Geoscience Reference
In-Depth Information
controlled conditions of the wind tunnel and involve the
observation of the onset of sediment motion with in-
creasing wind velocity (see, for example, Bagnold, 1941;
Kawamura, 1951; Zingg, 1953). While such studies have
helped in the development of relationships such as that
shown in Figure 18.11 they have not been able to greatly
improve our knowledge of thresholds in natural field en-
vironments where the identification of the onset of sedi-
ment ransport is complicated by a constantly varying wind
velocity.
However, the development of electronic grain impact
sensors such as the Sensit (Stockton and Gillette, 1990)
and the Safire (Baas, 2004) have enabled us to inves-
tigate entrainment thresholds in the field more fully. A
typical field data series of wind velocity and sediment
transport occurrence is shown in Figure 18.12 and demon-
strates the difficulties inherent in identifying one single
entrainment threshold. While a threshold value could
be considered as the wind velocity at which the first
grains are seen to be entrained into transport (A in Fig-
ure 18.12), it is also clear that there are many instances
where no sediment transport is detected despite winds
being far greater than this. In order to provide a quan-
titative means by which a single threshold can be de-
termined from such data series, Davidson-Arnott, Mac-
Quarrie and Aagaard (2005) recommend calculating the
mean of the five minimum windspeeds at which trans-
port is observed and the mean of the five maximum
windspeeds at which no sand transport is observed, thus
providing a range in the value of possible entrainment
thresholds.
Stout and Zobeck (1996, 1997) and Stout (2004) tackled
the problem by developing the time fraction equivalence
method, which has begun to be applied in a variety of
field situations (Wiggs, Atherton and Baird, 2004; Wiggs,
Baird and Atherton, 2004; Davidson-Arnott, MacQuar-
rie and Aagaard, 2005). The technique is based on the
observation that sand transport is intermittent (when mea-
sured at a frequency of
1 Hz) and the assumption that
the fraction of time in which sand transport events oc-
cur (described by the intermittency factor,
∼
) should be
equivalent to the fraction of time that the wind velocity is
equal to or greater than the threshold value for a particular
surface and at a particular time. By converting measure-
ments of sand transport and wind speed into two binary
series, the threshold value that gives rise to this equiva-
lence can be determined by iteration (Stout and Zobeck,
1996). A modification of the procedure employing a cu-
mulative frequency plot of wind velocity has also been
presented by Wiggs
et al.
(2004a) and is demonstrated in
Figure 18.13.
An entrainment threshold calculated using the time
fraction equivalence method is shown as B in Figure
18.11. While the technique allows a good objective
method by which to calculate a threshold, it is clear
that the inherent variability in sand transport data again
results in a threshold whereby there are instances of
both sand transport events below threshold and no sand
γ
60
50
40
B
30
A
20
10
0
2
0
1
3
4
u
m s
-1
5
6
7
8
Figure 18.12
Time series data of wind speed and detected sand transport (at 1 Hz sampling frequency). Entrainment thresholds
defined by (A) the onset of sediment transport and (B) the time fraction equivalence method are shown (from Wiggs, Atherton
