More than two thirds of European cities are regularly confronted with climate change-induced disasters. Early warning systems (EWS) play a crucial role in mitigating the effects of such disasters by detecting conditions that forecast the onset of a catastrophe and to compute its impact. EWSs provide alarming, decision support and information services to governments, companies and the general public. UrbanFlood is a European project investigating the use of sensors within flood embankments to support an online early warning system, real-time emergency management and routine asset management. It creates an internet-based hosting platform for EWSs. The EWS platform is run as an internet service and connects via the internet to sensor networks, to online sources of information and other EWSs. Co-ordinated in the Netherlands, the project also involves researchers from Poland, Russia and the United Kingdom. [login type="readmore"] The platform is able to host multiple EWSs, corresponding to various hazards and belonging to different organisations. Through the internet, additional computer resources required by the EWS platform are made available on demand. Artificial intelligence technologies detect alarming conditions in the objects monitored. In an alarming situation, a decision support system, making use of computational models of the physical environmental, informs stakeholders about the developing catastrophe. The core of the EWS hosting platform is a common information space (CIS). The CIS is run as an online service, supports web service technologies and connects via the internet with other CISs. The EWS is also able to process simulated data, turning the EWS into a simulator with which disaster mitigation scenarios can be developed and personnel can be trained. UrbanFlood validates the EWS framework in the context of dyke failures and ensued flooding in an urban environment. Dykes are equipped with sensor systems and the EWS service is built up from a series of dyke failure and flooding specific modules, including 3D/pseudo 4D dyke evolution and flood-spreading models. FLOOD RESEARCH IN PRACTICE “There are 48 sensors constantly telling us how a dam is feeling and how it’s behaving,” said Robert Meijer, a professor of applied sensor networks at the University of Amsterdam and co-ordnator of the Urbanflood project. “This technology makes this the smartest dam in the world.” The underground sensors, developed within the research project, transmit measurements to data centres in several countries. Boston, on the east coast of England, is one of the main test sites for the new technology. Its river rises dramatically with the sea tides, putting local citizens at risk. Mark Robinson, a senior coastal advisor at the UK Environment Agency said: “Boston has some good sea defences, but they’re not high enough. So when the surge comes through, it means the walls can get over-topped, and potentially could fail. So it’s at significant risk.” Engineers involved in the study have designed several types of sensors, reliable and affordable enough to be put in damp soil embankments. Robert Meijer explained: “A sensor is a measuring device that has electronic components inside a stainless steel container. It has a tiny opening through which water can enter. This device measures the amount of water in the dam and many other parameters such as the underground temperature, and the movement of the sensors inside the dam.” These indications can be followed in real time using any internet-enabled device. A touch screen panel installed in a Boston café allows the public to see how the system works. If the artificial intelligence anticipates a breach, it sends a warning to the responsible services that may otherwise be unaware of the weakening within the flood defences. LEAKAGES AND DEFENCE “Currently we use a visual walkover,” said Robinson. “It means probably two or three times a year we got people viewing the asset [the dam] from a visual point of view. This technology allows us to actually understand how the bank is performing, so it looks at how it’s moving under pressure, whether we’ve got leakages through the defence, and also we can see if there’s an incident in real time what’s happening.” The software calculates how fast the site will be flooded if the dam fails, and even suggests the best ways to move citizens into safer areas. Alexandra Topple, a senior flood-management engineer at HR Wallingford, said the system made it possible to show the outline of the flood mapping. “We can then run a life safety model which will show the evacuation process that people can take – this can be used by emergency planners, for them to be able to test various scenarios of evacuation to find the best evacuation procedure in the event of the failure,” he explained. If proved reliable, this system could be deployed in many urban areas in Europe and across the world to protect vulnerable cities from the imminent effects of climate change.  This article is based on materials supplied by the European Union.