Description

Disaster risk reduction has been moving from hazard event prediction (e.g., forecasting and early warning) to estimating impacts before the actual occurrence of the event. For tropical storm events for example this involves coupling storm path monitoring and modelling with pre-calculated scenarios of expected losses for different associated hazard types (wind, storm surges, rainfall, etc.) and different intensities. This allows a more targeted intervention, such as more focused and targeted preparatory action or evacuation, or disbursement of aid prior to the actual event (forecast-based financing). However, what effectively amounts to a last-minute risk assessment (= determination of expected losses) strongly hinges on the availability of the necessary data (data for a risk assessment, or availability of historic loss data), and is subject to various uncertainties. Storm path uncertainty strongly increases with distance to shore and lead time, causing a dilemma: early assessments can be done well in time, but the coastal area that may be affected will be very large. Conversely, last-minute assessments can make a more accurate landfall prediction, but leave less time for the analysis and for preventive measures.

Objectives and Methodology

The purpose of this project is twofold: carry out image-based damage assessment for an event where IBF was done, to determine the accuracy of the loss prediction, where possible for individual hazard types. Secondly, investigate the forecast timing dilemma, specifically the relation between large uncertainty of storm path and impact area for early assessments vs. the need for detailed/local loss prediction.

Further reading