Journal of Coastal Research
SI 56
297 - 301
ICS2009 (Proceedings)
Portugal
ISSN 0749-0258
Instantaneous and Mean Aeolian Sediment Transport Rate on Beaches:
an Intercomparison of Measurements from Two Sensor Types.
R. Davidson-Arnottπ, B. O. Bauer§, I. J. Walker∞, P.A. Hesp†, J.Ollerhead‡ and I. Delgado-Fernandezπ
π Dept of Geography,
University of Guelph,
Guelph,
ON Canada N1G2W1
[email protected]
§Earth and Environmental
Sciences & Geography,
University of British
Columbia, Okanagan,
Kelowna,
BC Canada V1V 1V7
[email protected]
∞Dept of Geography,
University of Victoria,
P.O. Box 3060, Station
CSC, Victoria, BC
Canada V8W 3R4
[email protected]
†Dept of Geography and
Anthropology,
227 Howe Russell, LSU,
Baton Rouge, LA
70803, USA
[email protected]
‡Dept of Geography,
Mount Allison University,
144 Main St., Sackville,
NB Canada E4L 1A7
[email protected]
ABSTRACT
DAVIDSON-ARNOTT, R., BAUER, B.,O., WALKER, I.,J., HESP, P.,A., and OLLERHEAD, J., 2009. Instantaneous and
mean aeolian sediment transport rate on beaches: an intercomparison of measurements from several sensor types.
Journal of Coastal Research, SI 56 (Proceedings of the 10th International Coastal Symposium), 297 – 301.
Lisbon, Portugal, ISSN 0749-0258
Recently several new instruments, such as the Saltiphone, Sensit, Safire and laser sensors, have made it possible
to measure aeolian transport in the field at a frequency of 1 Hz, allowing us to evaluate the relationship between
varying wind speed and instantaneous transport. The correlation between the two variables at this frequency is
often very low and the exponent can range from <2 to >5. Since several of the instruments can be used for longterm monitoring of coastal dunes, it is important that we understand the causes of this poor correlation and the
relationship to averages derived from trap measurements. In this paper we compare measurements from Safire
piezo-electric sensors and Wenglor laser sensors under conditions of intermittent and continuous transport. The
laser sensor generally measures a higher rate of transport than does the Safire and has fewer periods of zero
transport (lower intermittency). This may reflect detection of relatively slow moving grains which may not have
sufficient momentum for the impact to register on the Safire. Nevertheless, calibration of Safire output averaged
over a period of 15-20 minutes against trap data results in high R2 values. The fit of a power curve to saltation
intensity regressed against instantaneous wind speed is usually stronger for the laser sensor compared to a Safire
but both show a wide range in the exponent of the power function.
ADDITIONAL INDEX WORDS: Intermittency.
INTRODUCTION
It has been possible to obtain high frequency (1 Hz or faster)
measurements of sediment transport by waves and currents in the
nearshore for several decades and thus to study the relationship
between the transport rate and fluctuations in the fluid flows.
Beginning in the mid 1970s high speed measurements of flow in
the nearshore could be made using electromagnetic current meters
and these were then coupled with measurements of suspended
sediment concentration at the same location. using the Optical
Backscatterance Sensor (OBS). Their widespread use led to
considerable evaluation of the two types of instrumentation and
thus a good understanding of principles for calibration and the
limitations of both (ref). Aeolian sediment transport measurements
on beaches and dunes have lagged behind those in the water. In
laboratory wind tunnels sand movement has been measured using
lasers (NICKLING, 1981) balance traps (BUTTERFIELD, 1991)
and recently by high speed cameras using particle tracking
velocimetry (ZHANG et al., 2007). While it has been possible to
make high frequency measurements of wind flow with cup
anemometers, and in the past two decades with 2-D and 3-D sonic
anemometers, no single instrument comparable to the OBS or
ADCP has been developed and adopted for field measurements of
high frequency aeolian sand transport. Instead, researchers have
employed a variety of different instruments that have for the most
part been designed in house, or produced on a very limited scale
commercially. This makes it difficult to compare results from
different field studies and may account for the some of the
discrepancies between field and laboratory modelling of the
relationship between bed shear velocity and sediment transport
rates.
Three principal types of sensor have been deployed in the field
to measure instantaneous aeolian sand transport: (1) instruments
which record the acoustic signal produced by the impact of
saltating grains on the screen of a miniature microphone (SPAAN
and VAN DEN ABELE, 1991; LEENDERS et al., 2005 ELLIS,
2009); (2) instruments which record the impact of saltating grains
on a ring connected to a piezo-electric crystal (STOCKTON and
GILLETTE 1990; STOUT and ZOBECK, 1996; WIGGS et al.,
2004; BAAS, 2004); and 3) instruments which record the weight
of sediment collected in a trap using some form of load cell or
electronic balance (JACKSON, 1996; DAVIDSON-ARNOTT et
al., 2005). The first two types of instrument record impacts over
an area that is only 1-2 cm in diameter or height, and thus
effectively sample only a portion of the vertical flux. They are
therefore equivalent to an OBS in that they provide a measure of
sediment concentration at a particular height above the bed and
Journal of Coastal Research, Special Issue 56, 2009
297
Evaluation of Sensors for Measuring Aeolian Sand Transport
Figure 1 Instrument deployment associated with one H frame,
October 21, 2007. Wind data are taken from the lowermost 3-D
sonic anemometer. The Safire is deployed just in front of the right
hand upright and the laser sensor is immediately to the left of the
Safire with most of it buried below the sand surface.
ideally the total transport must be obtained by deploying several
instruments in a vertical array. The third group may sample a
greater proportion of the total transport, for example the horizontal
trap designed by JACKSON (1996) or they may make use of a
vertical trap that samples all or most of the effective vertical
profile (DAVIDSON-ARNOTT et al., 2005).
In this preliminary paper we compare instataneous sediment
transport measured by Safire piezo-electric sensors to
measurements of wind speed at the same frequency as well as
comparing mean values of transport intensity over periods of 1520 minutes with mass transport rates over the same period
collected by integrating vertical traps. Finally, we compare the
output from a Safire sensor to a co-located Wenglor laser sensor to
explore the differences between the two instruments.
INSTRUMENTATION AND
EXPERIMENTAL DESIGN
Data discussed here are taken from two field experiments
carried out at Greewich Dunes, Prince Edward Island, Canada in
October 2004 and October 2007. The October 2004 experiment is
described in detail in DAVIDSON-ARNOTT and BAUER (2009),
BAUER et al. (2009), and WALKER et al. (2009). Winds were
obliquely onshore for a period of ten hours or more blowing over a
flat, gently sloping beach with a surface moisture content on the
upper beach of 2-4 %. Data presented here are from two safires
deployed on the upper beach with the sensor ring at 2-4 cm above
the bed. They were colocated with R.M. Young cup anemometers
deployed at a height of 0.25 m. The Safires were oriented with
their ‘sweet spot’ – BAAS, 2004) facing the wind. Transport was
generally fully developed in this zone and there was a slight
decrease in wind speed towards the top of the beach due to
internal boundary layer development (DAVIDSON-ARNOTT
and BAUER (2009), BAUER et al. (2009). In the second
experiment winds were again obliquely onshore but there was a
well-developed flat berm with steep foreshore slope. Windflow
was measured using 2-D and 3-D sonic anemometers mounted on
4 H frames spaced 10 m apart landward of the berm crest and
aligned along the wind direction (Figure 1). Wind speed
measurements are taken from the lowermost 3-D sonic
anemometer centred at a height of 0.25 m. The Safire was again
mounted at a height of 2 cm with its sweet spot oriented in to the
wind. The laser sensor was deployed immediately adjacent to the
Safire.
The laser sensor is a commercial unit made by Wenglor Co.
Ltd., and used primarily for sensing objects on conveyor belts.
The instrument consists of a laser and photo sensor mounted
within a U shaped frame with a spacing (path length) of 3 cm
and a beam diameter of 1 mm. The instruments detects the drop
in voltage at the photo sensor resulting from the passage of
individual grains through the beam. The counting circuitry is
contained within the instrument and is capable of detecting 700800 grains per second. Data were recorded using an Onset
counting circuit and Onset Hobbo data logger. The Safire is made
by Sabajo in the Netherlands and records the impact of saltating
grains on a 2 cm high ring. The electronics are housed within the
2 cm diameter tube on which the ring is mounted and in this
deployment the voltage output was recorded with a nominal 200
impact per second corresponding to the maximum 5 volts output.
The instrument has been described in detail by BAAS (2004).
The instruments deployed here were later versions with
somewhat better sensitivity and reduced variation in output
around the ring than those described by BAAS (2004). In both
experiments cross calibration at the site of all sensors was carried
out at the end of the day.
RESPONSE OF SAFIRE PROBES
A major concern with the deployment of probes such as the
Safire piezo-electric sensors is the similarity of response between
instruments, especially given earlier tests which have shown
considerable variation in sensitivity around the circumference of
the probe (BAAS, 2004). In short-term experiments these can be
deployed so that they are always oriented with the same spot into
the wind which should reduce this effect, but there is still concern
that the sensitivity varies considerably from one instrument to
another, making it difficult to compare the response of instruments
deployed at different locations. Figure 2 shows results from a
cross calibration of three saltation probes carried out at the end of
the day of the field experiment described in DAVIDSONARNOTT and BAUER (2009) and BAUER et al. (2009). The
probes were deployed 0.1 m apart at the back of the beach and
allowed to record for one hour. There was a close correspondence
in the response of all three probes to fluctuations in wind speed
over this period (Figure 2a) and the mean saltation intensity for
ten minute intervals during the cross calibration period were very
simuilar (Figure 2b). Some deviation is expected between
instruments even with this close spacing (BAAS and SHERMAN,
2006). The differences between means is greater for higher mean
transport intensity at the beginning and end of the calibration
period, but it is notable that probe 8 had the highest mean value at
the beginning of the period and the lowest at the end of the period
so that there is no consistent difference in the response of the
probes. The results of the cross calibration showed that, for the
conditions under which the field experiment was carried out, the
response of the three probes are indistinuishable.
Journal of Coastal Research, Special Issue 56, 2009
298
Davidson-Arnott et al.
(a)
(a)
(b)
(b)
Figure 2. Comparison of the response of three saltation probes
during cross calibration in the field. The probes were spaced
0.1 m apart and transport recorded for one hour: a) plot of
saltation intensity over a 5 minute period for the three probes;
b) mean saltation intensity recorded over six 10 minute
intervals. Note that intensity here is recorded as the raw
voltage output – 5 volts corresponds nominally to 200 counts
per second
RELATIONSHIP BETWEEN SALTATION INTENSITY
AND U3
A key reason for the deployment of Safires is to examine the
relationship between instantaneous sand transport and wind speed
or bed shear velocity. Wind tunnel studies, as well as some field
experiments where transport is fully developed, show that
transport varies as a function of U*3 or U3. A number of studies
have shown a general relationship between wind speed
fluctuations and the saltation intensity recorded by piezo-electric
sensors such as the Sensit and Safire (WIGGS et al., 2004;
DAVIDSON-ARNOTT et al., 2008; DAVIDSON-ARNOTTand
BAUER, 2009) but much work still remains to be done to
determine the correspondence between mean values determined
from measurements over ten or fifteen minutes and measurements
made at a frequency of 1 Hz..
Figure 3 illustrates the general relationship between mean
saltation intensity and the cube of mean wind speed measured at
two locations on the upper beach for ten minute intervals in
October 2004 (DAVIDSON-ARNOTT and BAUER, 2009). The
relationship for both probes is significant at the 0.05 level and the
relationship appears to hold for conditions where transport is fully
developed over a substantial fetch distance. With a shorter fetch
transport is much more intermittent and the strength of the
relationship decreases rapidly.
Figure. 3. Comparison of mean saltation intensity measured by
saltation probes and the cube of mean wind speed measured by a
co-located cup anemometer for ten minute intervals taken from
continuous measurements over a period of 280 minutes: a) probe
salt 7 located at the back of the beach; b) probe 10, located about
6 m seaward of probe 7. The greater scatter for probe 10 may
reflect a much reduced fetch near the end of the recording period
and consequently full transport conditions may not have been
achieved.
RELATIONSHIP BETWEEN SALTATION INTENSITY
AND SAND TRANSPORT RATE
Rates of sand transport were measured on 5 occasions during
the field experiment using integrating vertical traps co-located
with the saltation probes, thus permitting a comparison of the
mean transport intensity measured by the saltation probes with the
mean transport rate measured by the traps (Figure 4). There is a
strong and significant relationship between the two measures of
sand transport which again serves to indicate that the mean
transport intensity measured by the probes is an adequate
representation of mean transport rates on the beach. The
calibration however is only valid for the particular bed conditions
under which the measurements were made.
Journal of Coastal Research, Special Issue 56, 2009
299
Evaluation of Sensors for Measuring Aeolian Sand Transport
counts per second) registered by the laser probe was about 3 times
higher than that for the Safire even though the nominal sensening
area is smaller, and this was consistent throughout the record. One
result of this was that as the transport rate dropped the Safire
registered many periods with zero trransport (high intermittency)
while the laser sensor registered continuous transport.
a)
Figure 4. Linear regression of mean sand transport intensity
against the sand transport rate measured by a co-located vertical
trap over the same period.
RELATIONSHIP BETWEEN INSTANTANEOUS
SALTATION INTENSITY AND WIND SPEED
While the relationship between saltation intensity and wind
speed appears to be quite robust for averages taken over ten
minutes or more, the instantaneous values show considerable
b)
c)
Figure 5. Plot of all non-zero values of instantaneous saltation
intensity measured by saltation probe 7 for a ten minute period.
scatter and are often not statistically significant, with R2 values
generally below 0.2. Where sand transport is fully developed, the
exponent for a power relationship ranges from <2 to > 4 (Figure
5). The relationship may be improved somewhat with a lag of one
second between saltaion intensity compared to wind speed but
decreases rapidly with a further increase in the lag.
COMPARISON OF SALTATION PROBE WITH COLOCATED LASER SENSOR
In the field experiment carried out on October 21, 2007
Wenglor laser sensors were deployed adjacent to Safires with
wind speed at 0.25 m measured by 3-D sonic anemmometers
(Figure 1). Sand transport was measured over a period of about 4
hours at a sampling frequency of 1 Hz. A two-minute portion of
the record from Station 3 located about 10 m landward of the berm
crest is shown in Figure 6a. In general the saltation intensity (grain
Figure 6. Comparison of saltation intensity measured by colocated Safire and laser probes: a) two-minute record showing
fluctuations in saltation intensity measured by the two
instruments and wind speed measured by the sonic anemometer
at 0.25 m; b) and c) relationship between instantaneous
measurements of saltation intensity and wind speed for the
Safire and laser probes respectively.
There was also a much weaker relationship between between
wind speed and instantaneous transport for the Safire as compared
to the laser sensor. This is illustrated in Figure 6 b, c for the 2
minute record shown in Figure 6a.
Journal of Coastal Research, Special Issue 56, 2009
300
Davidson-Arnott et al.
DISCUSSION
Our experience in the deployment of Safire piezo-electric
probes is that they can provide a useful indication of the presence
of sand transport and a qualitative indication of the relative
strength of sand transport intensity. With careful laboratory testing
to determine the location of the sector with the most sconsistent
response on the circumference of the probe, they can be deployed
for short-term experiments with that sector aligned into the wind.
Cross calibration in the field is essential if comparisons are to be
made between probes deployed at different locations. The strong
relationships found between mean saltation intensity measured by
the Safires and mean wind speed measured over ten minutes
indicates that they provide a robust measure of relative sand
ttransport. Likewise, the strong relationship between transport
intensity and total sand transport rate measured by integrating
vertical traps suggests that with this form of field calibration the
Safires could be used to provide a measure of the actual transport
rate rather than simple transport intensity. However, since the
shape of the vertical concentration profile is likely to vary with
different surfaces and different grain size distributions, field
calibration is always necessary.
The lower counts registered by the Safire probe during the
exoperiment in October 2007 may reflect in part the presence of a
loose dry surface at this location and thus likely a greater
proportion of low velocity (reptating) grains compared to the
relatively dam, hard surface associated with the 2004 experiment.
It is likely that many of these grains did not have sufficient
momentum to be registed on the piezo-electric sensor but they
would be detected by the laser sensor. The poor correspondence
between instantaneous transport and wind speed for the Safire
probes suggests that their sampling volume is too small to
adequately represent total transport at a frequency of 1 Hz though
they provide a good measure of transport at a lower frequency. It
also calls into question their use for calculating a transport
threshold using the time fraction equivalence method proposed by
STOUT and ZOBECK (1996).
LITERATURE CITED
BAAS A.C.W. 2004. Evaluation of saltation flux impact
responders (Safires) for measuring instantaneous aeolian sand
transport intensity. Geomorphology, 59, 99-118.
BAAS, A.C.W., SHERMAN D.J. 2005b. Spatiotemporal variability
of aeolian sand transport in a coastal dune environment.
Journal of Coastal Research, 22, 1198-1205.
BAUER, B.O., DAVIDSON-ARNOTT, R.G.D, HESP, P.A., NAMIKAS,
S.L., OLLERHEAD, J. and WALKER, I.J., 2009. Aeolian
sediment transport conditions on a beach: Surface moisture,
wind fetch, and mean transport rates. Geomorphology 105(12), 106-116.
BUTTERFIELD, G.R., 1991. Grain transport rates in steady and
unsteady turbulent airflows. Acta Mechanica (Suppl), 1, 97122.
DAVIDSON-ARNOTT, R.G.D and BAUER, B.O., 2009. Aeolian
sediment transport on a beach: Thresholds, intermittency and
high frequency variability. Geomorphology, 105, 117-126.
DAVIDSON-ARNOTT, R.G.D, MCQUARRIE, K, AND AAGAARD, T.,
2005. The effects of wind gusts on aeolian sediment transport
on a beach. Geomorphology, 68, 115-129
DAVIDSON-ARNOTT, R.G.D., YANG, Y., OLLERHEAD, J., HESP,
P.A., WALKER, I.J., 2008. The effects of surface moisture on
aeolian sediment transport threshold & mass flux on a beach.
Earth Surface Processes and Landforms, 33, 55-74.
ELLIS, J.T., MORRISON, R..F. AND PRIESTA, B.H., 2009.
Detecting impacts of sand grains with a microphone system in
field conditions. Geomorphology, 105, 87-94.
JACKSON, D.W.T., 1996. A new, instantaneous aeolian sand trap
design for field use. Sedimentology, 43, 791-796.
LEENDERS, J.K., VAN BOXEL, J.H. and STERK, G., 2005. Wind
forces and related saltation transport. Geomorphology, 71,
357-372.
NICKLING, W.G., 1988. The initiation of particle movement by
wind. Sedimentology, 35, 499-511.
SPAAN, W.P. and VAN DEN ABELE, G.D., 1991. Wind borne
particle measurements with acoustic sensors. Soil Technology,
4, 51-63.
STOCKTON, P. and GILLETTE, D.A., 1990. Field measurement of
the sheltering effect of vegetation on erodible land surfaces.
Land Degradation and Rehabilitation, 2, 77-95.
STOUT, J.E. and ZOBECK, T.M., 1997. Intermittent saltation.
Sedimentology, 44, 959-970.
WALKER, I.J., HESP, P.A., DAVIDSON-ARNOTT, R.G.D., and
OLLERHEAD, J., 2009. Responses of three-dimensional flow to
variations in the angle of incident wind & profile form of
dunes: Greenwich Dunes, PEI, Canada. Geomorphology
105(1-2), 127-138.
WIGGS, G.F.S., BAIRD, A.J. and ATHERTON, R.J., 2004.
Thresholds of aeolian sand transport: establishing suitable
values. Sedimentology, 51, 95-108.
ZHANG, W., WANG, Y. and LEE, S., 2007. Two-phase
measurements of wind and saltating sand in an atmospheric
boundary layer. Geomorphology, 88, 109-119.
ACKNOWLEDGEMENTS
Thanks to NSERC and the LSU Faculty Research Grant for
financial support, and to PEI National Park for giving us
permission to carry out the work. We especially thank Park
Ecologists Denyse Lajeunesse and Kirby Tulk for their assistance.
Journal of Coastal Research, Special Issue 56, 2009
301