Coastal REsistance: Alerts and Monitoring Technologies (CreamT)

Lead Research Organisation: National Oceanography Centre (WEF011019)
Department Name: Science and Technology

Abstract

A 1 m sea level rise is almost certain in the next century and it is estimated that 20% of England's coastal defences could fail under just half this rise. Ambitious climate mitigation and adaptation plans may protect 400,000 - 500,000 people, but flood and coastal erosion risks cannot be fully eliminated - we cannot build infinitely high sea walls. Worldwide 150 million people could be affected by sea level rise in the next 30 years. Better ways to measure, forecast, warn of and respond to coastal flooding are thus required. Using Penzance and Dawlish we will demonstrate a new monitoring system able to issue vital real-time hazard alerts and flood data to national government agencies.
Working with the Environment Agency (EA), Met. Office, Channel Coastal Observatory (CCO), Cornwall Council, Teignbridge District Council, Capgenimi and National Trust, we will build on previous research using digital communication, data networking and citizen science. Our recent project (WireWall) created a unique overtopping sensor that we will develop into a low-cost hazard monitoring system for long-term deployments using telemetry to transfer data. Another project (SWEEP) created a south west regional computer simulation that updates daily to forecast coastal hazard 3 days in advance. The CCO hosts both projects online alongside the Regional Coastal Monitoring Programmes (RCMP) across England. This project will incorporate our new hazard data into the SWEEP service through a new web-accessible, open source data staging web service, thus linking models and new monitoring to validate current hazard services. The new web service will expose existing, coastal, river and weather data, while the new system will include: 1) a novel wave overtopping sensor to measure water levels and waves just before they impact a sea wall in addition to the depth, volume and speed of the water as it overtops onto public access areas behind the sea defence; 2) cameras to validate wave conditions and confirm the occurrence of overtopping events; 3) laser measurements of the pre- and post-storm beach levels during an event; and 4) an international citizen science programme, CoastSnap, that monitors beach conditions over time through photographs. The system will use the UK's tide gauge network to trigger the measurement of potentially hazardous conditions when water levels reach the sea walls and return real-time alerts when flooding is detected. This information will allow validation of the SWEEP computer alert service. With the EA's flood forecast team we will use this information to refine their local hazard thresholds and to understand the uncertainty in local conditions at the sea wall sites due to their large (many km's) distance from national monitoring stations.
The measured, visual and audio data will be used in an interactive coastal walk, and made accessible through an Augmented Reality (AR) phone application, available for IOS and Android devices. The AR walk will guide people to CoastSnap photo posts, encouraging participation in the RCMP beach monitoring. Promotion of the walk through the Tourist Information Centres and Twitter will raise community awareness of changing coastal hazards and shoreline management initiatives such as #floodaware and #CoastSafe.
The team of oceanographers, engineers, data managers, a digital artist, a poet and a software developer will apply their expertise in different disciplines to significantly improve the accuracy and effectiveness of existing coastal hazard warning services. They will engage the public through an easily accessible phone app and participation in citizen science monitoring. Information will be archived at BODC and made available under the NERC Data Policy. This online catalogue is designed to be easily found by the Google dataset search engine and ensures our data are FAIR (Findable, Accessible, Interpretable and Re-usable).
Keywords: Hazard monitoring; Coastal forecasting; Flood aware; Hazard warning

Planned Impact

This project will demonstrate a new flood hazard nowcasting and alert system that has the potential to form a digital real-time hazard monitoring network. We will combine digital technology from the science and arts to collect, communicate about, visualise and interrogate the data that supports national flood and erosion risk management activities undertaken by stakeholders, e.g. local authorities (LAs) and government agencies such as the Environment Agency (EA). The proposal has been designed with input from our existing Wider Interest Group that includes members from academia, industry, business, charities, local authorities and government agencies. The more than 100-member strong group includes representatives from Australia, Netherlands, Norway and the USA.
The main outcomes of the project are listed below, along with the associated benefits to stakeholders. The more immediate benefits are listed first.
1) a robust real-time flood and erosion hazard monitoring system to provide crucial, but currently absent, data on local conditions and resulting hazards at the point-of-impact. These data will: (a) Allow real-time hazard alerts based on point-of-impact observations. (b) Provide validation/calibration information for existing EA flood hazard forecasts and alert thresholds. (c) Considerably reduce the current uncertainty in hazard forecasting, reducing or eliminating missed and false alerts. (d) Improve the publics' trust in hazard warnings. (e) In the longer term, allow much more cost-effective targeting of funds to the most vulnerable sea defences.
2) a new web service that will: (a) Aggregate existing data from the currently disparate coastal, river and meteorological national monitoring services into a single source, allowing, e.g. flood hazards from both coastal and riverine sources to be easily assessed jointly for the first time. (b) Integrate our new point-of-impact data with national monitoring data collected by, e.g. the EA and Regional Coastal Monitoring Programmes (RCMPs). This will, for the first time, allow LAs to fully assess the source-pathway-receptor risk management tool used in their cost/benefit analyses. (c) Incorporate and expose the hazard alerts to satellite navigation systems and online services such as Google Maps or Google Alerts (post-project). The near-universal use of such services will mean comprehensive public awareness of the hazard warnings.
3) an augmented reality (AR) phone application (app) that will guide users along coastal walks at our 2 demonstration sites. Working with the Tourist Information Centres and their wider contacts, the apps and walks will immediately begin to: (a) Communicate our stakeholders' messages about local hazards (past, present and future) and their management strategies. (b) Increase public awareness of existing forecast and monitoring services. (c) Build on the popularity of the South West Coastal Path. (d) Lead people to the international citizen science CoastSnap installations.
4) expand the recently-implemented UK CoastSnap beach monitoring initiative. Data from the monitoring will be used to: (a) Provide early warning of beach erosion that threatens the performance and integrity of local sea defences. (b) Improve model forecasts of flood hazards. (c) Ultimately, allow cost-effective targeting of funds. The initiative will be continued post-project by the RCMPs.
As the above data streams grow post-project they will provide training sets for Machine Learning (ML) approaches to predict local impact from nearby national monitoring stations. The efficiency of ML algorithms and easy access to national data through our new web service will form a basis on which to build Artificial Intelligence (AI) systems and phone applications to support a wide range of stakeholders, e.g. 1) the public and emergency responders when navigating through hazardous conditions; and, 2) asset managers to deliver cost-effective, targeted flood defence inspections and maintenance.

Publications

10 25 50