Characteristics, drivers, and predictability of flood events in the Tana River Basin, Kenya

Abstract

Study area: Tana River Basin in Kenya. Study Focus: Flood-related impacts and losses have been rising. Therefore, understanding flood characteristics, drivers, and predictability is critical for informed decisions in the ongoing flood early warning (FldEWS) projects. This study presents an in-depth analysis of hydro-meteorological, Sentinel Mission (SM), and ensemble hydrological model datasets. We examine flood characteristics using observed hydro-meteorological and SM datasets, followed by statistical analysis of climate drivers of flood events at inter-annual and sub-seasonal (S2S) time scales. Finally, reforecasts from Global Flood Awareness System (GloFAS) are assessed against observed river flows. New hydrological insights for the study region: There is a high inter-annual variability of flood events with flood peaks occurring in May and December. SM satellites have the ability to map flooded areas in near-real time. At inter-annual timescales, positive Indian Ocean Dipole (IOD) and warm El Niño Southern Oscillation (ENSO) drives short rains (October to December). At Sub-Seasonal (S2S) timescales, Madden Julian Oscillation (MJO; phases 2-4) is a notable driver of flood related extreme rainfall. GloFAS offers reliable forecasts depending on the flood magnitude, trigger probability, and 'anticipation window' and it meets the tolerable skill requirements for flood preparedness actions (FAR 50%) with up to a 20-day lead time for 1 and 2-year return periods. We subsequently discuss how our research findings can inform the development of FldEWS in Kenya, with an emphasis on improved co-production of flood forecast information with relevant stakeholders.

Keywords

• Madden Julian Oscillation
• El Niño Southern Oscillation
• Indian Ocean Dipole
• Global Flood Awareness System
• Flood early warning system