The spectrum of uncertainty in flood damage assessment

Abstract

The future of the world is becoming more uncertain owing to climate change. The unfolding impacts of climate change are affecting human societies and natural ecosystems. Projections of climate change impacts are associated with a cascade of uncertainties including greenhouse gas emissions scenarios, climate models, and associated processes. Climate models are essential for predicting flow and flood peaks, neces-sitating proper means of quantification and re-use to help refine the predictions made. This study presents an innovative new framework to quantify flood damage assessment as the climate changes. To integrate uncertainty in modelling catchment behaviour, the Quantile Flow Deviation (QFD) metric was used to attribute different sources of uncertainty, including those from variations in climate from point measurements and from extrapolation to flood peaks from the limited observed flows that are available. The square root of error variance (SREV) calculated from global climate model (GCM) precipitation outputs was used to quantify climate change uncertainty, thereby enabling the estimation of the uncertainty in modelled streamflow to allow the extent of change in flood damage to be assessed. Using data from the Leaf River catchment in the USA, this study presents the increase in flood damage uncertainty resulting from explicit consideration of uncertainty as well as the change in the climate as a function of global temperature rise.