The Utility Of Impact Data In Flood Forecast Verification For Anticipatory Actions
John Wiley & Sons Ltd
Download
Abstract: This article explores how flood impact data can enhance forecast verification and anticipatory action protocols in Uganda and Kenya. It evaluates five impact repositories including EM-DAT and local Red Cross data, comparing them with river-gauge observations and GloFAS forecasts. Verification metrics such as Type I and II errors and Probability of Detection are introduced, alongside recommendations for harmonizing impact reports, satellite imagery integration, and AI-driven expansion. Kenya?s flood basins (Tana, Nzoia, Athi) are analyzed to illustrate forecast-based financing opportunities and grassroots preparedness efforts.
Author:
Faith Mitheu, Elena Tarnavsky, Andrea Ficchì, Elisabeth Stephens, Rosalind Cornforth, Celia Petty
Theme/Sector:
Early Warning Systems, Climate Information Services, East Africa, Disaster Risk and Reduction
Year
2023