Abstract/Summary

Flooding has the highest frequency of occurrence of all types of disaster arising from natural hazards across the globe. The anticipation and forecasting of floods is a key component in managing, preparing for and mitigating the effects of severe events, from local to international scales. This research aims to explore ways to extend the predictability of flood hazard at the global scale and provide earlier indications of potential flood events. Two approaches for predicting river flow extremes on seasonal timescales are developed and tested; statistical forecasts based on the known influence of El Niño and La Niña on river flow and flooding at the global scale, and dynamical forecasts using numerical weather prediction systems. The statistical forecast development has shown that the likelihood of increased or decreased flood hazard during El Niño and La Niña events is much more complex than is often perceived and reported. The dynamical forecasts are shown to be more skilful than a long-term average climatology in many rivers worldwide, up to four months in advance in some cases. These approaches both have the potential to provide early warning information, and to support El Niño preparedness activities. As such, a comparison of the ability of the two forecasts to predict hydrological extremes during El Niño is undertaken, highlighting regions of the globe where each forecast is (or is not) skilful compared to a forecast of climatology, and the advantages and disadvantages of each approach. Both of these new seasonal hydro-meteorological forecasts are openly available, with the dynamical forecasts produced operationally as part of the Global Flood Awareness System (GloFAS-Seasonal), supported by the Copernicus Emergency Management Service. This research has provided a step change in moving from forecasts that were previously only available for precipitation, to global-scale forecasts of hydrological variables at extended lead-times.