126
Erosion and Sediment Transport Measurement in Rivers: Technological and Methodological
(Proceedings of Hie Oslo Workshop. June 2002). IA1 IS l'ubl. 283. 2003.
Advance.
Sampler size and sampling time affect bed load
transport rates and particle sizes measured with
bed load traps in gravel-bed streams
KRISTIN B U N T E & STEVEN R. A B T
Engineering
Research
Center,
Colorado
State
University,
Fort
Collins,
Colorado
80523,
USA
[email protected]
Abstract A bed load trap was designed for collecting gravel and cobble bed
load and tested in four mountain gravel-bed streams. The traps permit a long
sampling duration of about 1 h and have a large 0.3 x 0.2 m opening as well as
a large sampler bag. Samples collected with the bed load traps have welldefined power-function bed load rating curves with exponents of 8-16.
Compared to a Helley-Smith sampler with a 7.6 x 7.6 cm opening, bed load
transport measured with the traps is smaller during low flows and larger
during high flows. Bed load particle-size distributions obtained with the traps
coarsen more pronouncedly with flow than those obtained with a HelleySmith sampler. These differences in sampled bed load transport rates and size
distributions have implications for computations such as critical discharge,
annual load, and effective discharge.
Key words
b e d load sampling; bed load traps; flow c o m p e t e n c e ; gravel b e d load transport;
initial m o t i o n ; rating c u r v e ; R o c k y M o u n t a i n g r a v e l - b e d rivers; s a m p l i n g t i m e
INTRODUCTION
The relationship between gravel bed load transport rates and discharge (or any other
measure of flow) varies widely between streams. This is particularly true in mountain
streams and may be attributed to the stream-specific arrangements of bed-material par­
ticles in gravel- and cobble beds as well as site-specific conditions of sediment supply.
Similarly, critical discharge (or any other measure of critical flow) for incipient motion
of a given particle-size class varies among streams and is poorly predictable. There­
fore, if site-specific information is needed for a given mountain stream, measurements
of bed load transport seem to have the best potential for accurate results. The sampler
employed for measuring bed load in a given mountain stream should not only be port­
able but also specifically designed for sampling gravel and cobble particles. Automat­
ically recording portable hydrophones and similar devices (e.g. Bogen & Moen, 2003;
Downing et al, 2003; Froehlich, 2003; Gottesfeld & Tunnicliff, 2003; Habersack,
2003; Laronne et al, 2003; Mizuyama et al, 2003; Sear, 2003; Sear et al, 2003)
provide useful temporal records of the onset and cessation of motion and relative
transport rates. However, the exact particle-size distribution in motion or the exact
transport rate is usually unknown. To obtain this information, one still needs a sampler
that, for given flows, provides bed load samples that can be weighed and sieved.
The transport frequency of gravel and cobble particles of the size class that is just
beginning to move is very low; exactly how low is a matter of observer definition (or
patience), e.g. one particle per metre width per hour. For accurate samples of gravel
Sampler
size and sampling
time affect bed load transport
rates and particle
sizes
127
and cobble transport rates at the threshold of motion, the sampling duration must be
equal to or larger than the presumed transport frequency of the largest mobile particle
size to be collected. Other prerequisites for accurate sampling of gravel and cobble bed
load are a sampler opening large enough for cobbles to enter, while the sampler bag
must have a sufficiently fine mesh to retain small gravel particles. The bag size must
be large enough to hold the large quantity of smaller bed load particles that accompany
cobble transport at the onset of motion. Bed load samplers with those criteria were not
available. We therefore developed a bed load trap that has these attributes (Fig. 1).
Fig. 1 Bed load trap for collecting gravel and cobble bed load.
Fig. 2 (a) Bed load traps with ground plates at East St. Louis Creek in low flow.
(b)Emptying bed load traps at Little Granite Creek in bankfull flow. The open squares
indicate the positions of 4 of the 6 bed load traps which are completely submerged by
flow, (c) Using a 7.6 * 7.6 cm opening Helley-Smith sampler at Little Granite Creek
near bankfull flow.
128
Kristin
Bun te & Steven R. Abt
METHODS
The bed load traps (Fig. 1) consist of a 0.3 x 0.2 m aluminium frame onto which a
0.9 m long net with a 3.9 m m mesh width is attached. For the usual 1-h duration of
each sample, the frame is fastened onto a ground plate anchored on the stream bottom
(Fig. 2(a)). This arrangement frees the operator from holding the traps and avoids
unwanted particle pick-up during trap placement or retrieval from the bed surface. Sev­
eral bed load traps are deployed across the stream in 1-2 m increments (Fig. 2(b)). Bed
load was sampled with these traps in four mountain gravel-bed streams with plane-bed
or step-pool morphology at wadable flows (Bunte et al, 2 0 0 1 , 2003; Table 1).
Although not designed for coarse gravel and cobble bed load, a Helley-Smith
sampler (7.6 x 7.6 cm opening) (Helley & Smith 1971) is often used in mountain
streams, because this portable and readily available sampler is easy to use. In order to
compare the sampling results of the bed load traps with those of the Helley-Smith
sampler, bed load was also collected with a Helley-Smith sampler at each of the four
streams (Fig. 2(c)), using a standard sampling scheme (e.g. 2 minutes per location
spaced at 0.3-1 m increments across the stream). These data were pooled with
extensive Helley-Smith data sets collected by others (Ryan & Emmett 2002; Water
Resources Team of the Winema National Forest, Oregon; Ryan et al., 2002).
S A M P L I N G INTENSITY
Because bed load traps and Helley-Smith samplers collect over different time spans
and different fractions of the total width, we defined a relative sampling intensity /, in
order to quantify these differences:
w
act
'tot
where w and n are the width and number of traps (or sampling locations for the
Helley-Smith sampler) per cross section, At is the sampling time, w
is the active
stream width, and t, , is the total time increment allotted to one sample. We typically
used a sampling time of 1 h for the bed load traps in order to provide a (sufficiently)
long duration for the collection of infrequently moving (large) particle sizes near the
threshold of motion. A 1-h sampling time also matches the typical time required to
s
s
ucl
0
Table 1 Characteristics of the four streams sampled.
Stream
Drainage
Area
(km )
(nrV)
Wf
kk
Dso
14.3
9.5
6.5
3.7
69
49
53
108
2
Little Granite Cr.
55
5.7
Cherry Cr.
41
3.1
St. Louis Cr.
34
4.0
East St. Louis Cr.
8
0.76
* Rosgen ( 1994).
t Montgomery & Buffmgton (1997).
Surface
A»
(mm)
(m)
141
152
120
258
Stream
gradient
(-)
0.017
0.025
0.017
0.093
Stream type*,
morphology f
Range of
measmts
B4,
B4,
B4,
A3,
65-135
49-145
28-65
26-71
(% Qnr)
plane-bed
plane-bed
plane-bed
step-pool
Sampler
size and sampling
time affect bed load transport
rates and particle
sizes
129
collect a bed load sample with a Helley-Smith sampler. Sampling times longer than 1 h
may fail to have near-constant discharge during snowmelt highflows and thus prohibit
allocation of a bed load sample to a discharge value. W e typically deployed 4 to 6 bed
load traps across streams 4 - 1 2 m wide, resulting in sampling intensities of 1 5 - 3 0 %
(equation 1). Sampling the entire stream width continuously over the entire time period
allotted to one sample equals 100%. By contrast, sampling intensities for the HelleySmith sampler only attained 0.4-0.5%; a difference of a factor of about 50.
RESULTS
Steep and well-defined rating curves
The long sampling duration facilitated by temporarily mounting the bed load traps onto
ground plates permits collecting very small samples when transport rates are low. The
smallest possible sample that is obtained by collecting one 4-mm particle per hour per
trap equals a unit transport rate of 0.0002 g m" s" . However, the large opening size of
the bed load traps and the large bag size also permit the collection of very large sam­
ples during high transport rates. For example, the largest sample obtained (at Little
Granite Cr.) comprised 20 kg of gravel and cobbles collected over 6 min and had a unit
transport rate of 185 g m" s" . This range spans 6 orders of magnitude between the
lowest and the highest measured flows. Transport rates measured with the bed load
traps generally increased steeply with discharge and the data scatter was comparatively
low. Power function rating curves fitted to the bed load trap data were well-defined
(0.76 < r < 0.91) for all four streams and had high exponents between 8 and 16 (Table
2 and Fig. 3). These values are similar to rating curve exponents of 5-20 observed by
other researchers who measured gravel bed load transport using devices with long
sampling times and large sampler openings (e.g. Hassan & Church, 2 0 0 1 ; Wilcock,
2001; Bunte, 1996). By contrast, power function rating curves fitted to the HelleySmith data were less well defined (0.35 < r < 0.59) and had exponents ranging
between 1.7 and 3.6.
1
1
1
1
2
Comparison with Helley-Smith sampler
Detailed gravel transport rates obtained by the bed load traps and the Helley-Smith
sampler are shown for comparison in Fig. 4. Results obtained at East St. Louis Creek
2
Table 2 Exponents (Q = a-Q *) and coefficients of determination (r ) for bed load rating- and flow
competence curves from the bed load traps and the Helley-Smith sampler (HS).
b
Stream
Bed load transport rating curves
i
Exponent
r~
Traps
HS
Traps
HS
Flow competence curves
Exponent
r~
Traps
Traps
HS
HS
Little Granite Cr
Cherry Cr.
St. Louis Cr.
East St. L. Cr.
16.2
11.5
10.8
8.4
3.5
2.7
2.4
1.5
0.46
0.57
0.46
0.27
3.1
1.7
3.2
3.6
0.76
0.90
0.49
0.78
0.44
0.35
0.59
0.54
0.88
0.67
1.03
1.01
0.67
0.90
0.43
0.40
Kristin
Bunte
& Steven R. Abt
1000
3
Discharge (m /s)
Fig. 3 Bed load rating curves obtained from the bed load traps for all four streams
sampled. The respective rating curves for the Helley-Smith sampler are shown for
comparison as dashed lines keyed to the symbol for each site. Dotted lines extend bed
load trap rating curves to bankfull flow.
East St. Louis Creek
Helley-Smith:
o this
study
- Ryan's
Bedload
samples
traps
1E-4
0.1
3
Discharge (m /s)
Fig. 4 Bed load transport rates and rating curves obtained from samples collected with
bed load traps and a 3-inch Helley-Smith sampler at East St. Louis Creek. The greyshaded area represents the envelope of bed load transport rates computed from HelleySmith samples. Dotted line extends bed load trap rating curve to bankfull flow.
Sampler
size and sampling
time affect bed load transport
rates and particle
131
sizes
are typical of results obtained at all four streams. Gravel transport rates from the bed
load traps were lower and had less data scatter during moderate high flows than gravel
transport rates measured with a Helley-Smith sampler. The gap between bed load
rating curves of the traps and the Helley-Smith sampler reached 1-4 orders of magni­
tude at 5 0 % of bankfull flow. When high flows exceeded approximately bankfull,
gravel transport rates collected with the bed load traps were higher than those from the
Helley-Smith sampler. Thus, bed load rating curves were generally steeper for the bed
load traps than for the Helley-Smith sampler and crossed near bankfull flow (Table 2
and Fig. 3). These differences are attributed to the long, 1-h sampling period which,
during infrequent particle motion, facilitates sampling much smaller transport rates for
the largest mobile size classes than the 1-2 min sampling time typical of the HelleySmith. The large opening of the trap allows large gravel and cobble particles to easily
enter the sampler, whereas these particles sizes may not contribute to the sample when
using a sampler with a 7.6 x 7.6 cm opening. In addition, the large bag of the traps also
allows sampling for a relatively long duration at high transport rates when cobbles start
to become mobile.
The differences in sampler dimensions and sampling time produce not only
different transport rates, but also different bed load particle-size distributions. Particlesize percentiles of samples collected with the bed load trap coarsened pronouncedly
with increasing discharge (Fig. 5). By contrast, the Helley-Smith samples contain more
coarse particles at very low flows, while at high flows, Helley-Smith samples remain
finer. The bed load D
particle sizes from the trap samples likewise increase steeply
with flow. The flow competence curves fitted to data from all four streams have larger
exponents for the traps than for the Helley-Smith (Fig. 6 and Table 2).
max
100
Little Granite
0
2
Creek
4
QWQ
8
10
3
Discharge class (m /s)
Fig. 5 Increase of bed load particle-size percentiles with grouped values of discharge
for samples obtained from the bed load traps (thick lines) and the Helley-Smith
sampler (thin lines).
132
Kristin Bun te & Steven R. Abt
1
r-
0.1
f-
••[
•!
t
\
I
t • I -f
V - - - I
1
\
[-- -!
-I
I
IH
10
3
Discharge (m /s)
Fig. 6 Flow competence curves determined from samples collected with the bed load
traps for all four streams sampled. The respective flow competence curves for the
Helley-Smith sampler are shown for comparison as dashed lines keyed to the symbol
for each site. Dotted lines extend flow competence curves up to bankfull flow.
IMPLICATIONS
The sampler-dependent differences in sampled transport rates and particle-size
distributions have implications for all computations that are based on measured bed
load transport rates. Critical discharge for gravel mobility is higher when computed
based on bed load trap samples compared to computations based on Helley-Smith
samples. Annual bed load discharge computed from bed load trap samples is smaller
for years of low flow and higher for years with high flows than annual bed load
discharge computed from Helley-Smith samples. Effective discharge, which in a
magnitude-frequency analysis determines the flow that transports the majority of bed
load in a given stream over the long run, is strongly affected by the rating curve
steepness. In streams with snowmelt regimes, effective discharge seems to be in the
neighbourhood of bankfull flow when computed from bed load rating curves with
exponents around 3 but shifts to progressively higher flows as exponents of the bed
load rating curves increase. This study has shown that sampling duration and sampler
dimensions greatly affect sampled transport rates and bed load particle-size
distributions. Consequently, other computations, such as incipient motion, annual load,
and effective discharge will also be influenced by these factors of equipment and
technique. The implications of this alternative way of looking at bed load data raise
fundamental questions that challenge our basic understanding of fluvial systems.
Sampler
size and sampling
time affect bed load transport
rates and particle
133
sizes
Acknowledgements Sandra Ryan and Bill Emmett kindly provided most of the
Helley-Smith data. Kurt Swingle, Sean McCoy, Paul Bakke and others helped with the
field work. John Potyondy provided support in all stages of the work. The project was
funded by the Stream Systems Technology Center of the U S D A Forest Service, Rocky
Mountain Research Station, Fort Collins, Colorado, USA.
REFERENCES
B o g e n , J. &
Moen,
Transport
K. ( 2 0 0 3 )
Measurement
Bedload
measurements
in Rivers:
with
Technological
and
a n e w passive acoustic sensor.
Methodological
Advances
I n : Erosion
and
Sediment
( e d . b y J. B o g e n , T . F e r g u s
D . E . W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
&
issue).
B u n t e , K . ( 1 9 9 6 ) A n a l y s e s o f t h e t e m p o r a l v a r i a t i o n o f c o a r s e b e d l o a d t r a n s p o r t a n d its g r a i n s i z e d i s t r i b u t i o n ( S q u a w
Creek, Montana, U S A ) . English translation of P h D dissertation, Freie University Berlin, G e r m a n y . U S D A Forest
Service, R o c k y M o u n t a i n Forest and R a n g e E x p e r i m e n t Station, G e n . Tech. Report
RM-GTR-288.
B u n t e , K . , A b t , S. R . & P o t y o n d y , J. P . ( 2 0 0 1 ) P o r t a b l e b e d l o a d t r a p s w i t h h i g h s a m p l i n g i n t e n s i t y for r e p r e s e n t a t i v e
s a m p l i n g o f g r a v e l t r a n s p o r t in w a d a b l e m o u n t a i n s t r e a m s . I n : Interagency
Sedimentation
(Proc. Seventh Conf,
R e n o , N e v a d a , U S A ) , vol. 2, III.24-III.31.
B u n t e , K . , A b t , S . R . , P o l y o n d y , .1. P . & R y a n , S . E . ( 2 0 0 3 ) M e a s u r e m e n t o f c o a r s e g r a v e l a n d c o b b l e t r a n s p o r t u s i n g a
p o r t a b l e b e d l o a d t r a p . , / . Hydraul.
Engineer
D o w n i n g , J . , F a r l e y , P . J . , B u n t e , K.. et al.
Measurement
in Rivers:
Technological
(submitted)
( 2 0 0 3 ) A c o u s t i c g r a v e l - t r a n s p o r t s e n s o r . I n : Erosion
and Sediment
Transport
and Methodological
Advances
(ed. b y J. B o g e n , T . F e r g u s & D . E. W a l l i n g )
( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
issue).
F r o e h l i c h , W . ( 2 0 0 4 ) M o n i t o r i n g o f b e d l o a d t r a n s p o r t in s t r e a m s b y u s e . I n : Erosion
and
Sediment
Measurement
in Rivers:
Technological
and Methodological
Advances
( e d . b y J. B o g e n , T . F e r g u s & D . E .
( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
issue).
Transport
Walling)
G o t t e s f e l d , A . S . & T u n n i c l i f f , .1. ( 2 0 0 3 ) B e d l o a d m e a s u r e m e n t s w i t h a p a s s i v e m a g n e t i c i n d u c t i o n d e v i c e . I n :
and
Sediment
Transport
Measurement
in Rivers:
Technological
and
Methodological
Advances
F e r g u s & D . E . W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
Erosion
( e d . b y J. B o g e n , T .
issue).
H a b e r s a c k , H . M . ( 2 0 0 3 ) U s e o f r a d i o - t r a c k i n g t e c h n i q u e s in b e d l o a d t r a n s p o r t i n v e s t i g a t i o n s . I n : Erosion
and
Sediment
Transport
Measurement
in Rivers:
Technological
and Methodological
Advances
( e d . b y J. B o g e n , T . F e r g u s & D . E.
W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
issue).
H a s s a n , M . A . & C h u r c h , M . ( 2 0 0 1 ) S e n s i t i v i t y o f b e d l o a d t r a n s p o r t in H a r r i s C r e e k : S e a s o n a l a n d s p a t i a l v a r i a t i o n o v e r a
c o b b l e - g r a v e l b a r . Water
Résout:
Res.
37(3), 8 1 3 - 8 2 5 .
H e l l e y , E . J. & S m i t h , W . ( 1 9 7 1 ) D e v e l o p m e n t a n d c a l i b r a t i o n o f a p r e s s u r e - d i f f e r e n c e b e d l o a d s a m p l e r . U S G e o l . S u r v e y
Water Resource Division, Open-File Report, M e n l o Park, California,
L a r o n n e , J . B . , A l e x a n d r o v , Y . , B e r g m a n , N . et al.
environments.
Advances
I n : Erosion
and
Sediment
USA.
( 2 0 0 3 ) T h e c o n t i n u o u s m o n i t o r i n g o f b e d l o a d flux in v a r i o u s
Transport
Measurement
in
Rivers:
Technological
and
fluvial
Methodological
(ed. b y J. B o g e n , T. F e r g u s & D . E. W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S
Publ.
2 8 3 {this
issue).
M i z u y a m a , T . , F u j i t a , M . & N o n a k a , M . ( 2 0 0 3 ) M e a s u r e m e n t o f b e d l o a d w i t h t h e u s e o f h y d r o p h o n e in m o u n t a i n t o r r e n t s .
I n : Erosion
and Sediment
Transport
Measurement
in Rivers:
Technological
and Methodological
Advances
( e d . b y .1.
B o g e n , T . F e r g u s & D . E . W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
issue).
M o n t g o m e r y , D . R . & B u f f m g t o n , J . M . ( 1 9 9 7 ) C h a n n e l - r e a c h m o r p h o l o g y i n m o u n t a i n d r a i n a g e b a s i n s . Geol.
Bull. 1 0 9 ( 5 ) , 5 9 6 - 6 1 1 .
R y a n , S . E . & E m m e t t , W . W . ( 2 0 0 2 ) T h e n a t u r e o f f l o w a n d s e d i m e n t m o v e m e n t in L i t t l e G r a n i t e C r e e k n e a r
W Y . U S D A Forest Service, R o c k y Mountain Research Station. Gen. Tech. Report
Soc.
Bondurant,
RMRS-GTR-90.
R y a n , S. E . , P o r t h , L. S. & T r o e n d l e , C . A . ( 2 0 0 2 ) D e f i n i n g p h a s e s o f b e d l o a d t r a n s p o r t u s i n g p i e c e w i s e r e g r e s s i o n .
Surf.
Processes
Landf
Event
bedload
h y d r o g r a p h s h a p e . I n : Erosion
Advances
yield
and
measured with
Sediment
Transport
169-199.
load cell b e d l o a d
Measurement
traps and
in Rivers:
prediction
Technological
of bedload
and
yield
Transport
for e s t i m a t i n g c o a r s e b e d l o a d t r a n s p o r t
Measurement
in Rivers:
Technological
in g r a v e l - b e d d e d
and
Methodological
s t r e a m s u s i n g t r a c e r s . I n : Erosion
Advances
W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S P u b l . 2 8 3 (this
and
Landf
26, 1 3 9 5 - 1 4 0 8 .
Method
Sediment
( e d . b y J. B o g e n , T . F e r g u s & D . E .
issue).
W i l c o c k , P . R . ( 2 0 0 1 ) T o w a r d a p r a c t i c a l m e t h o d f o r e s t i m a t i n g s e d i m e n t t r a n s p o r t r a t e s in g r a v e l - b e d r i v e r s . Earth
Processes
Publ.
issue).
S e a r , D . A . , L e e , M . W . E . , C a r l i n g , R . J . et al. ( 2 0 0 3 ) A n a s s e s s m e n t o f t h e a c c u r a c y o f t h e S p a t i a l I n t e g r a t i o n
(SIM)
from
Methodological
( e d . b y .1. B o g e n , T . F e r g u s & D . E . W a l l i n g ) ( P r o c . W o r k s h o p , O s l o , N o r w a y , J u n e 2 0 0 2 ) . I A H S
2 8 3 (this
Earth
27(9), 9 7 1 - 9 9 0 .
R o s g e n , D . L . ( 1 9 9 4 ) A c l a s s i f i c a t i o n o f n a t u r a l r i v e r s . Catenall,
Sear, D. A . (2003)
Am.
Surf