Towards integrating community and institutional flood early warning systems: A framework applied to an Andean tropical case

Flood Early Warning Systems are critical for reducing flood risk because they generate efficient and timely hazard warnings. The importance of involving local vulnerable communities in designing, implementing, and managing these systems has been recognized in research and practice. However, community-based flood early warning systems are sometimes initiated and operated by the communities themselves, ignoring, either deliberately or not, related institutional flood risk management efforts in the same area, which may bring conflicting warnings in times of flood occurrences. This is the case of an urban catchment in the Colombian Andes. The question, which, to the best of our knowledge, has not yet been answered, is how existing community and institutional flood early warning systems initiatives operating in the same area can be integrated and improved. This paper aims to answer this question by putting forward a methodology (The SWOT method) that identifies the strengths, opportunities, weaknesses, and threats of each early warning system separately and then identifies the aspects for their integration and improvement, following the four pillars of a modern early warning system (risk knowledge, monitoring and forecasting, warning dissemination, and response capabilities). This is done by building and analyzing matrices for each flood early warning system obtained from historical data, workshops, and interviews with relevant stakeholders. Results show that the pillars of monitoring, forecasting, and response capabilities can be enhanced by their integration, particularly by using cutting-edge monitoring technologies (low-cost water level sensors) and implementing citizen science projects (increasing the number of observers upstream to improve the redundancy for accurate warnings). Furthermore, other social, technical, and technological aspects can be considered to work toward an integrated early warning system.