Geoscience Reference
In-Depth Information
intensities and wander over the landscape delivering their
water along discrete tracks. In consequence, autumn and
spring storms are flashier, the flood timebase is shorter,
but,
ceteris paribus
, peak discharge is greater.
In areas of pronounced topography, orographic height-
ening of convective storms can have a significant effect
on the spatial pattern of rainfall, a factor that has to be
taken into account when assessing probable stream dis-
charge in basins that are not greatly separated geographi-
cally. Wheater
et al
. (1991b) demonstrate reasonably well-
defined direct relations between both number of raindays
and annual rainfall and altitude in the Asir escarpment
of southwest Saudi Arabia. Martınez-Goytre, House and
Baker (1994) suggest a rain-shadow effect in the Santa
Catalina Mountain range of southeastern Arizona after
analysing the magnitude of palaeofloods in basins dis-
posed at different locations around the massif.
3.0
(a)
2.5
2.0
1.5
1.0
0.5
0
0
50
100
150
200
250
300
Flood peak discharge (m
3
s
-1
)
400
(b)
300
200
100
0
13.2.2
Flash flood hydrograph
0
1
2
3
4
5
Maximum 20-minute rainfall intensity (mm/10min)
The peculiarities of storms in desert areas mean that the
relations between rainfall and runoff are extremely com-
plex, almost precluding hydrological modelling (Wood-
ing, 1966; Renard and Lane, 1975; Srikanthan and McMa-
hon, 1980). Wheater and Brown (1989) (Figure 13.3) have
illustrated poor relations between flood peak discharge
and maximum rainfall intensity and between flood runoff
and storm rainfall volume in Wadi Ghat, a 597 km
2
basin
draining towards the Red Sea in southwestern Saudi Ara-
bia. The most convincing relation they show is, ironically,
between flood runoff volume and flood peak discharge,
suggesting some consistency between basic hydrograph
shape and flood magnitude. In contrast, Reid, Powell and
Laronne (1998) show a fairly convincing relation between
flood runoff and rainfall volumes for the Nahal Eshtemoa,
a 112 km
2
catchment in the semi-arid northern Negev. In
fact, although the information base is small for modelling
purposes, despite the 15-year record, Alexandrov
et al
.
(2009) provide a fairly clear-cut flood magnitude proba-
bility relation for the same catchment (Figure 13.4). The
emergence of this relation is surprising. Despite reports
of nonstationary relations for runoff elsewhere (specifi-
cally, the Wadi Wahrane basin of northern Algeria; see
Benkhaled and Remini, 2003), it might be used to give
water resource managers at least some hope that stochas-
tic flood prediction is feasible, even in semi-arid regions.
It might encourage greater investment in flow gauging,
despite the infrequence and uncertainty of river floods.
For catchments of modest size, there are three char-
acteristics apparent in the flood hydrographs of streams
3
(c)
2
1
0
0
10
20
30
Rainfall volume (mm)
Figure 13.3
(a) Surprisingly good relation between flash-
flood volume and peak discharge, Wadi Ghat, southwest Saudi
Arabia. (b and c) Lack of relations between flash flood param-
eters and storm rainfall, Wadi Ghat, southwest Saudi Arabia
(after Wheater and Brown, 1989).
steep rising limb that incorporates a bore (Figures 13.5
and 13.6). McGee (1897) gave a remarkable first descrip-
tion of a bore, though for an unconfined sheetflood on a
bajada fronting a mountain range in the Sonoran Desert.
Since then the phenomenon has been noted in ephemeral
channels in different deserts by various workers, including
Hubbell and Gardner (1944), Leopold and Miller (1956),
Renard and Keppel (1966), Gavrilovic (1969) and Fro-
stick and Reid (1979). The bore is rarely the 'wall of
water' often referred to in popular literature about flash
floods. Hassan (1990a) (Figure 13.7) has related the speed
