What happens downstream of a dam during a flush? Insight through laboratory experiments

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/384253737

What happens downstream of a dam during a flush? Insight through

laboratory experiments

Conference Paper · September 2024

CITATIONS

0

READS

92

4 authors, including:

Robin Meurice

Catholic University of Louvain

13 PUBLICATIONS 35 CITATIONS

SEE PROFILE

Brieuc Bosseler

Catholic University of Louvain

1 PUBLICATION 0 CITATIONS

SEE PROFILE

Sandra Soares-Frazão

Catholic University of Louvain

138 PUBLICATIONS 3,371 CITATIONS

SEE PROFILE

All content following this page was uploaded by Robin Meurice on 23 September 2024.

The user has requested enhancement of the downloaded file.1 INTRODUCTION

While 5% to 10% of the global capacity of dam reservoirs is already occupied by deposed sedi-

ment and 0.5% to 1% is probably lost each year, dam flushing has proved to constitute a conven-

ient way to extend the lifetime of dam reservoirs facing sedimentation (Lehner et al. 2011, White

2001). Under appropriate conditions, flushing may indeed prove very efficient. Yet, such an ex-

ceptional transient flow highly laden with sediment can have major impacts on the downstream

reach (Kondolf et al. 2014). While the populations of aquatic species can be immediately severely

impacted, their habitats, and river morphology more generally, can be seriously altered (Doretto

et al. 2022, Khakzad & Elfimov 2015).

The hydromorphodynamics of flushed reservoirs has received a vast interest from researchers,

from real field cases to laboratory physical models and numerical modelling (Haun & Olsen 2012,

Kantoush & Schleiss 2009, Petkovšek 2023, White 2001). On the opposite, investigations about

the downstream reach have been rather limited. Several field studies have mainly focused on

reporting survival rates among fish or invertebrates’ populations, while a few numerical simula-

tions have tried to reproduce the hydromorphodynamics (Espa et al. 2015, Liu et al. 2004). How-

ever, very few, if any, laboratory experiments have been reported to explicitly trying to reproduce

a dam flush and investigate the downstream hydromorphodynamics. The description of these pro-

cesses therefore usually remains rather simple. In particular, observations of the dynamics under

the water surface during the flush are lacking in the literature.

To overcome these limitations, we conducted two-dimensional idealised dam flushing experi-

ments in a laboratory flume. Both pressure and drawdown flushing modes were reproduced. Four

configurations with different upstream initial deposit depths were tested. Water levels, morpho-

logical changes, and concentrations were evaluated throughout the whole experiment.

First, the experiment will be described, including the laboratory flume, the instrumentation,

and the features of the different configurations. A particular focus will be put on the novel non-

What happens downstream of a dam during a flush? Insight

through laboratory experiments

R. Meurice, B. Bosseler, G. Noël & S. Soares-Frazão

Université catholique de Louvain, Institute of Mechanics, Materials and Civil Engineering, Louvain-la-

Neuve, Belgium

ABSTRACT: Dam flushing is often seen as a convenient way to cope with reservoir sedimenta-

tion. Yet, dam flushes may severely impact the downstream reach, which has received less atten-

tion from the scientific community than its upstream counterpart, especially in the experimental

field. We conducted idealised flushing experiments with varying initial deposit depths to enhance

understanding of the downstream hydromorphodynamics of dam flushes. Monitoring water and

bed levels, alongside concentrations, revealed distinctions between pressure and drawdown hy-

dromorphodynamics, emphasising the influence of the upstream deposit's depth. Moreover, we

propose a non-intrusive and innovative image processing approach to calibrate the explicit inver-

sion method employed in the acoustic backscattering framework to derive concentration profiles.

Although further work is required to quantify method uncertainties, our results offer a qualitative

narrative of controlled dam flushes, shedding light on the downstream dynamics.intrusive calibration method used to provide concentration profiles. Then, the results obtained

will be presented. The hydrographs through the outlet will be discussed, followed by the morpho-

logical changes and the concentration profiles. Finally, conclusions about this experimental cam-

paign will be drawn.

2 METHODS

The experiments were carried out in the installations of the Laboratoire Essais Mécaniques, Struc-

tures et génie Civil (LEMSC) at UCLouvain. The experimental flume is represented in Figure 1.

Water was fed to the flume through the upstream water tank, with a constant discharge 𝑄 = 1 𝑙/𝑠.

The flume was split into two parts by a dam with a single cavity standing as the outlet of the dam.

This dam was located at 𝑥 = 2.14 𝑚, delimiting with a porous plate (set at 𝑥 = 1.40 𝑚) a res-

ervoir filled with water and sediment stretching over 74 𝑐𝑚. The dam outlet (opening through the

plate representing the dam, see Fig. 1c for dimensions), could be open in less than 0.2 s. The

flume was horizontal, except for the last metre, where the slope reached 𝑆0 = 0.026. Downstream

of the flume, a thin crested weir set at 𝑧 = 8.5 𝑐𝑚 led the flow to a stilling basin.

Table 1. Positions of the instrumentation devices.

Dam B1 B2 B3 UVP Laser

Position (m) 2.14 1.72 2.03 2.45 2.28 2.28

Four different configurations were tested during the experimental campaign. In all configura-

tions, sand covered the first 2 𝑚 downstream of the dam over ℎ𝑠,𝑑𝑠 = 5.5 𝑐𝑚, i.e. the level of the

invert of the outlet. The water depth up- and downstream were respectively set to ℎ𝑤,𝑢𝑠 = 18 𝑐𝑚

and ℎ𝑤,𝑑𝑠 = 8.5 𝑐𝑚. Only the sediment depth upstream ℎ𝑠,𝑢𝑠 varied (see Table 2). Configuration

C1 has no sediment upstream and is seen as a reference configuration. The sediment depth up-

stream ℎ𝑠,𝑢𝑠 is then respectively inferior, equal and superior to the ceiling of the outlet, i.e. 𝑧 =

12.5 𝑐𝑚.

Table 2. Sediment depth upstream of the dam in all tested configurations.

Configuration C1 C2 C3 C4

ℎ𝑠,𝑢𝑠 (𝑐𝑚) 0 9 12.5 16

Figure 1. Experimental flume in (a) plan and (b) elevation views. Panel (c) is a sketch of the

metal plate used as a dam with its outlet. The positions of the instrumentation devices are

indicated in Table 1. Dimensions in metres.A single sediment mixture was used for the different configurations. The saturated density was

equal to 𝜌𝑠 = 2640 𝑘𝑔/𝑚³. The grain size distribution can be considered as lognormal (Fig. 2),

with a granulometric dispersion 𝜎𝑑 = √(𝑑84/𝑑16 ) = 1.56 and a median grain diameter 𝑑50 =

0.359 𝑚𝑚.

Figure 2. (a) Cumulative Density Function (CDF) and (b) Probability Density Function (PDF) of the sand

used in the experiments. The CDF was interpolated with a step of 1 𝜇𝑚 to derive the PDF. The few sieves

used to establish the CDF explain the stepped appearance of the PDF. The lognormal fit uses 𝜇𝑙𝑜𝑔 =

log (0.5𝑑50/√1 + 𝛿2) and 𝜎𝑙𝑜𝑔 = √log (𝛿2 + 1) as distribution parameters, following Thorne & Hurther

(2014), with 𝛿 = 0.4.

Opening the bottom outlet starts the flush of the reservoir. During the flush, the upstream water

discharge, flowing from the tank to the flume, was maintained constant (𝑄 = 1 𝑙/𝑠), but the dis-

charge flowing through the outlet was larger, so that the water level in the reservoir declined. At

the beginning, the water level was superior to the ceiling of the outlet (𝑧 = 12.5 𝑐𝑚) and the flush

was in pressure mode. After some time, the water level became lower than the outlet's ceiling and

the flush entered the drawdown phase. Eventually, the discharge through the outlet became equal

to that fed upstream and the flow became steady.

Ultrasonic Baumer sensors, labelled B1 to B3, provided a time series of point measurements

of the water levels. The evolution of the topography was estimated by laser profilometry (Meurice

et al. 2022). A linear laser sheet was projected through the water surface and the evolution of its

projection onto the bed was captured by a camera located outside the transparent flume (Fig. 1a).

It was therefore possible to follow the evolution of the bed during the flow along a linear profile.

Moreover, photogrammetry was used to measure the final topography, once the water had been

evacuated. Finally, a single-frequency 4 𝑀𝐻𝑧 Ultrasonic Velocity Profiler (UVP) transducer,

connected to a UVP-DUO monitoring unit (Met-Flow 2011), used acoustic backscattering to es-

timate concentration profiles, following Pedocchi & García (2012). The UVP transducer was in-

clined with a counter-clockwise angle of 40° and its nozzle was set at 𝑧 = 9.5 𝑐𝑚.

Although it was developed initially for the marine environment, we used the theoretical frame-

work defined by Thorne & Hurther (2014) to translate echo measurements made by the UVP-

DUO into concentration measurements. The fundamental equation to establish these concentra-

tion profiles along the water column is as follows:

𝑀(𝑟) = (𝜓(𝑟)𝑟

𝑘𝑠𝑘𝑡 )2

𝑉 𝑟𝑚𝑠

2 (𝑟)𝑒4𝑟𝛼(𝑟) (1)

where 𝑀(𝑟) (𝑔/𝑙) is the concentration at some range 𝑟 (𝑚) from the transducer’s nozzle, 𝜓(𝑟)

the near-field corrector as defined by Downing et al. (1995), 𝑘𝑠 = 2.24 the sediment backscatter-

ing coefficient determined thanks to the general formulation of Moate & Thorne (2012) for natural

particles, 𝑘𝑡 the instrumentation coefficient (not determined here), 𝑉 𝑟𝑚𝑠(𝑟) (𝑉) the root mean

square value of the electrical tension obtained by the UVP-DUO system for the backscattered

echo, and 𝛼(𝑟) = 𝛼𝑤 + 𝛼𝑠(𝑟) the attenuation coefficient with 𝛼𝑤

= 0.405 𝑁𝑝/𝑚 the component

due to water according to François & Garrison (1982) considering a temperature 𝑇 = 20 °𝐶, and

𝛼𝑠(𝑟) the component due to suspended particles that can be defined as:𝛼𝑠(𝑟) =

1

𝑟 ∫ 𝜉𝑀(𝑟)𝑑𝑟

𝑟

0 (2)

with 𝜉 = 4.77 × 10−4 𝑚2/𝑘𝑔 following Moate and Thorne (2012).

If the attenuation due to suspended particles 𝛼𝑠(𝑟) cannot be neglected, Equation 1 is implicit

for 𝛼𝑠(𝑟) depends on 𝑀(𝑟). An implicit inversion algorithm was extensively used in the literature

to obtain both 𝑀(𝑟) and 𝛼𝑠(𝑟). Yet, this algorithm may diverge in some cases. Furthermore, this

method requires the calibration of 𝑘𝑡, as proposed by Betteridge et al. (2008), which could not be

done during this experimental campaign. Instead, we used the explicit method proposed by Lee

& Hanes (1995) based on the derivation of 𝑀(𝑟) and that does not rest on 𝑘𝑡. Its final solution is

as follows:

𝑀(𝑟) =

𝛽²

𝛽𝑟𝑒𝑓

2 /𝑀𝑟𝑒𝑓−4 ∫ 𝛽²𝑑𝑟

𝑟

𝑟𝑟𝑒𝑓

(3)

with 𝛽 = 𝑉 𝑟𝑚𝑠𝜓𝑟𝑒2𝛼𝑤𝑟/𝑘𝑠 and 𝑟𝑟𝑒𝑓 being a reference range, typically at the top of the ultrasonic

profile. The reference concentration 𝑀𝑟𝑒𝑓 is often estimated with an intrusive sampling device

like a pump (Holdaway & Thorne 1997). Considering the dimensions of the flume and the depth

of the flow, direct sampling would probably have seriously influenced the flow. Instead of direct

measurements, we took advantage of the LPT system to estimate 𝑀𝑟𝑒𝑓 with image processing, in

a non-intrusive way. The method of Capart & Fraccarollo (2011) was adapted to determine the

concentration inside a measurement volume 𝑉 𝑀 = 311.28 𝑚𝑚³

, close to the transducer’s nozzle

and illuminated by the laser above the flume (Fig. 3). Considering all particles equal to 𝑑50, their

number N was evaluated so that:

𝑀𝑟𝑒𝑓 =

1

𝑉𝑀

4𝜋

3 (𝑑50

2 )3

𝜌𝑠𝑁 (4)

Figure 3. (a) Plan view of the final topography with respect to the camera and laser’s positioning. The parts

of the laser that are not visible in (b) are represented by dashed lines. The nozzle of the UVP sensor is in

brown. (b) Picture captured by the camera positioned as in (a) during an experimental run. The water surface

is in blue and the UVP sensor in brown. The visible parts of the laser illuminating the bed and the direction

of the flow are respectively indicated in red and green. Reference points 1 and 2 correspond to those of (a).

A research window (purple) delimits the measurement volume 𝑉 𝑀.

3 RESULTS

3.1 Hydrographs

We computed the hydrographs through the outlet using Bernoulli’s equation and water levels

recorded by the Baumer sensors. In pressure mode, the water depth at the outlet was equal to ℎ𝑜 =

0.07 𝑚, i.e. the height of the outlet. In drawdown mode however, it depended on the Froude

number. For sub- and supercritical flows, ℎ𝑜 = ℎ𝐵3 = 𝑧𝐵3− 0.055 𝑚 and ℎ0 = ℎ𝑐 respectively,

where the critical depth ℎ𝑐 (𝑚) is estimated by ℎ𝑐 = 2/3 (𝑧𝐵1− 0.055), with sensor B1 being

considered as far upstream (Matthew, 1963). The discharge can then be computed with Ber-

noulli’s equation:

𝑄 = 𝜇𝑂𝑤𝑜ℎ𝑜√2𝑔(𝑧𝐵2− 𝑧𝐵3) (5)

where 𝑤𝑜 = 0.08 𝑚 is the width of the outlet, 𝑔 = 9.81 𝑚/𝑠² the gravitational acceleration, and

𝜇𝑜 the discharge coefficient of the outlet. Neglecting lateral head losses, 𝜇𝑜 ≃ 1 in drawdownmode. In pressure mode however, we calibrated 𝜇0 to ensure the continuity of the hydrograph in

C1 and obtained 𝜇0 = 0.7. Generally, Figure 4 shows an expected decline of 𝑄 with time, as the

reservoir was emptying. This is very clear for C1 and C2 that behave rather similarly. Although

C3 shows the same general tendency, its hydrograph sees a strong drop between 𝑡 = 100 𝑠 and

𝑡 = 150 𝑠, that even reaches negative amplitudes. This is due to morphological evolutions that

lead to 𝑧𝐵3 > 𝑧𝐵2 for a while. Configuration C4 shows a similar discharge local minimum around

𝑡 = 50 𝑠. Eventually, the hydrographs of all configurations converge towards the input discharge

𝑄 = 1 𝑙/𝑠. During the experimental campaign, we observed the flow transitioned from pressure

to drawdown when 𝑧𝐵2 became lower than 𝑧𝐵2 = 12.9 𝑐𝑚. The transitions are indicated by the

vertical dashed lines in Figure 4. It appears the deeper ℎ𝑠,𝑢𝑠, the sooner the transition, all other

things being equal.

Figure 4. Hydrographs of all configurations based on Bernoulli’s equation. The vertical dashed lines corre-

spond to the pressure-drawdown transitions of each configuration.

3.2 Morphological changes

As evidenced by Figure 5, the different configurations show both similar and specific morphody-

namical patterns that can be highlighted using three zones of the downstream sand cover. The

first zone corresponds to the primary erosion pit, occurring in pressure mode and clearly visible

in the final topographies of C1 and C2. In C3 and C4, that are still very dynamic in the drawdown

mode, this zone receives deposition from the sediment flushed from upstream. Zone 2 is mainly

constituted by deposits of the sediment eroded from zone 1 during the pressure step. The front of

this zone looks to recede with ℎ𝑠,𝑢𝑠, which suggests less erosion in pressure mode in advanced

configurations. Finally, zone 3 is the sand cover that is barely affected by the flushing process.

Figure 5. Digital Elevation Models (DEMs) of the final topography of the downstream sand cover estimated

with photogrammetry for (a) C1, (b) C2, (c) C3, and (d) C4. The first five centimetres of that sand cover

were not available in the DEMs. The red curves delimit the different zones of interest, as measured in the

reference experiment C1. The noise in (b) is due to water that had not been properly evacuated when the

pictures used for the DEM of C2 were captured.

These assumptions based on the final topography can be confirmed using the backscattered

echo received by the UVP sensor. At the bed, the backscattered echo indeed strongly rises. If thesignal has not been attenuated too much up to the bed, it is sometimes possible to estimate the bed

level as a local maximum. In Figure 6a, the bed level in the transducer axis, located in zone 1 (see

Fig. 3a), sees a strong reduction in C1 and C2 in the early pressure mode. In C3, soon after the

transition to drawdown (see Fig. 4), the bed level starts to rise. In C4 however, we cannot see

erosion from Figure 6a. To the contrary, it even suggests that there is some erosion in zone 1 in

the late drawdown mode, which is contradictory to the observations. This advocates for keeping

a qualitative approach when analysing morphological changes using this method.

The UVP sensor is a convenient tool to evaluate the bed level in turbulent and turbid conditions

but its spatial coverage is very limited. Laser profilometry was hence used to monitor the evolu-

tion of the banks of the erosion pit and its contours (Fig. 6b). As evidenced by Figure 6a, the

erosion of zone 1 occurs very rapidly. Because of turbidity and turbulence, no pictures were avail-

able until 𝑡 = 22 𝑠 in C1 but it appears the banks of the pit continue to stabilise until 𝑡 = 90 𝑠

around a stability angle 𝛼𝑟 ≃ 35° as can be seen in Figure 6b.

3.3 Concentrations

Using the measurement window of Figure 3b and Equation 4, 𝑀𝑟𝑒𝑓 can be obtained at the top of

the UVP transducer’s axis (Fig. 7a). As expected, 𝑀𝑟𝑒𝑓 increases with ℎ𝑠,𝑢𝑠. After some early

oscillations, 𝑀𝑟𝑒𝑓 stabilises eventually in C1 and C2. To the contrary of C2, a surge in 𝑀𝑟𝑒𝑓 is

observed at the pressure-drawdown transition in C3. Unfortunately, the laser could not be detected

in C4. There is hence no estimation available for 𝑀𝑟𝑒𝑓.

The time series of 𝑀𝑟𝑒𝑓 can be used by the explicit inversion method (Eq. 3) to establish ver-

tical concentration profiles for C1, C2, and C3 (Fig. 7b-d). As expected, the amplitudes of the

concentration profiles tend to decrease with time in C1 and C2. As suggested by Figure 6, the

morphodynamics slows down very quickly in these configurations. Even though 𝑀𝑟𝑒𝑓 showed

similar orders of magnitude from 𝑡 = 30 𝑠 to 𝑡 = 100 𝑠, the concentrations in C3 peak at a lower

level than in C2 during that period. Again, this advocates for caution when analysing the UVP

results. From 𝑡 = 100 𝑠 to 𝑡 = 300 𝑠, the bed level keeps rising in C3 (see Fig. 6a) and so do the

concentration peaks.

Figure 6. (a) Evolution of the bed level along the UVP transducer’s axis for configurations C1 to C4. (b)

Evolution of the left bank of the erosion pit in C1.Figure 7. (a) Time series of 𝑀𝑟𝑒𝑓 for C1, C2, and C3. (b)-(d) Evolution of the vertical concentration profile

for configurations C1-C3. The profiles were smoothed with a 10 channels-wide moving average window.

4 CONCLUSION

We presented the methods and results of an experimental campaign aimed at gaining a deeper

understanding of the downstream hydromorphodynamics of dam flushes, which has received less

focus from the scientific community compared to its upstream counterpart. The experiments con-

sisted in flushing a reservoir containing layers of sand and water through a single outlet, which

was narrower than the flume, into a downstream reach already covered with sand, extending up

to the invert of the outlet. Overall, four different configurations with a varying initial upstream

deposit’s depth were investigated.

Ultrasonic Baumer sensors provided measurements of the water levels that were used to esti-

mate the discharge passing through the outlet of the dam (Fig. 4). Laser profilometry, photogram-

metry and acoustic backscattering were combined to monitor the downstream morphodynamics.

In particular, the calibration required by the traditional explicit inversion method of acoustic

backscattering was achieved following a non-intrusive approach. We indeed used the laser-cam-

era system installed for laser profilometry and image processing to evaluate the concentration

near the nozzle of the UVP used to establish concentration profiles (Fig. 7). This innovative

method appears particularly appropriate in a laboratory framework, considering the shallowness

of the flows and the unavoidable disturbances that would arise from using more common direct

sampling methods for this calibration.

The results showed that pressure flushing leads to the excavation of an erosion pit downstream

of the outlet, with deposition of the eroded sediment on the contours of the pit. This pit may then

be filled during the drawdown step. Also, the deeper the initial upstream deposit, the shorter the

pressure mode, leading to a limited depth for the erosion pit when the initial upstream deposit’s

depth is near or exceeds the ceiling of the outlet.

More work is required to refine the understanding of the hydromorphodynamical processes

during pressure and drawdown flushing and to assess the impact of different factors on the down-

stream hydromorphodynamics. Also, the uncertainties around the applications of the UPV for

acoustic backscattering in these experimental configurations must be evaluated for a proper quan-

titative assessment of the downstream morphodynamical changes. Yet, the complementary infor-

mation delivered by the different instrumentation devices already provides more insight into the

downstream hydromorphodynamical processes of a dam flush.ACKNOWLEDGMENTS

The authors address their gratitude towards Francisco Pedocchi and Rodrigo Mosquera for their

Matlab codes and help regarding the treatment of UVP data. They also thank Olivier Mariette, for

assisting them to use the UVP-DUO unit as appropriately as possible, and Hervé Capart for its

kind help in determining the reference concentration with image processing. Finally, they

acknowledge the F.S.R.-FNRS for the financial support of Robin Meurice as an FNRS fellow.

REFERENCES

Betteridge, K. F., Thorne, P. D., & Cooke, R. D. 2008. Calibrating multi-frequency acoustic backscatter

systems for studying near-bed suspended sediment transport processes. Continental Shelf Research,

28(2): 227-235.

Capart, H., & Fraccarollo, L. 2011. Transport layer structure in intense bed‐load. Geophysical Research

Letters, 38(20).

Doretto, A., Espa, P., Salmaso, F., Crosa, G., & Quadroni, S. 2022. Considering mesohabitat scale in eco-

logical impact assessment of sediment flushing. Knowledge & Management of Aquatic Ecosystems,

(423): 2.

Downing, A., Thorne, P. D., & Vincent, C. E. 1995. Backscattering from a suspension in the near field of

a piston transducer. The Journal of the Acoustical Society of America, 97(3): 1614-1620.

Espa, P., Crosa, G., Gentili, G., Quadroni, S., & Petts, G. 2015. Downstream ecological impacts of con-

trolled sediment flushing in an Alpine valley river: a case study. River Research and Applications, 31(8):

931-942.

Francois, R. E., & Garrison, G. R. 1982. Sound absorption based on ocean measurements: Part I: Pure water

and magnesium sulfate contributions. The Journal of the Acoustical Society of America, 72(3): 896-907.

Haun, S., & Olsen, N. R. B. 2012. Three‐dimensional numerical modelling of the flushing process of the

Kali Gandaki hydropower reservoir. Lakes & Reservoirs: Research & Management, 17(1): 25-33

Holdaway, G. P., & Thorne, P. D. 1997. Determination of a fast and stable algorithm to evaluate suspended

sediment parameters from high resolution acoustic backscatter systems. 7th International Conference on

Electronic Engineering in Oceanography, Conference Publication No 439, Southampton, 23-25 June

1997. London: IEE.

Kantoush, S. A., & Schleiss, A. J. 2009. Channel formation during flushing of large shallow reservoirs with

different geometries. Environmental technology, 30(8): 855-863.

Khakzad, H., & Elfimov, V. I. 2015. Environmental reviews and case studies: a review of environmental

characteristics and effects of Dez Dam flushing operation on downstream. Environmental Practice,

17(3): 211-232.

Kondolf, G. M., Gao, Y., Annandale, G. W., Morris, G. L., Jiang, E., Zhang, J., ... & Yang, C. T. 2014.

Sustainable sediment management in reservoirs and regulated rivers: Experiences from five continents.

Earth's Future, 2(5): 256-280.

Lehner, B., Liermann, C. R., Revenga, C., Vörösmarty, C., Fekete, B., Crouzet, P., ... & Wisser, D. 2011.

High‐resolution mapping of the world's reservoirs and dams for sustainable river‐flow management.

Frontiers in Ecology and the Environment, 9(9): 494-502.

Liu, J., Minami, S., Otsuki, H., Liu, B., & Ashida, K. 2004. Prediction of concerted sediment flushing.

Journal of Hydraulic Engineering, 130(11): 1089-1096.

Matthew, G. D. 1963. On the influence of curvature, surface tension and viscosity on flow over round-

crested weirs. Proceedings of the institution of civil engineers, 25(4): 511-524.

Met-Flow. 2011. UVP Monitor User’s Guide. Met-Flow SA, Lausanne, Switzerland.

Meurice, R., Martínez-Aranda, S., Ebrahimi, M., García-Navarro, P., & Soares-Frazão, S. 2022. Laser Pro-

filometry to measure the bed evolution in a dam-break flow. Journal of Hydraulic Research, 60(5): 725-

737.

Moate, B. D., & Thorne, P. D. 2012. Interpreting acoustic backscatter from suspended sediments of differ-

ent and mixed mineralogical composition. Continental Shelf Research, 46: 67-82.

Pedocchi, F., & García, M. H. 2012. Acoustic measurement of suspended sediment concentration profiles

in an oscillatory boundary layer. Continental Shelf Research, 46: 87-95.

Petkovšek, G. 2023. Monitoring and modelling of sediment flushing: a review. Scientific Research Com-

munications, 3(1).

Thorne, P. D., & Hurther, D. 2014. An overview on the use of backscattered sound for measuring suspended

particle size and concentration profiles in non-cohesive inorganic sediment transport studies. Continental

Shelf Research, 73: 97-118.

White, R. 2001. Evacuation of sediments from reservoirs. Thomas Telford.

View publication stats