HR: 1340h
AN: H43A-0490 [Abstracts]
TI: Spatially distributed model calibration of flood inundation guided by consequences such as loss of
property
AU: Pappenberger, F
EM: f.pappenberger@lancaster.ac.uk
AF: Lancaster University, Environmental Science/Lancaster Environment Centre
, Lancaster, LA1 4YQ
United Kingdom
AU: * Beven, K J
EM: k.beven@lancaster.ac.uk
AF: Lancaster University, Environmental Science/Lancaster Environment Centre
, Lancaster, LA1 4YQ
United Kingdom
AU: Frodsham, K
EM: k.frodsham@lancaster.ac.uk
AF: Lancaster University, Environmental Science/Lancaster Environment Centre
, Lancaster, LA1 4YQ
United Kingdom
AU: Matgen, P
EM: matgen@crpgl.lu
AF: Centre de Recherche Public-Gabriel Lippmann, 162a, avenue de la Faiencerie, Luxembourg, L-1511
Luxembourg
AB:
Flood inundation models play an increasingly important role in assessing flood risk. The growth of 2D inundation models that
are intimately related to raster maps of floodplains is occurring at the same time as an increase in the availability of 2D
remote data (e.g. SAR images and aerial photographs), against which model performancee can be evaluated. This requires new
techniques to be explored in order to evaluate model performance in two dimensional space. In this paper we present a
fuzzified pattern matching algorithm which compares favorably to a set of traditional measures. However, we further argue
that model calibration has to go beyond the comparison of physical properties and should demonstrate how a weighting towards
consequences, such as loss of property, can enhance model focus and prediction. Indeed, it will be necessary to abandon a
fully spatial comparison in many scenarios to concentrate the model calibration exercise on specific points such as
hospitals, police stations or emergency response centers. It can be shown that such point evaluations lead to significantly
different flood hazard maps due to the averaging effect of a spatial performance measure. A strategy to balance the different
needs (accuracy at certain spatial points and acceptable spatial performance) has to be based in a public and political
decision making process.
DE: 1821 Floods
DE: 1846 Model calibration (3333)
DE: 1873 Uncertainty assessment (3275)
SC: Hydrology [H]
MN: Fall Meeting 2005