FloodCitiSense: citizen-led mitigation of urban flooding

Alerts that enable faster response times

Vulnerability to flooding can be reduced with better warning systems that alert the public and enable faster incident response. Through our involvement in the FloodCitiSense European research project and working with residents, RPS has developed innovative early-warning systems for urban pluvial flooding.

Bringing together global expertise from academia, industry, local councils and start-up companies, the project team set up urban living labs in three European pilot cities: Rotterdam, Brussels and Birmingham.

Low-cost, pragmatic and data-driven flood forecasting approaches were developed for each pilot. Novel methods were tested, including machine-learning based (using Eigenface models commonly used for facial recognition) and an analogue system in which the weather forecast is matched to historic climatic conditions. RPS’s deep expertise in flood modelling fed into the operational implementation and testing of both these techniques.

Residents fundamentally shaped the project. They co-created the free project app, which is used to collect live rainfall and flood data to generate flood alerts. Our team supported UK resident engagement activities through participation at workshops and conceptual design of the forecasting system.

Predictability driven by data

FloodCitiSense has demonstrated that current data, models and technology enable the ready implementation of data-driven pluvial flood forecasting models. The project data helps us to better understand who is at risk and when.

Project partners at Severn Trent Water are in the early stages of developing a flood forecasting system for their region, drawing on the lessons from this project, and ultimately reducing vulnerability of their customers and staff.