Uncertainty in flood damage estimation*

The authors gratefully acknowledge the helpful advices and comments from Matthijs Kok, Anne Barendregt, Maarten‐Jan Kallen, and two anonymous referees.

Abstract

The aim of this article is to incorporate uncertainty into the currently used method for estimating the economic damage of floods in the Netherlands. Using a high‐water information system, this so‐called Standard Method computes the expected flood damage per damage category on the basis of the number of objects subject to flooding, the maximum damage per object, and the relative damage depending on the water depth. Using a probabilistic approach, the Standard Method is extended by taking account of the uncertainties in the maximum damage per object (maximum possible damage per unit object; e.g., house, production facility, meter road) and the damage function (describing the proportion of the maximum damage incurred to an object due to flooding as a function of the water depth). In estimating the damage in a flooded area, it is also important to take account of the spatial dependence between flood damages at different locations. For this purpose, the following three types of spatial dependence are considered: complete spatial dependence, spatial independence, and partial spatial dependence (for which the damage functions are completely dependent for independent water‐depth classes). Using Monte‐Carlo simulation, probability distributions of the flood damage are determined. The new uncertainty‐based model for predicting flood damage is applied in a case study and the results of a sensitivity analysis are reported.

Keywords:

• Flood damage
• high‐water information system
• uncertainty
• flood‐risk analysis
• probability distributions
• spatial dependence
• Monte‐Carlo simulation

Notes

The authors gratefully acknowledge the helpful advices and comments from Matthijs Kok, Anne Barendregt, Maarten‐Jan Kallen, and two anonymous referees.