Probabilistic flood hazard mapping: effects of uncertain boundary conditions

摘要

Comprehensive flood risk assessment studies should quantify the globaluncertainty in flood hazard estimation, for instance by mapping inundationextents together with their confidence intervals. This appears of particularimportance in the case of flood hazard assessments along dike-protectedreaches, where the possibility of occurrence of dike failures mayconsiderably enhance the uncertainty. We present a methodology to deriveprobabilistic flood maps in dike-protected flood prone areas, where severalsources of uncertainty are taken into account. In particular, this paperfocuses on a 50 km reach of River Po (Italy) and three major sources ofuncertainty in hydraulic modelling and flood mapping: uncertainties in the(i) upstream and (ii) downstream boundary conditions, and (iii) uncertaintiesin dike failures. Uncertainties in the definition of upstream boundaryconditions (i.e. design-hydrographs) are assessed through a copula-basedbivariate analysis of flood peaks and volumes. Uncertainties in thedefinition of downstream boundary conditions are characterised by uncertaintyin the rating curve with confidence intervals which reflect dischargemeasurement and interpolation errors. The effects of uncertainties inboundary conditions and randomness of dike failures are assessed by means ofthe Inundation Hazard Assessment Model (IHAM), a recently proposed hybridprobabilistic-deterministic model that considers three different dike failuremechanisms: overtopping, piping and micro-instability due to seepage. Theresults of the study show that the IHAM-based analysis enables probabilisticflood hazard mapping and provides decision-makers with a fundamental piece ofinformation for devising and implementing flood risk mitigation strategies inthe presence of various sources of uncertainty.