TITLE:
Flood Risk for Embanked Rivers
AUTHORS:
Ewa Bogdanowicz, Witold G. Strupczewski, Krzysztof Kochanek, Iwona Markiewicz
KEYWORDS:
Flood Frequency Analysis, Levee Break, Flood Duration, Maximum Likelihood, Nonstationarity
JOURNAL NAME:
Journal of Geoscience and Environment Protection,
Vol.2 No.3,
June
13,
2014
ABSTRACT:
Flood frequency analysis (FFA) concentrates
on peak flows of flood hydrographs. However, floods that last years devastated
large parts of Poland lead us to revision of the views on the assessment of
flood risk in Poland. It turned out that it is the prolonged exposure to high
water on levees that causes floods, not only the water overflowing the levee
crest. This is because, the levees are weakened by water and their disruption
occurs when it seems that the danger is over, i.e. after passing culmination.
Two main causes of inundation of embanked rivers, namely over-crest flow and
wash out of the levees, are combined to assess the total risk of inundation.
Therefore the risk of inundation is the total of risk of exceeding embankment
crest by flood peak and risk of washout of levees. Hence, while modeling the
flood events in addition to the maximum flow one should consider also the
duration of high water in a river channel, Analysis of the frequency of annual
peak flows based on annual maxima and peaks over threshold is the subject of
countless publications. Therefore we will here mainly modeling the duration of
high water levels. In the paper the two-component model of flood hydrograph
shape i.e. “duration of flooding-discharge- probability of nonexceedance”
(DqF), with the methodology of its parameters estimation for stationary case
was developed as a completion to the classical FFA with possible extension to
non stationary flood regime. The model combined with the technical evaluation
of probability of levees breach due to the d-days duration of flow above alarm
stage gives the annual probability of inundation caused by the embankment
breaking. The results of theoretical research were supplemented by a practical
example of the model application to the series for daily flow in the Vistula
River in Szczucin. Regardless promising results, this method is still in its
infancy despite its great cognitive potential and practical importance.
Therefore, we would like to point to the usefulness and necessity of the DqF
models to the one-dimensional analysis of the peak flood hydrographs and to
flood risk analysis. This approach constitutes a new direction in FFA for
embanked rivers.