Geoscience Reference
In-Depth Information
that will cope with all but the most extreme events.
The frequency of heavy rainfall events may also be of
significance in affecting slope processes and landslips. It
is important to remember, however, that the figures are
only probabilities, derived from average conditions over
a specific period. It is quite possible for two storms with
an average recurrence interval of fifty years to occur on
successive days!
Another way of expressing information on rainfall
variation is to plot annual rainfall totals on similar graphs.
Thus in
Figure 5.10
the frequencies of annual rainfall for
Sheffield and Timimoun are shown. We can see that there
is a 50 per cent probability of at least 820 mm of rainfall
occurring in any year at Sheffield, while at Timimoun the
equivalent total is 19 mm. It is also apparent from the
graphs that the variability at Sheffield is fairly small
compared with that at Timimoun, although the latter is,
on average, much drier.
Again this type of data may be very useful. For example,
a particular crop may grow satisfactorily only if the annual
rainfall exceeds 600 mm. From information on annual
rainfall frequencies it is possible to determine the likeli-
hood of receiving that amount of precipitation. If the
probability is, say, 90 per cent, the farmer may well think
it worth while to grow the crop; if it is only 20 per cent,
it is unlikely to be worth the risk. Similarly, it is possible
to determine in the same way how often, on average, it will
be necessary to irrigate crops.
Rainfall variability may also be expressed statistically
by the coefficient of variation (CV). This is calculated
from the formula:
CV = (
s
/
¯
)
100 %
where
¯
is the average rainfall and
s
is the
standard
deviation
. This defines the variability relative to the mean.
With a standard deviation of 200 mm and a mean annual
precipitation of 1,600 mm, the coefficient of variation
would be 12·5 per cent, but with the same standard
deviation and a mean of only 400 mm the coefficient of
variation would rise to 50 per cent. This is a useful
measure, since it gives an indication of the importance of
the variability. In Britain coefficients of variation of annual
rainfall range between 10 per cent in north-west Scotland
and about 20 per cent in south-east England.
Spatial variations of precipitation
We all know that annual rainfall totals vary from one part
of the world to another, even when altitude is allowed for.
1100
140
SHEFFIELD
(1883 - 2006)
1000
120
100
900
800
80
TIMIMOUN (Algeria)
(1912 - 1960)
700
60
600
40
500
20
400
0.01
0
0.1
1
5 0 20
50
90
99
99.9
99.99
Cumulative percentage
Figure 5.10
Cumulative percentage frequency graphs of annual precipitation at Sheffield (1883-2006) and Timimoun, Algeria
(1912-60).
