Geoscience Reference
In-Depth Information
0.6
(a)
0.4
0.2
0
-0.2
-0.4
(b)
0.4
0.2
0
0
10
20
30
40
50
60
70
80
90
Gauge separation (km)
Figure 13.2
(a) Cross-correlation of hourly rainfall as a function of gauge separation, Wadi Yiba, southwest Saudi Arabia. (b)
Probability of simultaneous occurrence of hourly rainfall as a function of gauge separation, Wadi Yiba, southwest Saudi Arabia.
Both reflect the spotty nature of rainfall in desert regions (after Wheater
et al
., 1991b).
There is therefore less likelihood of identifying a
char-
acteristic
outfall hydrograph for an ephemeral system, in
contrast with perennial streams affected by widespread
frontal rain (Wheater and Brown, 1989). Besides this, the
seemingly fixed maximum storm-cell size means that the
fraction of a river basin that will be affected by rain will
fall as the basin size increases. Although it is conceivable
that several storms will wet a large basin more or less
simultaneously, atmospheric dynamics will dictate that
they are widely separated geographically. Sharon (1974)
has suggested that the distance between storm centres may
average 40-60 km.
One other factor that has considerable bearing on the
nature of river discharge is that storm cells migrate as
they deliver their rain (Sharon, 1972; Frostick and Reid,
1977; Frostick, Reid and Layman, 1983). Again, there are
implications for the shape of the flood hydrograph, since
a storm moving upstream across a drainage basin will
cause runoff from lower tributaries to occur earlier than
those of headwater regions. Indeed, the fact that over-
land flow is generated within minutes in arid environ-
ments (Yair and Lavee, 1976; Reid and Frostick, 1986)
encourages this temporal separation of contributions from
the multipeak nature of flash floods reflects, among other
things, the piggybacked contributions of individual trib-
utary water catchments as each receives and disposes of
rainfall. Countering this in part, Ben-Zvi, Massoth and
Schick (1991) have demonstrated for ephemeral streams
in Israel that the downstream travel time of flood crests is
governed most significantly by crest height, as would be
expected from hydraulic considerations of wave celerity.
This would suggest that the contributions from individual
tributaries would overrun each other in a manner dictated
by the spatial inhomogeneity of storm rainfall, as much as
by other factors. Because spatiotemporal rainfall patterns
are highly unpredictable, this has to add to the complexity
and changeability of the river outfall hydrograph (Reid,
Laronne and Powell, 1998).
Alexandrov, Laronne and Reid (2007) have used the
climatically transitional nature of the Levant to highlight
seasonal differences in catchment runoff response that
arise from different types of storm. Here, atmospheric
conditions dictate that winter months are characterised by
frontal rainstorms, which have low rain intensities and
deliver water fairly ubiquitously throughout a catchment.
In contrast, autumn and spring are characterised by con-
