The 2016 European Geosciences Union General Assembly saw us co-convene our second PICO (Presenting Interactive Content) session on operational forecasting and warning systems for natural hazards. Along with several other sessions in hydrological forecasting there were plenty of take home points regarding the challenges and innovation of natural hazard early warning.
There is a growing effort in the area of real-time mapping of the consequences of hazards to support emergency response. In the Philippines, Project NOAH is responsible for the development of a storm surge advisory service with the purpose of warning for expected inundation of storm surges using hazard mapping (Santiago) – see image. A similar concept being adopted in the Chapere watershed in Bolivia involves river forecasts linked to real-time inundation scenarios to allow civil protection authorities to estimate the impact on the population (Rossi). For approaches in forecasting of surface water flooding in the UK the hazard footprint is combined with an indication of the likelihood to provide an overall indication of risk (Cole) – as seen with our approach to the Commonwealth Games pilot – which this is now being adopted on a countrywide scale in England and Wales.
A method of staged decision-making based on probabilistic forecasting was presented highlighting that a cost-loss approach was possible (Booister). The work, partly based on research in Scotland with the flood forecasting service, opens up possibilities for more flexible decision making compared to more traditional deterministic approaches.
The use of crowd sourced information to improve predictions is another area of innovation. As reported last year initiatives such as Floodtags or WeSenseIt have provided excellent examples of such approaches. This year an approach to relying on indigenous knowledge for flood and drought disaster forecasting in Ghana was presented, with the system using local co-ordination linked with satellite-based warnings (Jungermann). Coupling flood forecasting and social media was also covered in the context of the European Flood Awareness System (EFAS) and the use of the monitoring application Global Flood News (Kalas).
Challenges still remain with flash flood prediction although some innovative approaches are being trialled. On a global-scale developers of the Global Flood Awareness System (GloFAS) are setting up a system using 4-day ensemble surface runoff forecasts with a flash flood index based on climatology (Baugh). On a regional-scale high resolution convective-permitting quantitative precipitation forecasts are being used to support a new Vigicrues flash flood service in France (Dermargne).
The emergence of multi hazard platforms are proving beneficial to responders with one such case in Southeastern France – RHYTMME – providing a warning and mapping system for floods, landslides and debris flows (Fouchier). Risk management strategies are now supporting multi-hazard approaches and are particularly beneficial where hazards interact both spatially and temporally (Gill) as seen with Storms Desmond (UK) and Synne (Norway) with landslide and flooding impacts being supported through national hazard assessments (Boje).
This is a unique international forum for sharing ideas on the prediction and early warning of natural hazards and we look forward to further sharing of research and operational system development at the EGU in 2017.