Previsico provides real-time, street-level surface water flood forecasts to save lives and livelihoods globally
Please briefly describe your Water ChangeMaker journey
Previsico addresses the global challenge of protecting lives and livelihoods from surface water flooding – the single greatest cause of floods in the UK and the fastest growing cause in Europe (EASAC, 2018). Hourly changes in weather patterns can cause floods which are not detected using existing forecasting systems due to spatial resolution and analytical limitations. Internationally peer-reviewed research that underpins Previsico has found that flooding impacts emergency response times in England, with vulnerable population facilities – care homes, schools, nurseries and sheltered accommodation – becoming unreachable within the 7-minute window for life-threatening incidents. Some 84% of the English population can be reached by ambulance within the mandatory timeframe. This drops to 70% for a 30-year flood,and 61% for a 100-year event. The proportion of the elderly population that can be reached during a 30-year event is 65% (80% in normal conditions). Climate change is expected to increase the magnitude and frequency of flood events which will make responding to emergencies more difficult. By 2030, an estimated 147M people a year will be affected, costing over $700B, (WRI 2020). This is further exacerbated by the continued practice of developing in high flood risk areas. Third-generation flood prediction technology is crucial for global resilience.
Please describe the change that your initiative created and how was it achieved
Loughborough University’s Dapeng Yu and Rob Wilby have advanced knowledge in flood modelling, communicating flood risk information and evaluating the accuracy of flood nowcasting since 2001. They began working with the UK Met Office in 2016 to utilise high-intensity rainfall nowcasts to predict short-duration surface water flooding at street level with up to 36 hours warning. This led to the development of FloodMap Live, a novel method for surface water flood nowcasting which utilises real-time weather nowcast products in high-resolution surface water flood modelling to predict emerging risks around the clock. Leicestershire and Rutland Resilience Forum (LRRF) was FloodMap Live’s first adopter. Leicester City is ranked 16th out of 4215 UK settlements for surface water flood risk with 36,900 properties occupying flood-prone areas (Defra, 2009). Existing flood forecasting systems were insufficient to support its response to major events in 2007 and 2012 due to their low resolution and poor accuracy. In June 2019, FloodMap Live helped LRRF’s Operations Team to mobilise resources and coordinate its response ahead of flood events across the county – cutting response times and reducing flood damage. Further pilots followed with Birmingham, London and Manchester to provide live data feed via the Cabinet Office’s ResilienceDirect platform for emergency responders in 2019. The same year, spinout company, Previsico, was launched to exclusively licence FloodMap Live. Partnerships with global brands in the insurance (information is currently confidential) software (IBM) and engineering (WSP) sectors are enabling Previsico to build capability to deliver flood nowcasting services in the US, and worldwide.
How did your initiative help build resilience to climate change?
Previsico provides continuous modelling up to 48 hours in advance, updated every three hours, using the latest IBM Weather rainfall forecasts. It uses ‘hyperlocal’ (25m resolution = property-level) forecasts for surface water, river and coastal flooding to provide bespoke and off-the-shelf analytical solutions showing the accessibility of operations teams, enabling organisations to prepare for flooding by moving assets to safety or erecting defences. With its source code a trade secret, FloodMap Live continuously improves with every flood, sourcing Big Data for calibration and developing machine-learning capability. The technology gives critical asset managers, insurance companies, public sector and humanitarian organisations advanced information about expected surface water flood risks in cities, potentially avoiding loss of life or substantial social and economic costs. The technology is validated in 50 cities across 5 continents, including China where it was deployed in a live drill to prepare millions of citizens for potentially catastrophic typhoon flooding. (Supporting evidence case studies are provided.)
What water-related decisions did your initiative influence or improve?
Surface water flood warnings have been exceptionally crude lacking spatial resolution to help government and emergency services prepare and respond to flood events. Luana Avagliano, Head of ResilienceDirect, Cabinet Office describes FloodMap Live as “immensely assisting our resilience community in making informed decisions for planning and response to flood events and impact”. This is achieved by providing real-time access to a new 48-hour flood forecast that predicts flooding at 25m resolution in urban areas (100m elsewhere). The ResilienceDirect impact has led to further work with London Fire Service (LFS) to provide accessibility mapping, as a fire-truck cannot be driven through more than 30cm of flood water. The service delivers real time actionable insight around flooding to inform where LFS should position its fire-trucks to enable them to respond to emergency calls within the required 6-minute response time. Previsico is working to implement the property flood resilience code of practice funded by Aviva, Environment Agency, Welsh Government, Department for Infrastructure and Scottish Government; and is supported by Chartered Institution of Water and Environmental Management (CIWEM), Institution of Civil Engineers (ICE) and Royal Institution of Chartered Surveyors (RICE). Previsico’s flood warnings play a vital role in taking action to prevent flood loss.
What were some of the challenges faced and how were they overcome?
Flood forecasting has used a library of events to predict different flood events over different return periods e.g. 1 in 100 years. This approach enabled the Environment Agency to develop a good river and coastal forecast, however surface water proved much more difficult as every storm is different, and with increased urbanisation, predicting surface water flooding – now the UK’s biggest flood risk – is much more complex. The complexity was such that in 2016 the Cabinet Office, EA and Met Office invited Loughborough University’s Professor Yu to develop a high-resolution surface water nowcast. Professor Yu’s challenge was to prove the service's efficacy, after these organisations had been unsuccessful in developing a surface water forecast for many years. This resulted in numerous meetings with various interested parties across national and local government, transportation companies and engineering consultancies, who were all keen to solve this problem. Compounding the issue was the requirement for the solution to integrate with ResilienceDirect, the Cabinet Office’s bespoke portal that was accessed by councils, emergency services, utilities, transportation companies and various government agencies. The service needed to be designed to be plugged into ResilienceDirect and be accessible and usable across this huge range of differing end users.
In your view: Will the change that was created by your initiative continue?
Nineteen years of research has produced a globally proven and unique solution for surface water flooding that is considered to be the 3rd generation in flood forecasting. Published research, most recently in Nature Sustainability, confirms the technology is groundbreaking and that it will lead to future developments that build on this breakthrough innovation. The published research demonstrates the innovation’s value and suggests there is minimal risk of the technology failing to perform. The bigger risk is that other parties may build a better version of this 3rd-generation technology or maybe produce a 4th generation. The risk of the technology being adopted has been minimised by Loughborough spinning out Previsico to commercialise the technology. Previsico spinning out has also been central to helping the change. The excitement in the technology has resulted in Previsico raising £1.2m to build a successful business that can support continued development of the technology.
What did you learn during the initiative or after? And is it possible that others could learn from you?
We initially believed most businesses in flood risk areas would already have plans and resilience measures. However, we realised the need to educate the market on an integrated approach which required combining flood plans, resilience measures and warnings. Individually, each had some benefit but combined they were far more effective. Conducting research, the Cabinet Office, Environment Agency and Met Office taught us the different approach taken by central local government and the emergency services compared to commercial organisations. Even though the Cabinet Office provided the service free-of-charge via their portal, local government and emergency services would have their own preferred systems. Finally, data availability has proved much more challenging than expected. Whilst flood events can create over £1bn of damage there’s little openly available data on which properties flooded, to what depth and level of damage. This is important for Previsico to prove its efficacy in reducing flood loss globally.