ESurfEarth Surface DynamicsESurfEarth Surf. Dynam.2196-632XCopernicus PublicationsGöttingen, Germany10.5194/esurf-4-461-2016Armor breakup and reformation in a degradational laboratory experimentOrrúClarac.orru@tudelft.nlBlomAstridhttps://orcid.org/0000-0001-7988-3165UijttewaalWim S. J.Department of Hydraulic Engineering, Faculty of Civil Engineering and
Geosciences, Delft University of Technology, P.O. Box 5048, 2600
GA, Delft, the NetherlandsClara Orrú (c.orru@tudelft.nl)9June2016424614705January201620January20164May201624May2016This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://esurf.copernicus.org/articles/4/461/2016/esurf-4-461-2016.htmlThe full text article is available as a PDF file from https://esurf.copernicus.org/articles/4/461/2016/esurf-4-461-2016.pdf
Armor breakup and reformation was studied in a laboratory experiment using a
trimodal mixture composed of a 1 mm sand fraction and two gravel fractions
(6 and 10 mm). The initial bed was characterized by a stepwise downstream
fining pattern (trimodal reach) and a downstream sand reach, and the
experiment was conducted under conditions without sediment supply. In the
initial stage of the experiment an armor formed over the trimodal reach. The
formation of the armor under partial transport conditions led to an abrupt
spatial transition in the bed slope and in the mean grain size of the bed
surface, as such showing similar results to a previous laboratory experiment
conducted with a bimodal mixture. The focus of the current analysis is to
study the mechanisms of armor breakup. After an increase in flow rate the
armor broke up and a new coarser armor quickly formed. The breakup initially
induced a bed surface fining due to the exposure of the finer substrate,
which was accompanied by a sudden increase in the sediment transport rate,
followed by the formation of an armor that was coarser than the initial one.
The reformation of the armor was enabled by the supply of coarse material
from the upstream degrading reach and the presence of gravel in the original
substrate sediment. Here armor breakup and reformation enabled slope
adjustment such that the new steady state was closer to normal flow
conditions.
Introduction
The formation of an armor has two different origins
e.g.,. A static armor is created when
there is a lack of sediment supply and limited shear stress values enable
entraining of only the finer grains present at the bed surface. A mobile armor
forms when a sediment mixture governed by a range in grain sizes is supplied
from upstream. The over-representation of the coarse grains at the bed
surface then serves to increase their transport rate. The coarsening of the
bed surface of a static armor is mainly caused by the winnowing or washing-out of fines from the bed e.g.,.
For a mobile armor the coarsening is mainly related to kinematic sorting or
the infiltration of fines into the bed
e.g.,.
Armoring processes have been mainly investigated under controlled laboratory
conditions. Many authors have focused on the characteristics of the bed
structure during armoring studying the bed arrangement and the stability of
cluster particles
e.g.,.
The stability of a gravel bed can be increased by the presence of cluster
particles
e.g.,.
The armoring process is influenced by the nature of the sediment supply
e.g.,. The experiments by
showed that a decrease in the sediment supply rate
corresponds to a more efficient armor development. Also, the grain size
distribution of the bed material and, in particular, the increase in sand
content plays a role in the armoring process by reducing vertical sorting
or by reducing surface roughness and facilitating the
rearrangement of the bed particles . Another aspect is
the influence of different flow rates on the development of the armor
. indicated a general
increase in the armoring degree with an increase in the flow rate. They also
showed that temporally varying flows lead
to stronger armoring than steady flows.
Still little is known on the behavior of an armor under peak flow
. Studies have demonstrated that, during peak
flow, armored surfaces can be stable and persist or broken and reform during
the waning phase of the flood event. were the first
to present a field case on armor persistence during peak flow. Using their
surface-based transport model predicted the
persistence of an armor of a certain grain size distribution using transport
rates measured in the field. described a gravel bed reach
of the Colorado River with a persistent armor. The sediment supplied from
upstream provided the sediment for the replacement of the entrained
particles. Grains from a persistent armor can exchange grains with the
transported load . Despite these examples of armor
persistence, there are also examples in which the armor does not persist
during peak flow. described armor breakup and reformation
in a large river regulated by a dam. Their field measurements illustrated
that the armor did not persist under peak flow and reformed at smaller flow
values.
Various causes of armor breakup have been distinguished. The laboratory
experiments by provided detailed measurements of armor
breakup after a flood wave under bedform-dominated conditions. Here armor
breakup appeared to result from the turbulence created in the trough zones of
the migrating bedforms. A finer armor reformed in the waning phase of the
flood wave. conducted a laboratory experiment studying
armor breakup under a shortage of sediment supply. The armor was created
under base flow and its stability was tested under a stepwise increase in the
discharge. Armor breakup was due to an increased mobility of the coarse
particles and led to a sudden increase in the bed load transport. Other
causes of armor breakup are the sediment supply from upstream and the
presence of bedload sheets. The supply of finer material can lead to an
increased mobility of the coarse sediment and can therefore mobilize the
armor
.
The presence of bedload sheets can reduce armor stability by reducing the bed
roughness and increasing the flow velocity
. The
conditions and parameters that determine the persistence or breakup of an
armor are still unclear. One of the problems is the fact that it is difficult
to measure temporal changes in the bed surface texture during peak flow in
the field .
The objective of this paper is to study the mechanisms of armor breakup and
their consequences by providing detailed measurements of spatial and temporal
changes in the bed surface texture under controlled laboratory conditions.
The changes in the bed surface texture were measured during flow using the
technique developed by . We examine the
stability of an armor under peak flow under a limited sediment supply. The
experimental setup was characterized by an initial streamwise fining
pattern, following a previous laboratory study on armor formation in which a
reach characterized by an initially gradual fining pattern under a limited
sediment supply rate and partial transport conditions, developed into a more
abrupt spatial transition in grain size and slope .
The results presented by suggest a similarity with a
gravel–sand transition
e.g.,,
yet the mechanisms prevailing in that experiment were likely not comparable
to the ones governing natural gravel–sand transitions. Relevant mechanisms in
the development of a gravel–sand transition are the progradation of a gravel
wedge , basin subsidence
, base level change
, and suspended load
, which do not play a role in either the
experiment or in the current laboratory
experiment. Nevertheless, the current experiments may provide insight in the
coexistence of an armor and a sand patch during peak flow.
Experimental setupExperimental settings
The experiment was carried out at the Water Laboratory of the Faculty of
Civil Engineering and Geosciences of Delft University of Technology. The
experiment was conducted in a tilting flume that was 14 m long,
0.40 m wide, and 0.45 m high. The upstream water supply was
controlled by a water pump and the downstream water level was set by a
tailgate located at the downstream end of the flume. No sediment was supplied
from upstream. At the downstream end of the flume the transported sediment
was collected in a sediment trap.
We used a trimodal sediment mixture that was composed of a sand fraction
(D50,1= 1 mm) and two gravel fractions, a medium fraction
(D50,2= 6 mm) and a coarse fraction
(D50,3= 10 mm). The sediment fractions were painted in
different colors to enable measurements of the grain size distribution of the
bed surface using the image analysis technique of . The
fine fraction was left with its natural color, the medium fraction was
painted yellow-green, and the coarse fraction was painted medium turquoise.
Flume setup and initial bed for Experiment T1. The red numbers
indicate the volumetric gravel fraction content in each compartment.
As mentioned above the experimental setup was similar to the one by
. Here we use a trimodal mixture rather than a bimodal
one. The initial bed was installed with an imposed stepwise fining pattern.
The upstream reach was composed of the trimodal mixture (the trimodal reach)
and the downstream reach was composed of sand (the sand reach). The trimodal
reach was characterized by 10 compartments characterized by a length of
0.40 m (Fig. ), and the length of the most upstream
compartment was 0.88 m. Each compartment was characterized by a
different initial volumetric fraction content of the three fractions
(Fig. ). The sand content increased in streamwise direction
in steps of 10 % for each compartment. The bed slope was set equal to
0.0022. We refer the reader to for details regarding the method to
install the initial bed.
Water discharge, Qw, and base level (i.e., water surface
elevation at the downstream end), ηw, imposed in Experiments
T1 and T2. The water surface elevation was measured at x=10.62m.
An initial experiment (T1) was conducted to create an armor under base
flow conditions. The flow conditions were increased in Experiment T2 to
assess the stability of the armor under peak flow
(Fig. ). The flow regime was subcritical in both
experiments. During the initial experiment the water discharge was equal to
0.0465 m3s-1. The downstream water surface elevation was
adjusted during the first few flow hours and maintained constant for the
remainder of the experiment. The total duration of Experiment T1 was
16 h, and the time interval of experiment T1 is from -16 to
0 h. At the beginning of Experiment T2 the water discharge was set
equal to 0.0547 m3s-1 and the downstream water surface
elevation was decreased through lowering the tailgate. For the remainder of
the experiment water discharge and water surface elevation were maintained
constant (Fig. ). Experiment T2 lasted 4 h, and the
time interval of Experiment T2 is from 0 to 4 h.
Measurements
The water discharge was constantly measured at the input water pipe using a
flow meter that measured the travel time of an acoustic signal. Two laser
instruments mounted on a carriage were used to measure longitudinal profiles
of the bed and water surface elevations at the center of the flume
approximately every 20 min. The laser instrument used to measure bed
elevation was placed in a watertight eye-shaped box that was slightly
submerged to avoid reflections of the signal on the water surface. At the
downstream end of the flume (at x=10.62 m) the water surface elevation
was continuously measured using a pitot tube. A linear position sensor was
connected to the pitot tube by a hose and positioned beside the flume. The
transported sediment was caught in a sediment trap at the downstream end of
the flume. The sediment was pumped to a small tank which was placed on a
scale for measuring the submerged sediment mass.
The grain size distribution of the bed surface was measured during flow over
the entire observation section (≈ 10m). The measurements
were taken using the image analysis technique developed by
, which is based on color segmentation,
i.e., the division of the pixels of an image into color groups
(Fig. ). To this end, each grain size fraction was painted in a
different color. The equipment used to take the images of the bed surface was
composed of a carriage that enabled moving the equipment along the flume and
a floating device (Fig. ). A camera was connected to the
carriage. The floating device, which was slightly submerged, was connected to
the carriage using PVC pipes. The lower pipes had a smaller diameter than the
upper ones to allow for vertical motion of the floating device and automatic
adjustment of the level of the floating device to spatial changes in the
water surface elevation. The design and material of the floating part of the
measurement equipment were here optimized compared to the one presented by
to reduce its submersion. For this purpose the bottom of
the upstream V-shaped part of the floating device was designed with a small
inclination to obtain a lift force from the flow. The floating device was
made out of thin transparent Plexiglas®. Six
small LED lights were mounted on the frame of the floating device to
illuminate the bed surface. The images of the bed surface were processed
using a Matlab algorithm that provided the areal fraction content of a
surface covered by a certain color (i.e., grain size). The areal fraction
contents resulting from image analysis were converted into volumetric
fractions by applying the conversion model of .
Example of color segmentation for an image of the bed surface
applied to determine the grain size distribution of the bed surface.
Floating part of the equipment for measuring the grain size
distribution of the bed surface during the laboratory experiments. Here the
bottom and the walls of the floating part are made of thin
Plexiglas® plates and higher
Plexiglas® sheets are attached along the
walls to avoid water overflow. The numbers indicate (1) position of the
camera, (2a) upper pipe, (2b) lower pipe, (3) carriage, (4) LED light
attached to the cooling plate, and (5) floating part.
Imposed and measured geometric mean grain size of the bed surface
sediment at various times for Experiment T1.
Formation of the initial armor (Experiment T1)
In this section we briefly describe Experiment T1, conducted to create the
armor under the imposed supply limited conditions. Over the trimodal reach
the initially high rate of sand entrainment combined with the slightly mobile
gravel fractions quickly resulted in a coarse bed surface
(Fig. ). A limited amount of sand was available at the bed
surface. In the remaining part of Experiment T1 the coarser fractions were
less or no longer mobile (partial transport conditions) due to the formation
of an armored bed structure that enhanced particle stability. The prevailing
mechanisms were winnowing and the kinematic sieving of sand. Bed surface
coarsening was observed between x=1 and x=4.5m
(Fig. ). Some randomly arranged irregularities over the
armor created gaps between the gravel particles where the finer substrate was
exposed.
Measured water surface and bed elevation profiles of Experiment
T2 at t=0h 07 min, 2 h 08 min,
3 h 08 min, and 4 h 08 min. The enlarged
window shows the degradation occurring during the armor breakup. Flow is from
left to right. The profiles were smoothed using a Gaussian filter.
The armoring occurring over the trimodal reach limited the sediment supplied
to the sand reach, which resulted in a strong bed degradation over the sand
reach (Fig. ), as was observed by . This
spatial difference in degradation resulted in a sudden decrease in the bed
elevation between the trimodal and the sand reach. Adjusting to limited
sediment supply conditions the bed approached a final state that was
characterized by zero sediment transport. For the sand reach, the state of
zero sediment transport was governed by a much smaller flow velocity (and
thus
larger flow depth) than for the upstream trimodal reach, which resulted in
the observed step in bed elevation. This sudden decrease in bed elevation
resulted in a streamwise increase in the water surface elevation, which is a
Bernoulli effect e.g.,. The trimodal reach of
Experiment T1 was characterized by the presence of an M1 backwater curve
due to the different bed slopes and thus flow depths between the two reaches.
The final stage of the bed of Experiment T1 was characterized by an abrupt
transition in slope and bed surface texture between the two reaches
(Fig. ).
The upstream section of the trimodal reach was governed by an imbricated
structure (Fig. ). This structure formed in the initial part
of the experiment when the gravel fractions were still slightly mobile. The
particles quickly found a stable position that enhanced the armor stability.
The armor was considered fully developed after 16 flow hours (i.e., at time
0 h) when no relevant changes in the bed surface texture were
observed and the sediment transport rate reached approximately zero.
Imbrication of the bed surface sediment at the end of Experiment
T1. Flow is from left to right.
Breakup and reformation of the armor layer (Experiment T2)Bed surface texture
At the beginning of the armor breakup experiment, Experiment T2, the flow
velocity was increased by increasing the water discharge by 18 % and
lowering the downstream water level (Fig. ). Armor
breakup and reformation covered a short period. After the increase in the
flow velocity, the armor started to break up in several sections of the
trimodal reach. In the upstream part of the trimodal reach the substrate was
coarser due to a limited amount of sand and the bed was highly imbricated
(Fig. ). This imbrication enhanced armor stability and
consequently some sections of the armor did not break up. Figure
shows a section of the trimodal reach where some part of the armor was
broken. Initially, the dislodgement of a few gravel particles enabled the
entrainment of the finer subsurface material over a small section of the bed
(Fig. a). Subsequently, the sand entrainment appeared to enhance
the gravel mobility, which may have extended the breakup exposing the
subsurface over a wider section (Fig. b). Blurriness in the images
shown in Fig. a and b indicates the entrainment of sand. The
measurements of the bed surface texture show a fining of the bed between x=1 and x=4.5m (Figs. –). The
measurement taken after 7 min (point 3 in Fig. )
corresponds to the bed state of Fig. c. As we observed that the
bed surface at the moment of the breakup was finer (Fig. b), we
hypothesize that the bed surface at the moment of the breakup was finer
(point 2 in Fig. ) than the moment of Fig. c
(point 3 in Fig. ).
Armor breakup and reformation between Compartments 6 and 7:
(a) initial local breakup, (b) widening of the breakup and
exposure of the finer substrate, (c) reformation of the mobile
armor. Flow is from left to right. Blurriness in images (a)
and (b) indicates the entrainment of sand (orange arrows).
Measured geometric mean grain size of the bed surface sediment at
various times for Experiment T2. Points 1 to 4 in the enlarged window show
the temporal change in the bed surface. Point 2 indicates a hypothetical
(i.e., not measured; see Fig. b) finer surface at the moment of
the breakup. Point 3 corresponds to the coarser bed surface shown in
Fig. c right after the reformation of the armor.
Measured change in the geometric mean grain size of the bed surface
sediment at various locations in Experiments T1 and T2. The blue dotted
line indicates the moment of the increase in flow rate (t=0h) and
the red dashed line indicates the moment of armor breakup
(t=4min).
The breakup and bed surface fining were quickly followed by the formation of
a mobile armor that was coarser than the initial one (point 4 of
Fig. ). The coarse sediment supplied from upstream enabled
the formation of a new armor and the presence of gravel particles in the
substrate aided this armor reformation. After armor reformation a limited
amount of sand was present at the bed surface (Fig. ). The
fact that the reformed armor was slightly coarser than the initial one
resulted in a slight downstream coarsening in the gravel reach.
Bed elevation
The total amount of degradation due to armor breakup depended on the texture
of the substrate material. In the upstream part of the trimodal reach, the
substrate was coarser, which limited bed degradation. We observed a lateral
variation in degradation characterized by a stronger degradation at one side
of the flume that is not evident in our measurements since bed profiles were
measured only at the center of the flume. The breakup led to a fast
degradation which was arrested by the reformation of the mobile armor
(Fig. ). The degradation was not uniform in streamwise
direction. Over the reach that suffered from the breakup the slope decreased
to adjust to a situation with a shortage of sediment supply. The
redistribution of the sediment led initially to aggradation downstream of the
breakup area and subsequently to the progradation of the front between the
trimodal reach and the sand reach (Fig. ). The progradation had
ceased at the end of the experiment.
The decrease in the slope made the bed locally approach normal flow
conditions, which indicates a state in which the slope of the water surface
is equal to the slope of the channel bed. If boundary conditions of a reach
(i.e., upstream water discharge and sediment supply rate, as well as the
downstream water surface elevation) are constant for a sufficiently long
time, it will reach normal flow conditions (provided that particle abrasion,
tributaries, and subsidence or uplift do not play a role). However, under
conditions of partial transport, in which the coarse fractions of the
sediment are immobile, the associated armor can prevent this adaptation of
the bed and its approach to normal flow conditions (see Experiment T1 and
). Here armor breakup enabled adjustment of the bed
slope such that the bed slope became closer to the water surface slope and
the final bed configuration was closer to normal flow.
Hydraulic conditions of experiment T1 and T2:
(a) flow depth, (b) flow velocity, (c) Froude
number, (d) Shields stress for the sand fraction (logarithmic
scale), (e) Shields stress for the fine gravel fraction (logarithmic
scale), and (f) Shields stress for the coarse gravel fraction
(logarithmic scale). The Shields stress values were determined accounting for
a sidewall correction due to .
During Experiment T1 and T2 the hydraulic conditions varied not only
spatially but also temporally (Fig. ). The adjustment to the
limited sediment supply led to a coarsening over the trimodal reach and a
larger flow depth over the sand reach (Fig. a). Due to the
temporal increase in the water discharge and a lowering of the downstream
water surface elevation, flow velocities were increased in Experiment T2
relative to Experiment T1. The highest flow velocities were present at the
moment of armor breakup (Fig. b). Both experiment were
characterized by a subcritical flow regime (Fig. c). Shields
stress values for each sediment fraction were determined accounting for the
sidewall correction of . Highest Shields stress values over
the trimodal reach were observed when the armor broke up
(Fig. d, e, f). Despite the high Shields stress values observed
over the entire trimodal reach, the breakup was local and not uniform over
the reach. Reformation of the mobile armor was associated with a decrease in
the Shields stress (Fig. d, e, f).
Measured sediment transport rate (a) computed from the
migration of the front between the trimodal reach and the sand reach using
Eq. (), (b) at the downstream end of the flume.
Horizontal line intervals indicate time-averaged values.
Figure shows the local sediment transport rate computed from
the migration of the front between the trimodal reach and the sand reach
(Fig. a), as well as the sediment transport rate measured at
the downstream end of the flume (Fig. b). We determined the
local sediment transport rate at the position of the front,
qfront, from the streamwise migration speed of the front as
proposed by using the simple-wave relation:
c=qfrontcbΔ,
where c denotes the migration speed of the prograding front in streamwise
direction (here determined at the crest), cb=1-p (p being the bed
porosity; we assume p= 0.4), and Δ denotes the height of the
prograding front.
The sediment load at the front was composed of the three grain size fractions,
whereas the sediment load at the downstream end of the flume was composed of
mainly the sand fraction. The transport rates at the front and downstream end
of the flume are of the same order of magnitude (Fig. ). At
both locations the sediment transport rate shows a (sudden) increase, which
was followed by a gradual decrease. The peak in the sediment transport rate
at the downstream end of the flume shows a time lag. Also, the decrease in the
sediment transport rate was slower than at the front. The sediment transport
rate at the front increased as a result of the armor breakup. This increase
at the front coincided with the rapid entrainment of the substrate material
and the decrease in the sediment transport rate corresponded to the
reformation of the mobile armor. However, the sediment transport rate at the
downstream end seemed to be less affected by the breakup and seemed to
respond to primarily the increased flow rate.
Discussion
Our experiment indicates that armor breakup and reformation can be a fast
process and that the resulting changes in bed elevation may be limited.
Information on the timescales of armor breakup and reformation and the order
of magnitude of resulting bed elevation changes may be relevant to flood risk
and navigation studies over armored reaches, for instance in the upstream
part of the IJssel branch of the Dutch Rhine. The dislodgement of armor
particles may expose and release fine sediment from the substrate, and such a
sudden supply of fine sediment may result in local aggradation in the
downstream reach that creates problems for navigation. The stability of an
armor is also of interest to the design of granular filters aimed at
protecting structures from scour and in the operation of dams, for instance
in the design of flushing flows undertaken to release sediment from a
reservoir.
It is still difficult to predict armor breakup. This is due to (a) the
possible randomness associated with the position of the breakup and (b) the
fact that its mechanisms are not sufficiently clear. Here and in the studies
by and the increase in the flow
discharge led to an increase in the sediment mobility, which caused the armor
to break up. The results of our experiment showed an almost uniform increase
in the Shields stress over the trimodal reach; however, the breakup was local
and its position seemed to be random. Let us consider causes additional to an
increased sediment mobility that may have played role in the breakup process
in our experiment. Sediment supplied from upstream may induce breakup by
destabilizing the armor . Such mobilization has been
encountered when finer material is supplied to the armor surface
. By
filling the gaps of the coarse surface the fine sediment reduces the bed
friction, which increases the flow velocity. A similar process occurs when
bed friction is reduced due to the transport of finer material in bedload
sheets .
These potential causes can be ruled out in our experiment because the
material supplied from the upstream slightly degrading section was mainly
coarse. We expect that the destabilization of the armor may also be ascribed
to the impact of transported particles onto the gravel particles that are at
rest. attributed the destabilization of the armor in his
experiments to turbulence originated by migrating bedforms. In our plane bed
experiment, additional turbulence may have been created by irregularities at
the armor surface. Turbulence and the resulting pressure fluctuations due to
these irregularities may have caused or increased entrainment of sand from
the sandy substrate in the more downstream part of the trimodal reach, which
may have enhanced gravel mobility and facilitated the lengthening of the
breakup.
The moment of the breakup was characterized by an increase in the bed load
transport rate, which was also observed in the laboratory experiments of
and . The increase in the sediment
transport rate was rapid and sudden and corresponded to the mobility of the
armor particles and the entrainment of the finer substrate material. A
temporal fining of the bed surface characterized the moment of the breakup
and it was followed by a coarsening due to the reformation of an armor that
was coarser than the initial one. Similar results were presented by
; however, in their case armor reformation occurred only
under base flow. In their field case the increased degree of armoring was
caused by partial transport conditions. In our experiment the armor reformed
under continued peak flow conditions. The coarser upstream section acted as a
source of sediment for the finer downstream reach. This supply provided the
replacement for the entrained material, likely resulting in a quick
reformation of the armor. The gravel present in the original substrate may
have aided armor reformation. Possible causes of the fact that the reformed
armor was coarser than the initial one are (a) the supply from upstream
being mostly gravel, (b) the sand supplied from upstream not being trapped in
the zone of the broken up armor, and (c) the higher flow rate not allowing
for the sand to deposit and remain between the gravel particles.
Similar to the experiment of , Experiment T1 showed
the development of a reach characterized by a gradual fining pattern and
uniform slope into an abrupt transition in bed surface grain size and slope
as encountered in natural gravel–sand transitions
e.g.,.
Here Experiment T2 provides new insights into the coexistence of an armor
and a sand reach during peak flow, such as the starvation and degradation of
the sand reach, progradation of the gravel front, impact of a reduced slope
on the progradation of the gravel front, and armor reformation.
Conclusions
A flume experiment was conducted to investigate the stability of an armor
under conditions with a limited sediment supply. The armor formed over a bed
initially characterized by a gradual fining pattern under base flow
conditions. The armor broke up after an increase in the flow rate and rapidly
reformed under continued peak flow conditions. Despite the fact that the
Shields stress almost uniformly increased over the trimodal reach, the
breakup was local and not uniform over the reach. Besides the increased flow
rate, multiple factors may have contributed to the armor breakup such as the
impact of coarse sediment supplied from upstream and turbulence created due
to irregularities in the armored surface.
The breakup was characterized by a temporal fining of the bed surface due to
local degradation and the exposure of finer substrate sediment. Despite the
limited sediment supply conditions after the armor breakup, a new armor, which
was coarser than the initial one, quickly formed. Coarse sediment supplied by
the upstream degrading reach provided the sediment required for the armor to
reform, which was aided by the gravel in the substrate sediment. Armor
breakup coincided with a sudden and local increase in the sediment transport
rate due to the entrainment of the finer substrate material. This was
followed by a gradual decrease, which corresponded to the armor reformation.
The breakup led to a decrease in the Shields stress and the local bed slope.
Partial transport conditions can prevent the adjustment of the bed and the
approach to normal flow conditions (i.e., the equilibrium state was
characterized by a backwater). Here armor breakup enabled the adjustment of
the bed slope such that the final bed configuration was closer to normal
flow, i.e., the bed slope was closer to the water surface slope.
Acknowledgements
The authors especially thank the technicians of the Water Laboratory of Delft
University of Technology for their assistance during the
experiments. Edited by: J. Willenbring
ReferencesAndrews, E. D. and Erman, D. C.: Persistence in the size distribution of
surficial bed material during an extreme snowmelt flood, Water Resour.
Res., 22, 191–197, 10.1029/WR022i002p00191, 1986.Bacchi, V., Recking, A., Eckert, N., Frey, P., Piton, G., and Naaim, M.: The
effects of kinetic sorting on sediment mobility on steep slopes, Earth
Surf. Proc. Land., 39, 1075–1086, 10.1002/esp.3564,
2014.
Bagnold, R. A.: The physics of blown sand and desert dunes, Methuen, New
York, 1941.Church, M., Hassan, M. A., and Wolcott, J. F.: Stabilizing self-organized
structures in gravel-bed stream channels: Field and experimental
observations, Water Resour. Res., 34, 3169–3179, 10.1029/98WR00484, 1998.Clayton, J. A. and Pitlick, J.: Persistence of the surface texture of a
gravel-bed river during a large flood, Earth Surf. Proc. Land., 33, 661–673,
10.1002/esp.1567, 2008.Curran, J. and Waters, K. A.: The importance of bed sediment sand content for
the structure of a static armor layer in a gravel bed river, J. Geophys.
Res.-Earth, 119, 1484–1497, 10.1002/2014JF003143, 2014.
Dietrich, W., Kirchner, J., Ikeda, H., and Iseya, F.: Sediment supply and the
development of the coarse surface layer in gravel-bedded rivers, Nature, 340,
215–217, 1989.
Douglas, J., Gasiorek, J., Swaffield, J., and Jack, L.: Fluid Mechanics, 5th
Edn., Pearson Prentice Hall, Harlow, UK, 2005.
Frings, R. M.: Sedimentary characteristics of the gravel–sand transition in
the River Rhine, J. Sediment. Res., 81, 52–63, 2011.Guney, M., Bombar, G., and Aksoy, A.: Experimental Study of the Coarse
Surface Development Effect on the Bimodal Bed-Load Transport under Unsteady
Flow Conditions, J. Hydraul. Eng.-ASCE, 139, 12–21,
10.1061/(ASCE)HY.1943-7900.0000640, 2013.Hassan, M. A. and Church, M.: Experiments on surface structure and partial
sediment transport on a gravel bed, Water Resour. Res., 36, 1885–1895,
10.1029/2000WR900055, 2000.Hassan, M. A., Egozi, R., and Parker, G.: Experiments on the effect of
hydrograph characteristics on vertical grain sorting in gravel bed rivers,
Water Resour. Res., 42, W09408, 10.1029/2005WR004707, 2006.Heays, K. G., Friedrich, H., and Melville, B. W.: Laboratory study of
gravel-bed cluster formation and disintegration, Water Resour. Res., 50,
2227–2241, 10.1002/2013WR014208, 2014.Iseya, F. and Ikeda, H.: Pulsations in Bedload Transport Rates Induced by a
Longitudinal Sediment Sorting: A Flume Study Using Sand and Gravel Mixtures,
Geogr. Ann. A, 69, 15–27, http://www.jstor.org/stable/521363, 1987.Jain, S.: Armor or Pavement, J. Hydraul. Eng.-ASCE, 116, 436–440,
10.1061/(ASCE)0733-9429(1990)116:3(436), 1990.
Klaassen, G. J.: Armoured river beds during flood, Tech. Rep. 394, Delft
Hydraulics, Emmeloord, the Netherlands, 1988.Kuhnle, R. A. and Southard, J. B.: Bed load transport fluctuations in a
gravel bed laboratory channel, Water Resour. Res., 24, 247–260,
10.1029/WR024i002p00247, 1988.Mao, L., Cooper, J. R., and Frostick, L. E.: Grain size and topographical
differences between static and mobile armour layers, Earth Surf. Proc. Land.,
36, 1321–1334, 10.1002/esp.2156, 2011.Marion, A. and Fraccarollo, L.: Experimental investigation of mobile armoring
development, Water Resour. Res., 33, 1447–1453, 10.1029/97WR00705,
1997.Orrú, C., Chavarrías, V., Uijttewaal, W. S. J., and Blom, A.: Image
analysis for measuring the size stratification in sand-gravel laboratory
experiments, Earth Surf. Dynam., 2, 217–232, 10.5194/esurf-2-217-2014,
2014.
Orrú, C., Blom, A., Chavarrías, V., Ferrara, V., and Stecca, G.: A
new technique for measuring the bed surface texture during flow and application to a
degradational sand-gravel laboratory experiment,
Water Resour. Res., in review, 2016.
Paola, C.: Subsidence and gravel transport in alluvial basins, in: New
perspectives in basin analysis, 231–243, Springer, 1988.Paola, C., Parker, G., Seal, R., Sinha, S. K., Southard, J. B., and Wilcock,
P. R.: Downstream Fining by Selective Deposition in a Laboratory Flume,
Science, 258, 1757–1760, 10.1126/science.258.5089.1757, 1992.Parker, G.: Selective Sorting and Abrasion of River Gravel – I: Theory, J.
Hydraul. Eng.-ASCE, 117, 131–147,
10.1061/(ASCE)0733-9429(1991)117:2(131), 1991a.Parker, G.: Selective Sorting and Abrasion of River Gravel – II:
Applications, J. Hydraul. Eng.-ASCE, 117, 150–171,
10.1061/(ASCE)0733-9429(1991)117:2(150), 1991b.Parker, G. and Cui, Y.: The arrested gravel front: stable gravel-sand
transitions in rivers – Part 1: Simplified analytical solution, J. Hydraul.
Eng.-ASCE, 36, 75–100, 10.1080/00221689809498379, 1998.Parker, G. and Klingeman, P.: On why gravel bed streams are paved, Water
Resour. Res., 18, 1409–1423, 10.1029/WR018i005p01409, 1982.
Pickup, G. and Warner, R.: Geomorphology of tropical rivers – II:
Channel adjustment to sediment load and discharge in the Fly and lower
Purari, Papua New Guinea, Catena Supplement, 5, 18–41, 1984.Piedra, M. M., Haynes, H., and Hoey, T. B.: The spatial distribution of
coarse surface grains and the stability of gravel river beds, Sedimentology,
59, 1014–1029, 10.1111/j.1365-3091.2011.01290.x, 2012.Recking, A., Frey, P., Paquier, A., and Belleudy, P.: An experimental
investigation of mechanisms involved in bed load sheet production and
migration, J. Geophys. Res.-Earth, 114, F03010, 10.1029/2008JF000990,
2009.
Sambrook Smith, G. H. and Ferguson, R. I.: The gravel-sand transition along
river channels, J. Sediment. Res., 65, 423–430, 1995.Seal, R., Paola, C., Parker, G., Southard, J., and Wilcock, P. R.:
Experiments on Downstream Fining of Gravel I: Narrow-Channel Runs, J.
Hydraul. Eng.-ASCE, 123, 874–884,
10.1061/(ASCE)0733-9429(1997)123:10(874), 1997.Shaw, J. and Kellerhals, R.: The Composition of Recent Alluvial Gravels in
Alberta River Beds, Bulletin (Alberta Research Council), Alberta Geological
Survey and Alberta Research Council, available at:
http://books.google.nl/books?id=ByPNSgAACAAJ, 1982.Sklar, L. S., Fadde, J., Venditti, J. G., Nelson, P., Wydzga, M. A., Cui, Y.,
and Dietrich, W. E.: Translation and dispersion of sediment pulses in flume
experiments simulating gravel augmentation below dams, Water Resour. Res.,
45, W08439, 10.1029/2008WR007346, 2009.
Spiller, S., Rüther, N., Koll, K., and Koll, K.: Bed load movement over a
fully developed armor layer–A tracer experiment, in: River Flow 2012, CRC
Press, 465–471, 2012.Vanoni, V. and Brooks, N.: Laboratory Studies of the Roughness and Suspended
Load of Alluvial Streams, MRD Sediment Series, California Institute of
Technology Sedimentation Laboratory, available at:
https://books.google.nl/books?id=sccBYAAACAAJ, 1957.Venditti, J. G. and Church, M.: Morphology and controls on the position of a
gravel-sand transition: Fraser River, British Columbia, J. Geophys.
Res.-Earth, 119, 1959–1976, 10.1002/2014JF003147, 2014.
Venditti, J. G., Dietrich, W. E., Nelson, P. A., Wydzga, M. A., Fadde, J.,
and Sklar, L. S.: Mobilization of coarse surface layers in gravel-bedded
rivers by finer gravel bed load, Water Resour. Res., 46, W07506,
10.1029/2009WR008329, 2010a.Venditti, J. G., Dietrich, W. E., Nelson, P. A., Wydzga, M. A., Fadde, J.,
and Sklar, L. S.: Effect of sediment pulse grain size on sediment transport
rates and bed mobility in gravel bed rivers, J. Geophys. Res.-Earth, 115,
F03039, 10.1029/2009JF001418, 2010b.Venditti, J. G., Domarad, N., Church, M., and Rennie, C. D.: The gravel-sand
transition: Sediment dynamics in a diffuse extension, J. Geophys. Res.-Earth,
120, 943–963, 10.1002/2014JF003328, 2015.Vericat, D., Batalla, R. J., and Garcia, C.: Breakup and reestablishment of
the armour layer in a large gravel-bed river below dams: The lower Ebro,
Geomorphology, 76, 122–136, 10.1016/j.geomorph.2005.10.005, 2006.Wang, T. and Liu, X.: The Breakup of Armor Layer in a Gravel-Bed Stream with
No Sediment Supply, in: Advances in Water Resources and Hydraulic
Engineering, Springer Berlin Heidelberg, 919–923,
10.1007/978-3-540-89465-0_161, 2009.Wilcock, P. R. and DeTemple, B. T.: Persistence of armor layers in gravel-bed
streams, Geophys. Res. Lett., 32, L08402, 10.1029/2004GL021772, 2005.Yager, E., Kenworthy, M., and Monsalve, A.: Taking the river inside:
Fundamental advances from laboratory experiments in measuring and
understanding bedload transport processes, Geomorphology, 244, 21–32,
10.1016/j.geomorph.2015.04.002, 2015.Yatsu, E.: On the longitudinal profile of the graded river, Eos, Transactions
American Geophysical Union, 36, 655–663, 10.1029/TR036i004p00655,
1955.