Evaluation of reservoir sedimentation as a methodology for sediment yield assessment in the Mediterranean: challenges and limitations


21 févr. 2014 (il y a 3 années et 1 mois)

49 vue(s)

Evaluation of reservoir sedimentation as a methodology for
sediment yield assessment in the Mediterranean: challenges
and limitations

Joris de Vente, Jean Poesen & Gert Verstraeten

Laboratory for Experimental Geomorphology, K.U. Leuven, Belgium

1. Introduction

Reservoir sedimentation is a serious off-site consequence of soil erosion with large
environmental and economical implications. On the other hand however, reservoir
sedimentation also provides valuable information on erosion problems and sediment
transport within a drainage basin. A reservoir can be considered as a large scale
experiment, as the outlet of a giant erosion plot.
The objective of this paper is to first discuss briefly the methodology of
bathymetric surveys that are generally used to assess reservoir sedimentation and
advantages over other means to assess basin sediment export. Secondly, the data
obtained from reservoir surveys are discussed in relation to data obtained at other
scales, leading to a conceptual model of the relation between spatial scale and active
sources and sinks of sediment and sediment export at the basin scale. Yet, the next
paragraph will first illustrate why it is relevant to know more about reservoir
sedimentation and sediment export at the basin scale.

2. Problems related to reservoir sedimentation

Reservoirs around the world are losing on average about one percent of their storage
capacity annually (WCD, 2000), causing serious problems for water and electricity
supply, flood control but also for ecosystem development up-and downstream of
large dams. Consequences are especially precarious in (semi-) arid environments
where many reservoirs have been built for irrigation, water supply, flood control and
production of electricity. However, also in other areas sediment storage behind dams
can have large implications for ecosystem and coastal development downstream of
large river systems, as less sediments are delivered which will influence river and
coastal geomorphic processes (Syvitski, 2003; Vorosmarty et al., 2003; WCD, 2000;
Woodward, 1995). Furthermore, storage of possibly contaminated sediments in a
reservoir and subsequent chemical reactions occurring within the sediments due to
long term storage, cause serious problems for water quality and possibilities to
further use the sediments after dredging or flushing operations. Therefore, it is of
utmost importance to be aware of sediment yield at the basin scale, of the
composition of the sediments, and understand which factors determine the
sedimentation rate of reservoirs. This knowledge will allow to estimate the probable
lifespan of a reservoir and moreover to take proper measures against reservoir
sedimentation, water shortage and river bank and coastal erosion. At the moment
the prediction of sediment yield at the basin scale (> ~30 km²) is still one of the
largest challenges in soil erosion research. Measurements of the export of sediments

from a basin are relevant since at the moment very few models are available that
focus on assessments at spatial scales larger than ~30 km². Those models that do
focus at these scales always require measurements of sediment yield for calibration
and validation purposes.

3. Measurement of reservoir sedimentation

3.1 Bathymetric surveys

The methodology to assess the volume of sediments stored in a reservoir is currently
quite well developed and described. The main steps to be taken in a bathymetric
survey, as the methodology is called, are summarised below and are based on the
methodology used by the Spanish ‘Centro de Estudios Hidrograficos´ (CEH-CEDEX)
of the Ministry of Environment (Avendaño Salas and Calvo Sorando, 1994; Avendaño
Salas and Cobo Rayán, 1997; Avendaño Salas et al., 1995; CEDEX, 1992), and is
based on guidelines described by the International Commission on Large Dams
(ICOLD, 1989).

The first step in a bathymetric survey consists of the determination of the volume of
sediments in the reservoir. This is done by comparison of the initial reservoir volume
at the moment of dam construction (obtained from the construction plans) with the
present volume (at the moment of the capacity assessment). The present reservoir
volume is determined by a combination of photogrammetric and echo-sounding
methods. Photogrammetry is used to determine the topography of the part of the
reservoir not filled with water, and echo-sounding to determine the below-water
topography. Now, the present reservoir volume is calculated as the volume between
the contour lines of the topographical map, constructed from the photogrammetric
and bathymetric analysis. The volume of deposited sediments is defined by
subtraction of the initial from the present reservoir volume.

When beside sediment volume, also sediment mass is required, the bulk density of
the deposited sediments should be determined. Since, it is very complicated to take
undisturbed samples of submerged sediment, and the sediment bulk density changes
over time due to consolidation of the sediments over the years, usually an empirical
method is applied to determine its mean bulk density. Three factors basically
determine the bulk density of sediments, namely texture, reservoir operation and
age of the sediments. Therefore, sediment samples should be taken from the bottom
of the reservoir. It is important to take sufficient samples, to represent all parts of
the reservoir. All the sediment samples should be analysed for percentage sand, silt
and clay. The reservoir operation refers to the percentage of the total reservoir
volume that is normally filled with water. This is important, since sediments that are
often exposed to air will consolidate more than sediments that are always
submerged. Therefore, four types of reservoirs are generally distinguished: 1)
reservoirs that are normally full 2) reservoirs with moderate draw down 3) reservoirs
with considerable draw down 4) reservoirs that are normally empty. Now, the
density of the sediments is determined by combining the empirical equation of Miller
(Miller, 1953) (Eq. 1 and 2) and the equation of Lara and Pemberton (Eq. 3) (Lara
and Pemberton, 1965), which is an adaptation of the relation originally developed by
Lane and Koelzer (Lane and Koelzer, 1943).


+= 1)(ln*
pKpKpKK ++=


pWpWpWW ++=

Here, Wt refers to the average bulk density of sediments after t years of operation,
W0 stands for the mean bulk density after 1 year, pc, pm, ps, the percentages of
clay, silt and sand respectively and Wc, Wm, Ws are the coefficients of initial unit
weight for clay, silt and sand. K is the consolidation coefficient. Both the coefficients
of unit weight and the consolidation coefficient per fraction are empirical and were
reported by Lara and Pemberton (1965). The total mass of sediments present in the
reservoir is now calculated by multiplication of the average density (Wt) and the
volume of sediments present.

Now, in order to asses the basin sediment yield, the calculated sediment volume or
mass has to be corrected for the trap efficiency of the reservoir. The trap efficiency
refers to the percentage of incoming sediments that is retained in the reservoir, and
depends principally on the sediment characteristics, the stream flow velocity and the
reservoir operation (Vanoni, 1977). Different empirical relations have been
developed to estimate the trap efficiency. Usually the relation proposed by Brown
(Eq.4) (Brown, 1943) is applied, as this relation is suited for large reservoirs, and is
relatively easy to apply, but also other relations are available, that require some
more input data (Verstraeten & Poesen, 2000).

In this relation, TE, stands for trap efficiency (%), C for the capacity of the reservoir
(m³) and A for the drainage area of the basin (km²). D is a constant between 0.09
and 2.1 and depends on the reservoir type. For large reservoirs, the trap efficiency
will always be close to 100 percent. Finally, after correction for the trap efficiency,
the average area-specific reservoir sedimentation rate (t/km²/yr) can be calculated
by dividing the total mass of deposited sediments during the years of operation of
the reservoir and the surface area of the drainage basin.

3.2 Accuracy of bathymetric survey and advantage over alternatives

In general, reservoir surveys are seen as more accurate than alternative methods for
assessments of sediment export at the basin scale, since they provide direct
measurements instead of indirect estimates (Strand and Pemberton, 1987). Another
advantage of reservoir surveys is that they often provide information over long time
spans and represent both the effect of frequent and rare events. An alternative
method can be for example the development of a so called suspended sediment
rating curve. This curve is a correlation between discharge and suspended sediment
load measurements, and in general is rather accurate, but has two strong
drawbacks. First, bedload is not incorporated in the measurement, and second, since
measurements are usually collected only for a limited period (i.e. few years), these

relations often are not capable to represent the effect of extreme events, which can
have a significant effect on long-term sediment export.

Although bathymetric reservoir surveys are generally seen as more reliable than
other methods still some uncertainties are related to this method as well. For
example, errors can be made in the reservoir survey, in volume calculations, in the
conversion from sediment volume to mass and in the trap efficiency calculations. The
latter two are expected to be the most uncertain. As Verstraeten and Poesen (2001)
demonstrated, the empirical relations to determine bulk density based on fractions
sand silt and clay do not provide accurate estimates of sediment mass in small ponds
in Belgium, suggesting that it might be more reliable to use sediment volume than
sediment mass in sediment yield assessments. Nevertheless, the study by Salas and
Shin (Salas and Shin, 1999) suggested that the uncertainty related to estimation of
the sediment type and the trap efficiency are small compared to the estimation of
annual stream flow and sediment inflow, required for application of sediment rating
curves. Furthermore, though uncertainty in trap efficiency can be relatively high for
small reservoirs and ponds, for larger reservoirs, the error normally is much lower
(Verstraeten and Poesen, 2000; Verstraeten and Poesen, 2002). Altogether,
bathymetric surveys seem the most suited available method to make assessments of
long term average sediment export at the basin scale.

4. Know what you measure: some scale issues.

Reservoir sedimentation data provide a long-term average sedimentation rate. As
was discussed in the previous paragraph, the advantage of reservoir sedimentation
data is that in principle they can represent both the common and the rare events.
However, this also implies that interpretation and comparison with other
measurements of soil loss should be done carefully. Furthermore, it is important to
realise that reservoir sedimentation rates only give information on the sediment
export, and so on off-site effects of soil erosion. Reservoir sedimentation alone does
not provide information on spatial distribution of source areas of sediments nor of
sediment deposition within the catchment.

When comparing reservoir data with soil loss and sediment yield data at other scales
often large differences are apparent. The reason for this is that soil loss
measurements at one spatial scale are not representative for measurements at
another scale level. There are significant differences in sediment yield from
reservoirs and those measured in plot studies. Reservoir sedimentation represents all
active erosion and sediment transport processes in a basin. The fact that erosion
measurements at one scale are not representative for sediment yield at another
scale level is nicely illustrated by a comparison of sediment production rates at
different scales (de Vente and Poesen, submitted). From small plots to larger areas,
area-specific sediment yield generally increases due to the fact that more erosion
and transporting processes become active. However, from a certain area-threshold,
erosion becomes dominated by deposition, and the role of sediment sinks is more
important. This results in a negative relation between basin area and sediment yield.
However, there are of course exceptions to this general trend, and local conditions
determine the output at different scales. For example, in plot studies under extreme
conditions, very high sediment export can be found. Equally, the relation between
basin area and sediment yield is far from uniform. For example, a negative trend is

not always found and different groups can be identified with higher and lower
sediment yield. The explanation for this can be that though drainage area has a
strong influence on sediment transport, local conditions such as vegetation cover,
lithology and topography also have a strong influence on actual sediment export (de
Vente and Poesen, submitted). This was also suggested by various studies where
positive relations were explained by either the presence of densely vegetated slopes,
resulting in limited upland erosion and dominant bank erosion (Dedkov and
Moszherin, 1992), or by remobilization of Quaternary sediments (Church and
Slaymaker, 1989).

Reservoirs might be in operation already a long time when bathymetric survey are
applied. This means that in some areas, important land use changes might have
taken place, and erosion and sedimentation rates might have varied strongly over
time. Thus, it may perfectly be that most sediments were delivered in a short period
or during some extreme events. In this respect it is interesting to refer to various
sediment budget studies. Various examples are known where basin sediment export
was significantly influenced due to land use changes within the basin (Einsele and
Hinderer, 1997). However, several other studies (Prosser et al., 2001; Trimble,
1999; Walling, 1999) showed that though changed environmental conditions (i.e.
land use, climate) resulted in changed erosion rates and sediment production in
upslope areas, no significant changes were found in total sediment export at the
basin outlet. This suggests that basin sediment yield not necessarily responds
directly to changes within a basin, or not to the same extent. In other words, in
some cases, environmental change is primarily reflected in reorganisation within the
basin rather than in a direct change in sediment yield at the basin outlet (Phillips,

5. Further challenges and applications

As the previous paragraph makes clear, reservoir sedimentation data provide
valuable information on sediment yield and on sediment transport within a basin.
However, some issues deserve special attention. For basin management, beside
insight in the output of sediments, also information is required on the actual source
areas of the sediments, of the areas of most severe land degradation and areas
where sediment deposition takes place. Therefore, spatially distributed models and
measurements are preferred as they indicate where erosion and deposition occurs
within the catchment. However, the problem of spatially distributed models is that
model validation is troublesome since validation should be done not only at the basin
outlet but on the total spatial pattern of erosion and deposition (Jetten et al., 2003;
Takken et al., 1999). Possibilities for spatially distributed validation are by intensive
studies of overall sediment budgets, using tracers to establish source areas and
sediment dating to verify sediment storage (Walling, 1999; Wasson et al., 2002).
Another way might be by the measurement of sedimentation behind small dams in
sub-basins. Though, as the accuracy of these measurements is limited because of
the low and uncertain trap efficiency, these data merely serve as relative indications
rather than absolute measurements. Furthermore, detailed analysis of the
stratigraphy of sediments in a reservoir can be of value to link large sedimentation
periods to climate events or to periods of intensive land use.

6. Conclusions: Monitoring soil loss and sediment transport at the basin

Altogether, it can be concluded that reservoir sedimentation data are a valuable tool
for monitoring sediment dynamics at the basin scale and assess problems related the
loss of storage capacity and of sediment export to rivers downstream as well as to
coastal areas. Furthermore, reservoir sedimentation data also can provide insight in
sediment dynamics within the basin, though no information is obtained on the spatial
distribution of sediment sources. For insight in the spatial pattern of erosion and
sedimentation probably tracer studies in combination with modelling are the most
appropriate. Nevertheless, for modelling at the basin scale very few models exist
that are capable to incorporate both erosion and sediment delivery accurately, and
these models will require data on sediment yield at various scales for calibration and

The methodology of bathymetric surveys is well developed and is generally seen as
the best available method to assess basin sediment export, though with the
limitation that it can of course only be applied when a reservoir is present. Further,
in assessment of reservoir sedimentation it is important that the reservoir is in
operation long enough to include also the rare event and the trap efficiency should
be included in calculations. In order to assess the bulk density sufficient sediment
samples should be taken.

Basin sediment yield reflects all erosion and deposition processes within a basin and
thus reservoir sedimentation rates can not be compared directly with for example
soil loss measurements from runoff plots as different erosion and deposition
processes are measured. Important for interpretation is further that reservoir
sedimentation rates represent the long-term average sediment export of a basin and
therefore do not necessarily reflect all environmental changes occurring within a
drainage basin.


Avendaño Salas, C. and Calvo Sorando, J.P., 1994. La modelización matemática, ajuste y contraste del
coeficiente de entrega de sedimentos a los embalses. Aplicación al cálculo de la erosión de
cuencas fluviales., La Fundación Universidad Empresa and CEDEX, Madrid.
Avendaño Salas, C. and Cobo Rayán, R., 1997. Metodología para estimar la erosion de cuencas fluviales a
partir de la batimetría de embalses. In: J.J. Ibánez, B.L. Valero Garcés and C. Machado (Editors),
El paisaje mediterráneo a través del espacio y del tiempo. Implicaciones en la desertificación.
Geoforma Ediciones, Logroño, pp. 239-257.
Avendaño Salas, C., Cobo Rayán, R., Gómez Montaña, J.L. and Sanz Montero, E., 1995. Procedimiento
para evaluar la degradación específica (erosión) de cuencas de embalses a partir de los
sedimientos acumulados en los mismos. Aplicación al estudio de embalses Españoles. Ingeniería
Civil 99, 51-58.
Brown, C.B., 1943. Discussion of Sedimentation in reservoirs, by J. Witzig. Transactions of the American
Society of Civil Engineers 69, 1493-1500.
CEDEX, 1992. Informe sobre la situación de los estudios de aforo de solidos, batimetria y sedimentologia
de embalses hasta 1992. Centro de Estudios Hidrográficos del CEDEX, Madrid, pp. 38.
Church, M. and Slaymaker, H.O., 1989. Disequilibrium of Holocene sediment yield in glaciated British
Columbia. Nature 337, 452-454.
Dedkov, A.P. and Moszherin, V.T., 1992. Erosion and sediment yield in mountain areas of the world. In:
D.E. Walling, T.R. Davies and B. Hasholt (Editors), Erosion, debris flows and environment in
mountain regions. IAHS publication No. 209, Chengdu, pp. 29-36.

de Vente, J. and Poesen, J., submitted. Modelling soil erosion and sediment yield at the basin scale in
Mediterranean environments.
Einsele, G. and Hinderer, M., 1997. Terrestrial sediment yield and the lifetime of reservoirs, lakes, and
larger basins. Geologische Rundschau 86, 288-310.
ICOLD, 1989. Sedimentation Control of reservoirs, Guidelines, Bulletin 67. International Commission on
Large Dams, 159 pp.
Jetten, V., Govers, G. and Hessel, R., 2003. Erosion models: quality of spatial predictions. Hydrological
processes 17, 887-900.
Lane, E.W. and Koelzer, V.A., 1943. Density of sediments deposited in reservoirs, a case study of methods
used in measurement and analysis of sediment loads in streams. Report No. 9, Interagency
Committee on Water Resources.
Lara, J.M. and Pemberton, E.L., 1965. Initial unit weight of deposited sediments. In: A.R.S. USDA (Editor),
Federal Interagency Sedimentation Conference, pp. 818-845.
Miller, C.R., 1953. Determination of the Unit Weight of Sediment for Use in Sediment Volume
Computations, Bureau of reclamation, Denver, CO.
Phillips, J., 2003. Alluvial storage and the long-term stability of sediment yields. Basin research 15, 153-
Prosser, I.P., Rutherfurd, I.D., Olley, J.M., Young, W.J., Wallbrink, P.J. and Moran, C.J., 2001. Large-scale
patterns of erosion and sediment transport in river networks, with examples from Australia.
Marine Freshwater Research 52, 81-99.
Salas, J.D. and Shin, H.S., 1999. Uncertainty analysis of reservoir sedimentation. Journal of Hydraulic
Engineering 125(4), 339-350.
Strand, R.I. and Pemberton, E.I., 1987. Reservoir sedimentation. In: U.S.w. resources (Editor), Design of
small dams. United states department of the interior, bureau of reclamation, pp. 529-564.
Syvitski, J.P.M., 2003. Supply and flux of sediment along hydrological pathways: research for the 21st
century. Global and Planetary Change 39(1-2), 1-11.
Takken, I., Beuselinck, L., Nachtergaele, J., Govers, G., Poesen, J. and Degraer, G., 1999. Spatial
evaluation of a physically-based distributed erosion model (LISEM). CATENA 37(3-4), 431-447.
Trimble, S.W., 1999. Decreased rates of alluvial sediment storage in the Coon Basin, Wiscounsin, 1975-
93. Science 285, 1244-1246.
Vanoni, V.A., 1977. Sedimentation Engineering. ASCE manuals and reports on engineering practice No.
54, 745.
Verstraeten, G. and Poesen, J., 2000. Estimating trap efficiency of small reservoirs and ponds: methods
and implications for the assessment of sediment yield. Progress in Physical Geography 24(2),
Verstraeten, G. and Poesen, J., 2001. Variability of dry sediment bulk density between and within retention ponds and its impact
on the calculation of sediment yields. Earth Surface Processes and Landforms 26, 375-394.
Verstraeten, G. and Poesen, J., 2002. Using sediment deposits in small ponds to quantify sediment yield
from small catchments: possibilities and limitations. Earth Surface Processes and Landforms 27,
Vorosmarty, C.J., Meybeck, M., Fekete, B., Sharma, K., Green, P. and Syvitski, J.P.M., 2003.
Anthropogenic sediment retention: major global impact from registered river impoundments.
Global and Planetary Change 39(1-2), 169-190.
Walling, D.E., 1999. Linking land use, erosion and sediment yields in river basins. Hydrobiologia 410, 223-
Wasson, R.J., Caitcheon, G., Murray, A.S., McMulloch, M. and Quade, J., 2002. Sourcing sediment using
multiple tracers in the catchment of lake Argyle, Northwestern Australia. Environmental
Management 29(5), 634-646.
WCD, 2000. Dams and Development. A new framework for decision-making. Earthscan publications Ltd.,
London, 356 pp.
Woodward, J.C., 1995. Patterns of erosion and suspended sediment yield in Mediterranean river basins.
In: I.D.L. Foster, A.M. Gurnell and B.W. Webb (Editors), Sediment and water quality in river
catchments. Wiley, Chichester, pp. 365-389.