 |
| Titel |
Effect of uncertainty in Digital Surface Models on the boundary of inundated areas |
| VerfasserIn |
I. Nalbantis, I. Papageorgaki, P. Sioras, Ch. Ioannidis |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250066312
|
|
|
|
|
|
| Zusammenfassung |
| The planning, design and operation of flood damage reduction works or non-structural
measures require the construction of maps that indicate zones to be potentially inundated
during floods. Referring to floods due to heavy rainfall, the common procedure for
flood mapping consists of the following five computational steps: (1) Frequency
analysis of extreme rainfall; (2) construction of design hyetographs for various return
periods; (3) construction of the related direct runoff hydrographs; (4) routing of
these hydrographs through the hydrographic network; (5) mapping of the inundated
area that corresponds to the temporally maximum depth for each location in the
flood plain. Steps 3 through 5 require the use of spatial information which can be
easily obtained from a Digital Surface Model (DSM). The DSM contains grid-based
elevations of the ground or overlying objects that influence the propagation of flood
waves.
In this work, the SCS-CN method is used in step 3 in combination with a synthetic Unit
Hydrograph based on the SCS dimensionless Unit Hydrograph. In step 4, the full
one-dimensional Saint Venant equations for non-uniform unsteady flow on fixed bed are
used, which are numerically solved.
The impact of uncertainty in the DSM on the inundated area boundary is investigated. For
this the Monte Carlo simulation method is employed to produce a large number of
erroneous DSMs through introducing errors in elevation with a standard deviation
equal to Ïă. These DSMs are then used for delineating potentially flooded areas. The
standard deviation of the distance (from the riverbed axis) of the boundary of these
areas, herein denoted as ÏăF, is used as the measure of the resulting uncertainty. The
link between Ïă and ÏăF is examined for a spectrum of large return periods (100 to
10000).
A computer experiment was set up based on data from two drainage basins. The first
basin is located in East Attica and is drained by a branch of the Erasinos Torrent named the
South-East Kalyvia Torrent; it extends over an area of about 17 square kilometres. The
second basin is that of the Kerynitis River in north-western Peloponnesus; it covers an area of
89 square kilometres. In each one of the two basins hydrographs at the outlet of the upper part
of the basin are estimated with the aid of hydrological modelling, while, for the lower part
hydraulic routing is employed. The South-East Kalyvia basin is hilly, whereas the
Kerynitis Basin shows high ground slopes in its upper part and low slopes in the lower
part.
Graphs of Ïă vs. ÏăF and maps showing the mean position μF of the boundary of flooded
area along with limits of this boundary that reflect positions μF±2ÏăF help visualize the
impact of the uncertainty in DSM. To acquire a better feeling of the effect of DSM
uncertainty, results are compared to those obtained from uncertain rainfall depths of the
design hyetographs. |
|
|
|
|
|
|