16 Flood Forecasting and Control
369
seen from Fig. 16.7 that also the resulting flood forecasts are not in a too good
agreement with the flood hydro graph which was observed after the forecast. The
forecast hydro graph with the probability of non-exceedence of 0.6 (= 60 %) underestimates the peak while the volume is in reasonable agreement with that of the
observed hydro graph. The forecast hydro graph with the probability of nonexceedence of 0.99 (99 %) matches the observed peak, the volume, however, is
much too high as compared with the observed hydro graph. Nevertheless such
forecasts are a great help for flood warning or for the operation of flood protection
reservoirs.
3.2
1.2
0 .8Mean rainfall
(mm)
Pu = probability of
non·exceedance
o
~ ime (hours)
o
16
32
48
64
80
96
112
80
64 -
48
32 -
16r-07. 08. -i'-08. 08 . -+ -09. 08. -+-10. 08. -+ - 11.08.
- - Time of forecast
Runoff (m 3fs)
~- observed
,l(.
i
\
:
.
I
i
Pu = 0,99
Pu = 0,90
.. - --- Pu = 0,60
o
I
Time (hours)
o
16
32
48
64
80
96
112
r -07. 08. -i(-: : . 08.08. - ,.- 09.08. -+ -10. 08. + 11.08.
L_ Time of forecast
Fig. 16.7. Flood forecast based on radar rainfall measurement and rainfall forecast (QPF)