Development of a flood-damage-based flood forecasting technique

Abstract

Flood forecasting is a pre-emptive non-structural measure used to mitigate inundation. Most current flood forecasting techniques incorporate complex processes, such as training and optimization, before the technique can be applied. Conventional flood forecasting techniques, based on flood volume, provide alerts even if there is no significant risk of flood damage. In this study, a new flood forecasting technique has been developed based on likely flood damage using the multi-dimensional flood damage analysis method. This new flood forecasting technique overcomes the drawbacks of current flood forecasting techniques because it can be easily applied using rainfall data. The studied drainage area was divided into subareas, and the damage functions were obtained for each subarea using the flood volumes and damage information. Using these damage functions, the rainfall intensity when the flood damage initially occurred was calculated for each duration and subarea. The damage graph produced for flood forecasting in each subarea identified the rainfall intensities and durations that resulted from the initial occurrence of flood damage. This new flood forecasting technique could be used to save lives, valuable assets, and manage drainage areas.