Four European higher education institutions, including the Universidad Politécnica de Madrid’s Facultad de Informática through the Ontological Engineering Group, are developing an early warning system for natural disasters capable of predicting coastal flooding and forest fires.

The ontology-based system combines information taken from different sources and displays a status map for experts. Using this map, experts will be able to make better decisions on which localities to evacuate or what preventive measures to take to avert the disaster or minimize its effects on people and property.

The system is being developed for two particular domains, with potential extensions to other applications in the future. First, the system focuses on the prevention of coastal flooding caused by torrential rain or unexpected sea level rises. Second, it targets forest fire prevention.

Coastal flooding

The coastal flooding application is being developed for southern England, a region that is extremely sensitive to this phenomenon. Even so, it is extendible to any other coastline in the world suffering the effects of climate change.

To develop this part of the application, the system is to be built to use information from a system of alert boys strategically located along the southern English coast. It also employs information gathered through direct observation. Finally, it will utilize meteorological information supplied by the British Government’s Environment Agency.

The system integrates and combines information from all these sources with information provided by Google Maps and Google Earth. It then displays an on-screen and real-time status map to support expert decision making about which houses to evacuate first in a crisis situation or which roads it is best to use for transporting people and property in a state of emergency. This takes crisis management to a level of sophistication that would be impossible without the technologies used that also enable experts to check out new hypotheses in real time.

Forest fires

In the case of forest fires, research has focused on the Spanish experience in forestry. In this case, the system uses information from two sources: images offered by Deimos Space, the Spanish company that put the first private Spanish Earth observation satellite into orbit last July, and the information provided by sensor networks across the forest area.

The system integrates land information (provided by sensors) with satellite information. By combining the two sources of information with the information available in applications like Google Earth or supplied by the Spanish National Geographical Institute’s cartographic services, the system will give experts access to a status map of the forest area, visualizing the areas of greatest tree density, humidity concentrations and air temperature in real time. This way, it will be possible to establish, minute by minute, what part of a particular forest zone is at greatest risk from fire.

The system is not designed to detect fires started on purpose, but it does offer technicians a simple query language for accessing remote sensors and modifying their programming. This way, they can, depending on the detected risk, change the frequency or characteristics of the information issued, thereby fine tuning the prognosis of a possible forest fire outbreak.

SemsorGrid4Env: a Seventh Framework Programme

The project is called SemsorGrid4Env and is part of the European Union Seventh Framework Programme. It kicked off in 2008 and is due to end in 2011. The first phase is now complete, having identified the functions to be provided by each system. The second phase, scheduled for 2010, is to create prototype applications. Finally, in the third stage, developed in 2011, different user groups will test the application.The ultimate aim is to commercialize the developed application for use by institutions all over the world that are responsible for managing natural disasters.

The project coordinators are Asunción Gómez Pérez and Oscar Corcho, from the Ontological Engineering Group. The other SemsorGrid4Env project partners are the Universities of Manchester and Southampton (in the United Kingdom), and the National and Kapodistrian University of Athens (Greece). Another Spanish company, TechIdeas, and Emu Ltd., from the United Kingdom, join Deimos Space as the project’s industrial partners.