcas file between single quotation mark

• PROGRAMME PRINCIPAL = ’princi.f’

• FICHIER DE GEOMETRIE = your geo file.geo between single quotation mark

4.3 Initial conditions

Initial bottom elevations are given in the geoC and geo files (see section above). Initial free
surface elevation can be given in an external file or computed by M ASCARET.
Initial concentrations in C OURLIS - SUSPENSION can be given in an external file (cf §3.3) or
directly in the cas file (MODE D’ENTREE DES CONCENTRATIONS INITIALES POUR
COURLIS = 2):

• NOMBRE DE POINTS DECRIVANT LES CONC INITIALES POUR COURLIS :

Number of points

• ABSCISSES DES CONC INI : List of points’ abscissae (along x), for example 0.0;1.0;2.0

• CONCENTRATION EN VASE INI : List of points’ silt concentrations in g/L, for ex-
ample 0.0;0.0;0.0

• CONCENTRATION EN SABLE INI : List of points’ sand concentrations in g/L, for

example 0.0;0.0;0.0

4.4 Boundary conditions

Liquid boundary conditions are specified in the xcas file.
Solid boundary conditions are specified in the cas file. Variations of sediment concentrations
with time define boundary conditions laws which are given either in an external file (especially
for large number of points, cf §3.3) or directly in the cas file:

• NOMBRE DE LOIS DE CONCENTRATION : Number of different laws

• LOI CONC NUMERO X : Law ID, X is an integer from 1 to 15

• LOI CONC X NOM : Law name of law X between single quotation mark, for example
’Constant concentration’

• LOI CONC X MODE D’ENTREE : Input type for law X, external file 1 or directly in
cas file 2 (cf §3.3)

If an external file is used, the corresponding file name should be written in :

• LOI CONC X FICHIER : External file name for law X between single quotation mark
(cf §3.3)

Otherwise, the law X should be define thanks to the following keywords :

• LOI CONC X NOMBRE DE POINTS : Number of points

• LOI CONC X TEMPS : List of points’ time in seconds, for example 0.0;3600.0

• LOI CONC X CONCENTRATION : List of points’ concentrations in g/L. for example

3.14;3.14
4.4 Boundary conditions 23

Finally, laws affected to the upstream boundary are defined by :

• NUMERO LOI CONC AMONT VASE : Law ID for silt upstream inflow

• NUMERO LOI CONC AMONT SABLE : Law ID for sand upstream inflow

Similarly, it is possible to affect laws to lateral inflows (i.e. set a sediment inflow from affluents
where a liquid inflow has already been added in M ASCARET) in the same order as they have
been given in the xcas file:

• NUMERO LOI CONC APPORT VASE : Law IDs for silt lateral inflow, for example
3;4 to refer to the third law for the first affluent and the fourth law for the second affluent.

• NUMERO LOI CONC APPORT SABLE: Law IDs for sand lateral inflow, for example
5;6

With C OURLIS - BEDLOAD, it is possible to set the upstream solid discharge through an equi-
librium slope. A law must still be given to the upstream boundary but it will not be taken into
account.

This option relies on the following keywords : CONCENTRATION AMONT CALCULEE AVEC
PENTE EQUILIBRE set to TRUE to use an equilibrium slope instead of the values given by
the law and PENTE EQUILIBRE AMONT to specify this equilibrium slope, by default this
keyword is equal to 0.01 for a 1% slope.

By default, concentration are given without voids but by setting CONCENTRATION AMONT
SANS LES VIDES to FALSE, voids will be taken into account for the upstream law.
 5. Physical parameters

5.1 Sediment layers

According to the geometrical set up given in §4.2, the sediment layers are described by the
following keywords :

• NOM DES COUCHES : List of layers names, for example ’Silt deposition’;
’Sand deposition’;’Mixed sediment layer’ or ’Gravel bed’

• CONCENTRATION DES COUCHES : List of layers concentration, for example 1000.0;1000.0;1000.0

or 2650

5.1.1 C OURLIS - BEDLOAD

Several grain diameters can be defined accordingly to the beload formula used :

• D84 : Diameter for which 84% of grains are smaller (coarse fraction) in meters

• DIAMETRE MOYEN : Mean diameter in meters

• D16 : Diameter for which 16% of grains are smaller (fine fraction) in meters

• D50 DES SABLES : Median diameter in meters

• POROSITE : Porosity, by default 0.25.

Uknown grain sizes

If D84, DIAMETRE MOYEN or D16 are set to 0.0, the following formulae are used
to dertermine there values automatically:

• d84 = 2.1d50 ,

• dm = 1.1d50 ,

• d16 = 0.5d50 .
 5.2 Friction parameters 25

5.1.2 C OURLIS - SUSPENSION

• POURCENTAGE DE SABLE : List of layers sand proportion from 0 to 100, for example
0.0;100.0;50.0

• D50 DES SABLES : List of median diameters for sands in meters in each layers, for
example 1.0;700.E-6;700.E-6

• VITESSE DE CHUTE DES SABLES : List of settling velocity for sands in each layers

• VITESSE DE CHUTE DES VASES : Settling velocity for silts

• CONTRAINTE CRITIQUE D’EROSION DES VASES : List of critical shear stresses

for silt erosion τc,erosion for each layer, for example 5.0;5.0;5.0

• CONTRAINTE CRITIQUE DE DEPOT DES VASES : Critical shear stress for silt de-
position τc,deposition

• COEFFICIENT DE PARTHENIADES : List of Partheniades coefficient for each layers

5.2 Friction parameters

The friction coefficients (cf §1.2) are given by the following keywords :

• STRICKLER DE PEAU : List of skin friction coefficient per layer

• STRICKLER TOTAL : List of bottom friction coefficient per layer

5.3 Slope stability model

The slope stablility model is described in §1.2.3. The corresponding keywords are :

• SEDIMENT SLIDE OPTION : option to activate the option when it set to TRUE (by
default, it is set to FALSE)

• MODELE DE RUPTURE DES TALUS : 1 (this keyword should always define to this
value)

• PENTE DE STABILITE DES TALUS IMMERGES : Equilibrium slope for underwa-

ter sediment Istab,UN in %, for example 0.8

• PENTE DE STABILITE DES TALUS EMERGES : Equilibrium slope for emerged

sediment Istab,EM in %, for example 0.5
 6. Numerical parameters

6.1 Vertical discretization and clipping parameters

The vertical discretization step is set in the xcas file.
To limit the calls to the vertical discretization process, two options are proposed in C OURLIS:

• CLIP ABSOLU SUR L’EVOLUTION or ABSOLUTE CLIP EVOLUTION : The ver-

tical discretization process is done every time bottom variations are higher than the clip-
ping value in meters. For example, ABSOLUTE CLIP EVOLUTION = 0.05 means
the vertical discretization process will take place every time bottom variations are higher
than 5 centimenters. It is by default equal to 1e-5 m.

• The second option suggests a clipping depending on the water depth. If the CLIPPING
OPTION is set to TRUE (by default, CLIPPING OPTION = FALSE), the vertical dis-
cretization process is called every time bottom variation is more than a given percentage
(between 0 and 1) specified in CLIP EVOLUTION. For example, CLIP EVOLUTION
= 0.05 means the vertical discretization process will happen every time bottom varia-
tion represents more than 5% of the water depth.

For now, no option has been shown to be more efficient than the other. The clipping value can
be adapted for each simulation and a compromise between errors generated by this clipping and
computational time should be found.
To start, a clipping value of 1% is recommended.

6.2 Local slope option

The energy slope value used in bedload transport laws can be computed from the head values
along the reach (computed by M ASCARET) or from the Strickler formula (Equation 1.2) using
the mean water depth, the section geometry, the liquid discharge and the Strickler coefficient.
Historically, the first method was used in C OURLIS. Yet, in some cases, this option led to
high bed instabilities caused by the hydraulics computed by M ASCARET, sometimes locally
perturbed.
That is why the option PENTE LOCALE was developped.
Recommendation
It is highly recommended to set PENTE LOCALE = TRUE.
 6.3 Coupling parameters 27

6.3 Coupling parameters

The coupling parameters described in §1.2.4 n and p correspond to the keywords NOMBRE
D’ITERATIONS HYDRAULIQUE for the hydraulics number of iterations n and NOMBRE
D’ITERATIONS SEDIMENTO for the sediment transport number of iterations p.

6.4 Uncentered scheme for SARAP kernel

When using the permanent kernel of M ASCARET, SARAP, an option is available in the xcas
file to use an uncentered scheme on the energy slope (see Figure 3.2).
Traditionnally, the energy slope term J in the shallow water equations (Equation 1.1) was ap-
proximated by upstream and downstream values. This option replaces this scheme to consider
only the upstream value as described below :

2
J= 1 1
→ J = Jupstream (6.1)
Jdownstream + Jupstream

The use of this uncentered scheme with C OURLIS - BEDLOAD has shown good results and seems
to be less mesh-sensitive.

6.5 Printouts
The outputs files names (see §8 for description of these files) are given with the following
keywords :

• FICHIER LISTING COURLIS : The listing file name, for example results.listingcourlis
;

• FICHIER RESULTATS PROFIL EN LONG : The longitudinal profile results file name,
for example results.plong ;

• FICHIER RESULTATS PROFIL EN TRAVERS : The cross-sections results file name,

for example results.ptravers ;

For the values printed in the listing, different printouts are available :

• IMPRESSION DES PARAMETRES SEDIMENTAIRES : Printouts of sediment char-

acteristics

• IMPRESSION DES INTERFACES SEDIMENTAIRES : Printouts of sediment layers

interfaces

• IMPRESSION DES PARAMETRES DE COUPLAGE : Printouts of coupling parame-

ters

• IMPRESSION DES CONC INITIALES POUR COURLIS : Printouts of initial con-

centrations

• IMPRESSION DES LOIS DE CONCENTRATION : Printouts of boundary conditions

laws

• IMPRESSION DES APPORTS SEDIMENTAIRES : Printouts of sediment inflows

28 Chapter 6. Numerical parameters

An important parameter to limit computational time and the size of the results files is the printout
step. For M ASCARET, printouts periods are given in the xcas file. For C OURLIS, several
keywords are given to limit printouts during calculation :

• PAS D’IMPRESSION COURLIS : Period in number of sediment timesteps for C OURLIS

printouts in the listing file

• PAS DE STOCKAGE POUR LE PROFIL EN LONG : Period in number of sediment

timesteps for C OURLIS printouts in the plong file

• PAS DE STOCKAGE POUR LE PROFIL EN TRAVERS : Period in number of sed-

iment timesteps for C OURLIS printouts in the ptravers file

The boolean DEBUG CHARRIAGE permits to print bedload values in the listing for each section
at each iteration to detect inconsistencies.
 7. Calculation run

7.1 First run

From your working folder, in the same way as for hydraulics simulations with M ASCARET, the
user should call the M ASCARET script:

1 [bash:] mascaret.py <your_xcas_file.xcas>

An example of a C OURLIS - BEDLOAD simulation on a Debian 10 machine is given below :

1 Loading Options and Configurations
2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
3
4 < COURLIS VERSION >
5
6 Running your CAS file(s) for:
7 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
8
9
10
11 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
12
13 S10.gfortran.dyn:
14
15
16 +> Gfortran compiler 8.3.0 with open_mpi for a Debian 10
17
18 +> root: <your root folder>
19 +> module: ad / api / artemis / bief
20 damocles / gaia / gretel / hermes
21 identify_liq_bnd / khione / mascaret / nestor
22 parallel / partel / postel3d / sisyphe
23 special / stbtel / telemac2d / telemac3d
24 tomawac / waqtel
25
26
27 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
28
29
30 ... processing the steering file
31
32 ... checking parallelisation
33
 30 Chapter 7. Calculation run

34 ... handling temporary directories

35 ~+> Creating FichierCas.txt
36
37 ... checking the executable
38
39
40 Running your simulation(s) :
41 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
42
43
44
45 In <your simulation folder>:
46 mpirun -np 1 <your root folder>/builds/S10.gfortran.dyn/bin/mascaret
47
48
49 Fichiercas : FichierCas.txt
50
51 TELEMAC-MASCARET V8P2R0beta == Copyright (C) 2000-2020 EDF-CEREMA ==
52
53 Data File : <your xcas file.xcas>
54
55 Clip Evolution = <non absolute clipping value>
56
57 Pente locale : T
58
59 Parametres utilises seulement pour les formules de transport de Recking et
de Lefort
60 D84 = < >
61 Diametre moyen = < >
62
63 Parametre utilise seulement pour la formule de transport de Lefort
64 D16 = < >
65
66 Bedload option : T Bedload simulation
67
68 Suspension option : F
69
70 sediment slide option : F
71
72 planim clipping option : F
73
74 absolute clipping = <absolute clipping value>
75
76 bedload transport law is <transport law>
77 Study name : <your study name>
78
79 Hydraulic statistics (prior solve phase)
80 ------ Geometric parameters ------
81 Number of reach(es) = 1
82 Number of cross-sections = <nb sections>
83 Number of open boundaries = 2
84 Number of junction(s) = 0
85 Number of inflow(s) = 0
86 Number of lateral weir(s) = 0
87 Number of storage area(s) = 0
88 Number of link(s) = 0
89 Number of dam(s)/weir(s) = 0
90 ------ Numerical parameters ------
91 Number of 1D nodes (mesh) = <nb sections>
92 Computation Kernel = 1
93 Initial time step = <dt>
 7.2 Calculation continuation 31

94 Variable time step? = 0

95 Simulation time = <t max>
96
97 OptionCourlis T
98 Start the Simulation...

7.2 Calculation continuation

wip
Warning
When updating your version of C OURLIS, make sure to delete the precedent dic-
tionnary file before running your simulation again.
 8. The outputs

8.1 Mascaret outputs

The hydraulics variables values are given in the opt file like in M ASCARET simulations. Hy-
draulics variables are also printed in the lis file.

8.2 Dictionnary
All keywords available for C OURLIS for a given version are listed in the dictionnary file.

8.3 The listingcourlis file

The listingcourlis file wip

8.4 The plong file

wip

8.5 The ptravers file

wip

8.6 The damoc files

wip
 9. Post-treatment

9.1 Fudaa-mascaret
wip

9.2 Pretel
wip
 A. Transport formulae for C OURLIS-
bedload

A.1 Meyer-Peter & Müller (1948)

The most common formula, and still currently used nowadays, is the Meyer-Peter & Müller [10]
formula also noted MPM. It was elaborated from bedload experiments in flumes for median
grain sizes between 0.4 à 29 mm. Sediment transport is computed above a critical stress τc =
0.047 and the solid transport rate is computed from :
CMPM
q 3
Qs = B Rgd503 (θ ∗ − θ ) 2 (A.1)
c
1−n
where the adimensional efficient shear stress θ ∗ is defined as
  32
∗ K
θ = θ (A.2)
Kp

• CMPM is a coefficient for the MPM formula. In C OURLIS, CMPM = 8 ;

• n represents the sediment porosity, by default n = 0.25. Its value can be changed by the
user via the option POROSITE;

• d50 is the median grain diameter ;

ρs
• R the relative reduced density R = ρw − 1.

The Shields parameter θ is computed from

U2
θ= 1 (A.3)
RK 2 Rh3 d50

The MPM formula is not fit for sand transport nor high slopes problems. Moreover, it often
underestimates gravel transport during periods of low flow [9].

A.2 Lefort (2015)

In Lefort [9] equation, an adimensional diameter dm∗ is defined such as:
  31
gR
dm∗ = dm (A.4)
ν2
A.2 Lefort (2015) 35

where ν is the water kinematic viscosity.

The transition discharge Q0 which separates partiel bedload from general bedload is chosen
such as :
  13  −0.5
Q0 ∗ dm K
p = C(dm ) J −m0 (A.5)
B g(Rdm )3 B Kp
with C(dm∗ ) a function to describe the Shields stress variation with the mean grain diameter:
  ∗ 
∗ 15 d
C(dm ) = 0.0444 1 + − 1.5 exp − m (A.6)
1 + dm∗ 75
The shear stress increase with slope is done thanks to the exponent m0 such as
m0 = 1.6 + 0.06 log10 (J) (A.7)
The solid discharge is then estimated as :

d84 d50 0.2

  
1.7 mR+1 F(Q)
Qs = Q J 1.65 0.5 Cdune (A.8)
1−n R d50 d16 (R + 1)(1 − n)
with m = 1.8+ 0.08 log
 10 (J)
84 d50
The term 0.5 dd50 d16 relates to the particle size range.
The dune correction coefficient Cdune serves to model the increase in friction due to the devel-
opment and the transport of dunes. It is set for grain diameters above 0.6 mm. According to
Lefort [9], it has little influence on the results.
   2  0.5 
Q
1 − 1.4 exp −0.9 KKp if dm∗ < 14 and KKp < 0.63,

Cdune = Q0 (A.9)
1 otherwise


The discharge function F describes the partial bedload of fine particles and general tranport,
first bedload and then bedload and suspended load, via the following expression :
0.06CM QQ0

 if Q < Q0 ,
 0.284 1.66 mz
 "  #
F(Q) = Q0 (A.10)
 0.6 1 − 0.938 Q
 otherwise

with mz the add-on exponent for solid discharge during the transition between bedload and
suspension:   0.192
0.38 √Q
1 + d ∗ 0.45 if Q > 3.4Q0 ,

mz = m B gdm3 (A.11)

1 otherwise
and CM a reduction factor of transport for partial bedload :
  
 1 √ Q √Q 3 < 200,
CM = 200 B gJd 3 + 2.5 if
B gJdm (A.12)
m
1 otherwise


When the ratio KKp is unknown, an alternative is suggested and can be activated in C OURLIS
with the boolean ROUGHNESS RATIO WITH QSTAR FOR LEFORT.
When set to TRUE, the friction coefficient ratio is given by :
  0.23
K Q
√ √Q 3 < 200,
0.75 if

= 200B gJd843 B gJd84 (A.13)
Kp 
0.75 otherwise
 36 Chapter A. Transport formulae for C OURLIS-bedload

A.3 Recking (2013)

The Recking [13] transport law is based on the surface diameter d84 to characterize the bed
friction and its ratio with the median diameter to characterize the particle size range.
A mobility stress θm∗ is defined, to separate partial bedload from general bedload. This parameter
is also called the Recking number [9]. This mobility stress is supposed constant for sands
(dm < 2mm : θm∗ = 0.045) and is evaluated for other particles by :
 4.4√J−1.5
d84
θm∗ = (5J + 0.06) (A.14)
d50

The associate Shields stress is computed by :

∗ J
θ84 =   −2p  (A.15)
2 Q 3p−1
Rd84 B + 74p2.6 (gJ) p B d84

with
√Q
(
0.2275 if 3
< 100,
p= B gJd84
0.3025 otherwise
An Einstein number is then defined from this two stresses by :
∗ 2.5
14θ84
Φ=  ∗ 4 (A.16)
1 + θθm∗
84

Finally, the solid discharge can be written as :

B
q
Qs = Φ 3
gRd84 (A.17)
1−n

A.4 Recking (2015)

In 2015, new estimations for the mobility stress were added to the Recking formula to discrim-
inate riffle-pools and alternate bars (where it depends on dd84
50
) from other morphologies (espe-
cially rivers with flat beds and narrow channels)[15]. Examples of those different morphologies
are shown in Figure A.1.
For these latter, θm∗ expression becomes :

θm∗ = 1.5J 0.75 (A.18)

For sand bed rivers, this choice does not matter because the Shield number is generally high
enough so errors generated by taking the average stress over the section remains negligeable.
In C OURLIS, the expression of θm∗ is set by the option MORPHOLOGY OPTION RECKING
2015. By default, it is equal to 1 for riffle-pools and alternate bars but it can be set to 2 to use
the expression above (Equation A.18).
The Shields number is computed by default from the discharge (SHIELDS SHEAR STRESS
WITH DISCHARGE FOR RECKING 2015 = TRUE) :
 2p
∗
θ84
J
= 0.015  q
Q  1 (A.19)
R B gJd 3 p2.5
84
A.4 Recking (2015) 37

with
√Q
(
0.23 if 3
< 100,
p= B gJd84
0.31 otherwise
Otherwise, the Shields number is computed from the hydraulic radius :

∗ JRh
θ84 = (A.20)
Rd84

Figure A.1: (a) A step-pool stream (Tinée), (b) a plane-bed stream (Byasse), (c) a riffle-pool
stream (Drac), (d) a braided stream (Bléone), (e) a sand-bed river (Loire). c S. Rodrigez. from
Recking et al. [15]
 B. Examples of cas and xcas files

B.1 The xcas file

1 <?xml version="1.0" encoding="ISO-8859-1"?>
2 <!DOCTYPE fichierCas SYSTEM "mascaret-1.0.dtd">
3 <fichierCas>
4 <parametresCas>
5 <parametresGeneraux>
6 <versionCode>3</versionCode>
7 <code>1</code>
8 <fichMotsCles>depot_lefort_sarap.xcas</fichMotsCles>
9 <dictionaireCourlis>dico_Courlis.txt</dictionaireCourlis>
10 <progPrincipal>princi.f</progPrincipal>
11 <sauveModele>false</sauveModele>
12 <fichSauvModele>depot_lefort_sarap.tmp</fichSauvModele>
13 <validationCode>false</validationCode>
14 <typeValidation>1</typeValidation>
15 <presenceCasiers>false</presenceCasiers>
16 <optionCourlis>true</optionCourlis>
17 <fichierMotCleCourlis>depot_lefort_courlis_sarap.cas</
fichierMotCleCourlis>
18 </parametresGeneraux>
19 <parametresModelePhysique>
20 <perteChargeConf>false</perteChargeConf>
21 <compositionLits>1</compositionLits>
22 <conservFrotVertical>false</conservFrotVertical>
23 <elevCoteArrivFront>0.05</elevCoteArrivFront>
24 <interpolLinStrickler>false</interpolLinStrickler>
25 <debordement>
26 <litMajeur>false</litMajeur>
27 <zoneStock>false</zoneStock>
28 </debordement>
29 </parametresModelePhysique>
30 <parametresNumeriques>
31 <calcOndeSubmersion>false</calcOndeSubmersion>
32 <decentrement>false</decentrement>
33 <froudeLimCondLim>1000.0</froudeLimCondLim>
34 <traitImplicitFrot>false</traitImplicitFrot>
35 <hauteurEauMini>0.005</hauteurEauMini>
36 <implicitNoyauTrans>false</implicitNoyauTrans>
37 <optimisNoyauTrans>false</optimisNoyauTrans>
38 <perteChargeAutoElargissement>false</perteChargeAutoElargissement>
39 <termesNonHydrostatiques>false</termesNonHydrostatiques>
 B.1 The xcas file 39

40 <apportDebit>0</apportDebit>
41 <attenuationConvection>false</attenuationConvection>
42 </parametresNumeriques>
43 <parametresTemporels>
44 <pasTemps>10.0</pasTemps>
45 <tempsInit>0.0</tempsInit>
46 <critereArret>1</critereArret>
47 <nbPasTemps>2</nbPasTemps>
48 <tempsMax>360000.0</tempsMax>
49 <pasTempsVar>false</pasTempsVar>
50 <nbCourant>0.8</nbCourant>
51 <coteMax>0.0</coteMax>
52 <abscisseControle>0.0</abscisseControle>
53 <biefControle>1</biefControle>
54 </parametresTemporels>
55 <parametresGeometrieReseau>
56 <geometrie>
57 <fichier>canal_depot_dx10m.geo</fichier>
58 <format>2</format>
59 <profilsAbscAbsolu>true</profilsAbscAbsolu>
60 </geometrie>
61 <listeBranches>
62 <nb>1</nb>
63 <numeros>1</numeros>
64 <abscDebut>0.0</abscDebut>
65 <abscFin>900.0</abscFin>
66 <numExtremDebut>1</numExtremDebut>
67 <numExtremFin>2</numExtremFin>
68 </listeBranches>
69 <listeNoeuds>
70 <nb>0</nb>
71 <noeuds/>
72 </listeNoeuds>
73 <extrLibres>
74 <nb>2</nb>
75 <num>1 2</num>
76 <numExtrem>1 2</numExtrem>
77 <noms>
78 <string>limite1</string>
79 <string>limite2</string>
80 </noms>
81 <typeCond>1 2</typeCond>
82 <numLoi>1 2</numLoi>
83 </extrLibres>
84 </parametresGeometrieReseau>
85 <parametresConfluents>
86 <nbConfluents>0</nbConfluents>
87 <confluents/>
88 </parametresConfluents>
89 <parametresPlanimetrageMaillage>
90 <methodeMaillage>1</methodeMaillage>
91 <planim>
92 <nbPas>500</nbPas>
93 <nbZones>1</nbZones>
94 <valeursPas>0.01</valeursPas>
95 <num1erProf>1</num1erProf>
96 <numDerProf>91</numDerProf>
97 </planim>
98 <maillage>
99 <modeSaisie>2</modeSaisie>
100 <sauvMaillage>false</sauvMaillage>
 40 Chapter B. Examples of cas and xcas files

101 <maillageClavier>
102 <nbSections>0</nbSections>
103 <nbPlages>0</nbPlages>
104 <nbZones>0</nbZones>
105 </maillageClavier>
106 </maillage>
107 </parametresPlanimetrageMaillage>
108 <parametresSingularite>
109 <nbSeuils>0</nbSeuils>
110 </parametresSingularite>
111 <parametresApportDeversoirs/>
112 <parametresCalage>
113 <frottement>
114 <loi>1</loi>
115 <nbZone>1</nbZone>
116 <numBranche>1</numBranche>
117 <absDebZone>0.0</absDebZone>
118 <absFinZone>900.0</absFinZone>
119 <coefLitMin>25.0</coefLitMin>
120 <coefLitMaj>25.0</coefLitMaj>
121 </frottement>
122 <zoneStockage>
123 <nbProfils>0</nbProfils>
124 <numProfil>-0</numProfil>
125 <limGauchLitMaj>-0</limGauchLitMaj>
126 <limDroitLitMaj>-0</limDroitLitMaj>
127 </zoneStockage>
128 </parametresCalage>
129 <parametresLoisHydrauliques>
130 <nb>2</nb>
131 <lois>
132 <structureParametresLoi>
133 <nom>loi_1_hydrogramme</nom>
134 <type>1</type>
135 <donnees>
136 <modeEntree>1</modeEntree>
137 <fichier>depot_lefort_amont.loi</fichier>
138 <uniteTps>-0</uniteTps>
139 <nbPoints>-0</nbPoints>
140 <nbDebitsDifferents>-0</nbDebitsDifferents>
141 </donnees>
142 </structureParametresLoi>
143 <structureParametresLoi>
144 <nom>loi_2_limnigramme</nom>
145 <type>2</type>
146 <donnees>
147 <modeEntree>1</modeEntree>
148 <fichier>depot_lefort_aval.loi</fichier>
149 <uniteTps>-0</uniteTps>
150 <nbPoints>-0</nbPoints>
151 <nbDebitsDifferents>-0</nbDebitsDifferents>
152 </donnees>
153 </structureParametresLoi>
154 </lois>
155 </parametresLoisHydrauliques>
156 <parametresConditionsInitiales>
157 <repriseEtude>
158 <repriseCalcul>false</repriseCalcul>
159 </repriseEtude>
160 <ligneEau>
161 <LigEauInit>false</LigEauInit>
 B.2 The cas file 41

162 <modeEntree>2</modeEntree>
163 <formatFichLig>1</formatFichLig>
164 <nbPts>0</nbPts>
165 <branche>-0</branche>
166 <abscisse>-0</abscisse>
167 <cote>-0</cote>
168 <debit>-0</debit>
169 </ligneEau>
170 </parametresConditionsInitiales>
171 <parametresImpressionResultats>
172 <titreCalcul>depot_lefort_sarap</titreCalcul>
173 <impression>
174 <impressionGeometrie>false</impressionGeometrie>
175 <impressionPlanimetrage>false</impressionPlanimetrage>
176 <impressionReseau>false</impressionReseau>
177 <impressionLoiHydraulique>false</impressionLoiHydraulique>
178 <impressionligneEauInitiale>false</impressionligneEauInitiale>
179 <impressionCalcul>true</impressionCalcul>
180 </impression>
181 <pasStockage>
182 <premPasTpsStock>1</premPasTpsStock>
183 <pasStock>3601</pasStock>
184 <pasImpression>3601</pasImpression>
185 </pasStockage>
186 <resultats>
187 <fichResultat>depot_lefort_sarap_ecr.opt</fichResultat>
188 <postProcesseur>2</postProcesseur>
189 </resultats>
190 <listing>
191 <fichListing>depot_lefort_sarap.lis</fichListing>
192 </listing>
193 <fichReprise>
194 <fichRepriseEcr>depot_lefort_sarap_ecr.rep</fichRepriseEcr>
195 </fichReprise>
196 <rubens>
197 <ecartInterBranch>1.0</ecartInterBranch>
198 </rubens>
199 <stockage>
200 <option>1</option>
201 <nbSite>0</nbSite>
202 </stockage>
203 </parametresImpressionResultats>
204 <parametresVariablesCalculees>
205 <variablesCalculees>false false false false false false false false
false false false false false false false</variablesCalculees>
206 </parametresVariablesCalculees>
207 <parametresVariablesStockees>
208 <variablesStockees>true false false false false true true false false
false true false false false false false false false false true false
false true false false false false false false true false false false
false false false false false false false false false</variablesStockees>
209 </parametresVariablesStockees>
210 </parametresCas>
211 </fichierCas>

B.2 The cas file

B.2.1 C OURLIS - BEDLOAD example
1 FICHIER DES MOT-CLES = ’permanent.xcas’
2 PROGRAMME PRINCIPAL = ’princi.f’
3 FICHIER DE GEOMETRIE = ’canal_depot.geo’
 42 Chapter B. Examples of cas and xcas files

4
5
6 /
7 /
8 //----------------------------COURLIS--------------------------------
9 /
10 / PARAMETRES SEDIMENTAIRE (1)
11 /
12 OPTION CHARRIAGE = VRAI
13 LOI DE TRANSPORT = 2
14 MODE D’ENTREE DES CARACTERISTIQUES SEDIMENTAIRES = 2
15 FICHIER DES CARACTERISTIQUES SEDIMENTAIRES = ’’
16 NOMBRE DE COUCHES = 1
17 CLIP ABSOLU SUR L’EVOLUTION=0.05
18 PENTE LOCALE = TRUE
19 D84 =0.0573
20 DIAMETRE MOYEN =0.03
21 D16 =0.00573
22 D50 DES SABLES = 0.027285714
23 DEBUG CHARRIAGE = FALSE
24 CONCENTRATION AMONT CALCULEE AVEC PENTE EQUILIBRE = TRUE
25 PENTE EQUILIBRE AMONT = 0.01
26 /
27 / COUCHES SEDIMENTAIRES (2)
28 /
29 NOM DES COUCHES =
30 ’Fond fixe’
31 CONCENTRATION DES COUCHES =
32 2650.
33 POROSITE = 0.25
34 /
35 STRICKLER DE PEAU =
36 37.9
37 STRICKLER TOTAL =
38 25.
39 /
40 / TALUS (2)
41 /
42 MODELE DE RUPTURE DES TALUS = 1
43 /
44 PENTE DE STABILITE DES TALUS IMMERGES = 1.
45 PENTE DE STABILITE DES TALUS EMERGES = 1.
46 /
47 /
48 / GEOMETRIE SEDIMENTAIRE (1)
49 /
50 FICHIER DE GEOMETRIE COURLIS = ’canal_depot_dx10m.geoC’
51 /
52 / LOI DE CONCENTRATION (1)
53 /
54 NOMBRE DE LOIS DE CONCENTRATION = 1
55 /
56 / LOI CONC NUMERO 1 (2)
57 /
58 LOI CONC 1 NOM = ’qs_depot’
59 LOI CONC 1 MODE D’ENTREE = 2
60 LOI CONC 1 FICHIER = ’’
61 LOI CONC 1 NOMBRE DE POINTS = 2
62 LOI CONC 1 TEMPS =
63 0.0;360000.0
64 LOI CONC 1 CONCENTRATION =
 B.2 The cas file 43

65 3.13920472;3.13920472
66 /
67 /
68 / APPORTS DE CONCENTRATION (1)
69 /
70 / CONDITIONS AMONT (2)
71 /
72 NUMERO LOI CONC AMONT VASE = 1
73 NUMERO LOI CONC AMONT SABLE = 1
74 /
75 /
76 / IMPRESSION-RESULTAT COURLIS(1)
77 /
78 FICHIER LISTING COURLIS = ’depot_lefort_sarap.listingcourlis’
79 FICHIER RESULTATS PROFIL EN LONG = ’depot_lefort_sarap.plong’
80 FICHIER RESULTATS PROFIL EN TRAVERS = ’depot_lefort_sarap.ptravers’
81 /
82 / IMPRESSION COURLIS (2)
83 /
84 IMPRESSION DES PARAMETRES SEDIMENTAIRES = FAUX
85 IMPRESSION DES INTERFACES SEDIMENTAIRES = faux
86 IMPRESSION DES PARAMETRES DE COUPLAGE = FAUX
87 IMPRESSION DES CONC INITIALES POUR COURLIS = FAUX
88 IMPRESSION DES LOIS DE CONCENTRATION = faux
89 IMPRESSION DES APPORTS SEDIMENTAIRES = FAUX
90 /
91 / PAS DE STOCKAGE ET D’IMPRESSION COURLIS (2)
92 /
93 PAS D’IMPRESSION COURLIS = 36
94 PAS DE STOCKAGE POUR LE PROFIL EN LONG = 36
95 PAS DE STOCKAGE POUR LE PROFIL EN TRAVERS = 36
96 /
97 /
98 / COUPLAGE (1)
99 /
100 NOMBRE D’ITERATIONS HYDRAULIQUE = 1
101 NOMBRE D’ITERATIONS SEDIMENTO = 1

B.2.2 C OURLIS - SUSPENSION example

1 FICHIER DES MOT-CLES = ’hydro_Torr.xcas’
2 PROGRAMME PRINCIPAL = ’princi.f’
3 FICHIER DE GEOMETRIE = ’GaronneInterp.georef’
4 /
5 /
6 //----------------------------COURLIS--------------------------------
7 /
8 / PARAMETRES SEDIMENTAIRE (1)
9 /
10 OPTION SUSPENSION = VRAI
11 CALCUL AVEC SABLE = VRAI
12 MODE D’ENTREE DES CARACTERISTIQUES SEDIMENTAIRES = 2
13 FICHIER DES CARACTERISTIQUES SEDIMENTAIRES = ’’
14 NOMBRE DE COUCHES = 2
15 CLIP EVOLUTION=0.05
16 PENTE LOCALE = false
17 /
18 / COUCHES SEDIMENTAIRES (2)
19 /
20 NOM DES COUCHES =
21 ’depot des vases’;’depot des sables ’
22 CONCENTRATION DES COUCHES =
 44 Chapter B. Examples of cas and xcas files

23 1000.0;1000.0
24 POURCENTAGE DE SABLE =
25 0.0;100.0
26 D50 DES SABLES =
27 1.0;700.E-6
28 VITESSE DE CHUTE DES SABLES =
29 0;0
30 CONTRAINTE CRITIQUE D’EROSION DES VASES =
31 5.0;5.0
32 COEFFICIENT DE PARTHENIADES =
33 0.01;0.01
34 STRICKLER DE PEAU =
35 85;85
36 STRICKLER TOTAL =
37 25;25
38 / pas de depot de vase
39 VITESSE DE CHUTE DES VASES = 0.005
40 CONTRAINTE CRITIQUE DE DEPOT DES VASES = 0.1
41 /
42 / TALUS (2)
43 /
44 MODELE DE RUPTURE DES TALUS = 1
45 PENTE DE STABILITE DES TALUS IMMERGES = 0.8
46 PENTE DE STABILITE DES TALUS EMERGES = 0.5
47 /
48 / EQUATION DE TRANSPORT (2)
49 /
50 COEFFICIENT DE DIFFUSION DES VASES = 0
51 CONVECTION DES TRACEURS POUR COURLIS = vrai
52 OPTION DE CONVECTION POUR LES TRACEURS POUR COURLIS = 4
53 ORDRE DU SCHEMA DE CONVECTION VOLUMES FINIS POUR COURLIS = 3
54 PARAMETRE W DU SCHEMA DE CONVECTION VOLUMES FINIS POUR COURLIS = 0
55 LIMITEUR DE PENTE DU SCHEMA VOLUMES FINIS POUR COURLIS = vrai
56 /
57 /
58 / GEOMETRIE SEDIMENTAIRE (1)
59 /
60 FICHIER DE GEOMETRIE COURLIS = ’GaronneVide.georefC’
61 /
62 /
63 / CONDITIONS INITIALES COURLIS (1)
64 /
65 MODE D’ENTREE DES CONCENTRATIONS INITIALES POUR COURLIS = 2
66 NOMBRE DE POINTS DECRIVANT LES CONC INITIALES POUR COURLIS = 2
67 ABSCISSES DES CONC INI =
68 0.0;9079.16854889
69 CONCENTRATION EN VASE INI =
70 0.0;0.0
71 CONCENTRATION EN SABLE INI =
72 0.0;0.0
73 /
74 /
75 / LOI DE CONCENTRATION (1)
76 /
77 NOMBRE DE LOIS DE CONCENTRATION = 2
78 /
79 / LOI CONC NUMERO 1 (2)
80 /
81 LOI CONC 1 NOM = ’CL amont vase’
82 LOI CONC 1 MODE D’ENTREE = 2
83 LOI CONC 1 FICHIER = ’’
 B.2 The cas file 45

84 LOI CONC 1 NOMBRE DE POINTS = 2

85 LOI CONC 1 TEMPS =
86 0.0;900000.0
87 LOI CONC 1 CONCENTRATION =
88 10.0;10.0
89 /
90 / LOI CONC NUMERO 2 (2)
91 /
92 LOI CONC 2 NOM = ’CL amont sable’
93 LOI CONC 2 MODE D’ENTREE = 2
94 LOI CONC 2 FICHIER = ’’
95 LOI CONC 2 NOMBRE DE POINTS = 2
96 LOI CONC 2 TEMPS =
97 0.0;900000.0
98 LOI CONC 2 CONCENTRATION =
99 0.00;0.00
100 /
101 /
102 / APPORTS DE CONCENTRATION (1)
103 /
104 / CONDITIONS AMONT (2)
105 /
106 NUMERO LOI CONC AMONT VASE = 1
107 NUMERO LOI CONC AMONT SABLE = 2
108 /
109 / APPORTS SEDIMENTAIRES (2)
110 /
111 NUMERO LOI CONC APPORT VASE =
112 3;3
113 /
114 /
115 / IMPRESSION-RESULTAT COURLIS(1)
116 /
117 FICHIER LISTING COURLIS = ’./mascaret0.listingcourlis’
118 FICHIER RESULTATS PROFIL EN LONG = ’./mascaret0.plong’
119 FICHIER RESULTATS PROFIL EN TRAVERS = ’./mascaret0.ptravers’
120 /
121 /
122 / IMPRESSION COURLIS (2)
123 /
124 IMPRESSION DES PARAMETRES SEDIMENTAIRES = FAUX
125 IMPRESSION DES INTERFACES SEDIMENTAIRES = faux
126 IMPRESSION DES PARAMETRES DE COUPLAGE = FAUX
127 IMPRESSION DES CONC INITIALES POUR COURLIS = FAUX
128 IMPRESSION DES LOIS DE CONCENTRATION = faux
129 IMPRESSION DES APPORTS SEDIMENTAIRES = FAUX
130 /
131 / PAS DE STOCKAGE ET D’IMPRESSION COURLIS (2)
132 /
133 PAS D’IMPRESSION COURLIS = 1200
134 PAS DE STOCKAGE POUR LE PROFIL EN LONG = 1200
135 PAS DE STOCKAGE POUR LE PROFIL EN TRAVERS = 1200
136 /
137 /
138 / COUPLAGE (1)
139 /
140 NOMBRE D’ITERATIONS HYDRAULIQUE = 1
141 NOMBRE D’ITERATIONS SEDIMENTO = 1
142 &FIN
 [1] URL http://theses.univ-lyon2.fr/documents/getpart.php?id=
lyon2.2008.pintomartins_d&part=154405.

[2] B. Camenen. Simple and general formula for the settling velocity of particles. Journal of
Hydraulic Engineering, February 2004. pp 229-233.

[3] D. Chen, K. Acharya, and M. Stone. Sensitivity analysis of nonequilibrium adaptation

parameters for modeling mining-pit migration. Journal of Hydraulic Engineering, pages
806–811, 2010.

[4] F. Engelund and E. Hansen. A monograph on sediment transport in alluvial streams.

Technical University of Denmark., 1967.

[5] Nicole Goutale and Fabrice Zaoui. Mascaret theory guide v8p1, December 2019.

[6] A. Hauet, G. Buannic, G. Antoine, M. Jodeau, M. Esteves, and F. Thollet. In situ measure-
ment of advection-dispersion processes in a torrential river, comparison with numerical
modeling. In River Flow, 2014.

[7] S. M. Kashefipour and R.A. Falconer. Longitudinal dispersion coefficient in natural chan-
nels. Water Research, 36, 2002.

[8] R. Krone. 1962.

[9] Philippe Lefort. Morphodynamique fluviale - Approches théorique et expérimentale. Ponts

et Chaussées (Presses), June 2018. ISBN 978-2-85978-516-1.

[10] E. Meyer-Peter and R. Müller. Formulas for bed-load transport. Hydraulic Engineering
Reports, June 1948.

[11] C. Migniot. Tassement et rhéologie des vases. La Houille Blanche, 1-2, 1989.

[12] E. Partheniades. 1965.

[13] A. Recking. Simple method for calculating reach-averaged bed-load transport. Journal of
Hydraulic Engineering, 139, Issue 1, January 2013. doi: https://doi.org/10.1061/(ASCE)
HY.1943-7900.0000653.

[14] Alain Recking. Evaluation des formules de transport solide en rivière avec prise en compte
de l’échelle temporelle. Rapport PGRN, May 2010.
Bibliography 47

[15] Alain Recking, Guillaume Piton, Daniel Vazquez-Tarrio, and Gary Parker. Quantifying
the morphological print of bedload transport. Earth Surface Processes and Landforms,
41, Issue 6, November 2015. doi: https://doi.org/10.1002/esp.3869.

[16] Philippe Ung. Code mascaret-courlis - aspect charriage, March 2018.