Overall Conclusion and Discussion - Response of Flood Events to Land Use and Climate Change - page 223

• The digital elevation model that was used in the 2008 flood event was adjusted
by increasing the heights of the dyke by 0.5 m from that of the 1957 event. The
maximum estimated velocity for the 2008 flood in the channel was about *3m/
s and had a maximum water depth of 2.7 m. At cross-section 12, located exactly
at the Abattoir Bridge, the maximum water depth in the near bank was 1.7 m.
This result shows the best agreement between measured and observed data. This
means that the model calibration was done in the right way to access more
accurate and real results.
• Flow simulation results for 1982, 1995 and 2000 offered an idea about the area and
river change and show critical places in flow propagation of velocity and water
depth. River changes between 1957 and 2008 show that the river became narrow
especially in the middle. The dyke height in the city area also increased by 0.5 m
after the 1957 flood, preventing the urban area from any probable flood damages.
In all simulations, the right bank was much more sensitive to the velocity of water.
No inundation was observed in natural part of the river. Probably, the vegetation
on the banks has a small contribution to overall flow resistance.
Dyke scenario
• In case of dyke failure, it is essential to evaluate the possible effects in advance.
In this scenario (1), normal retention water depth and velocity were shown in the
channel—water flow was kept between the banks. The model indicated that the
maximum water depth and velocity in the Ubaye River were approximately
*2.7 m and *3 m/s, respectively. This corresponds to the field reports of the
2008 flood. The idea behind scenario (2) was to test the capacity of the channel to
the same hydrograph of the 2008 flood in case of dyke failure. The results showed
that the flow propagation was affected by the channel maintenance and structural
dyke integrity, and in several locations the flow overtopped the bank and the
floodplain was affected by inundation. The assumption was that the lack of
maintenance produced the dyke break and inundation even with a low amount of
water. Scenario (3) was used only as support scenario to compare the simulation
with other scenarios. The results obtained from scenario (4) showed that the
dykes overtopped on several sections in both banks, especially in the urban
sectional area. Critical overtopping occurred between cross-sections 11 and 12
on both the left and right banks (*2.9 km from upstream). This shows that at the
current river morphology and dyke structure, although this hydraulic structure
was a good strategy to protect the city part of the river in case of severe flood,
there is still concern about overtopping and inundation in case of dyke failure.
Model Scenario
• In this study two different flood modeling software packages were used to
simulate floods. Based on the user's purpose, it has to be determined which
model should be applied. If engineers have to decide on dams or house
enforcements against flash flows, the SOBEK results are more suitable due to
higher accuracy of the flow velocity results. However, if only the approximate
extent of water is important (e.g., for spatial planning procedures), the HEC-