Uncertainty in flood damage estimates and its potential effect on investment decisions

Abstract

Abstract. This paper addresses the large differences that are found between damage estimates of different flood damage models. It explains how implicit assumptions in flood damage functions and maximum damages can have large effects on flood damage estimates. This explanation is then used to quantify the uncertainty in the damage estimates with a Monte Carlo analysis. The Monte Carlo analysis uses a damage function library with 272 functions from seven different flood damage models. The paper shows that the resulting uncertainties in estimated damages are in the order of magnitude of a factor of 2 to 5. The uncertainty is typically larger for flood events with small water depths and for smaller flood events. The implications of the uncertainty in damage estimates for flood risk management are illustrated by a case study in which the economic optimal investment strategy for a dike segment in the Netherlands is determined. The case study shows that the uncertainty in flood damage estimates can lead to significant over-or under-investments.

Estimation of river flood damages in Jakarta, Indonesia

“…For further research, we recommend to also use decision tree analysis as used in Merz et al (2013) and to design the surveys accordingly. We tend to agree with Middelmann-Fernandes (2010) and Wagenaar et al (2016), who emphasized the consideration of multiple methods in assessing flood damage. Thus, a combination of a damage assessment based on survey and flood model based on expert assessment such as damage scanner provides a better picture of flood damage in a flooding-prone area.…”

“…All VIF values for five independent variables are less than 10, meaning that there is no multicollinearity or no correlation between e i and the independent variables. Moreover, this also indicates that the linear relationship among the independent variables leads to reliable regression estimates (Verbeek 2012). However, we rejected the null hypothesis of homoscedasticity because the NR 2 (76.83) is higher than v 2 (5) at 5% significant level (9.49).…”

“…The OLS produces unbiased and efficient estimated coefficients in the linear model (Greene 2012;Verbeek 2012). Based on Eq.…”

“…Following Wagenaar et al (2016), uncertainties in the survey arise because flood damage was assessed based on only one single event and limited to certain ranges of depth and duration. In the damage scanner model, uncertainties appear in the construction of damage curve, the asset values associated with the curve and difference in methodological framework (Jongman et al 2012;Merz et al 2010;Wagenaar et al 2016). In the damage scanner, aleatory uncertainty occurs because it used average data in valuing the maximum damage per object, assumed a uniform damage per object in each land-use class and defined expected damage by expert judgement with artificial depth scenarios.…”

“…These factors lead to potential errors in the damage curve. To reduce this uncertainty, the average between different functions could be use as damage function (Wagenaar et al 2016). Additionally, Pistrika and Jonkman (2010) found that significant uncertainty arises in damage curves when the curves are applied in small case study areas.…”