SEMINAR | Julian Eleutério (Federal University of Minas Gerais, Brazil)

Flood warning and evacuation systems (WES) aim to reduce flood-related risk by promoting the adoption of protective actions among exposed populations. Although significant progress has been made in hydrometeorological forecasting, the effectiveness of WES depends on a complex chain of processes that remains insufficiently explored. This chain includes forecasting, decision-making, warning dissemination, and evacuation—all with their human components. Understanding and optimizing the timing associated with each of these stages is a major challenge for risk management.

This study is part of recent research efforts to optimize WES through the integrated simulation of their various components. It evaluates, in different contexts, the ability of simulation models to estimate the impacts of floods and the effectiveness of WES in reducing loss of life. In Europe, the case of the catastrophic July 2021 floods in Germany is analyzed. This event caused 134 deaths in the Ahr Valley, in a context marked by failures in warning content, issuance, and dissemination. Alternative early warning and evacuation scenarios are assessed using the multi-agent model LifeSim in order to estimate potential variations in fatalities.

In addition, three cases of dam-break flooding in Brazil are examined: two real events (Miraí in 2007 and Brumadinho in 2019) and one hypothetical scenario. The Miraí case illustrates the effectiveness of hazard identification and evacuation procedures, which prevented loss of life despite significant material damage, whereas the Brumadinho disaster caused 272 deaths in the complete absence of warning. The hypothetical scenario allowed for a deeper analysis of the temporal and spatial dimensions of WES.

Simulations conducted for all these cases demonstrate the strong potential of WES models as decision-support tools to improve their effectiveness. They also highlight the importance of considering the entire chain of WES actions, particularly the integration of human behavior—an essential factor in achieving significant risk reduction and a central focus of ongoing research.