Prediction software helps in disaster control
EU-funded scientists have developed new statistical methods and software to help generate information on pollution and environmental threats. The system complements existing European-based air quality monitoring stations, which are effective but costly to set up. The INTAMAP ('Interoperability and automated mapping') project was funded EUR 1.86 million under the 'Information society technologies' (IST) Thematic area of the EU's Sixth Framework Programme (FP6). Air pollution monitors are used across Europe to generate information on nuclear and chemical spills, groundwater pollution, traffic fumes and other environmental issues, but their expense hinders mass installation. Data generated at the relatively few location monitoring points is, nevertheless, fundamental to the subsequent decisions made by authorities on the right way of dealing with these problems. Another disadvantage is that they provide considerable information about pollution in their immediate surrounds, but little about the status of conditions beyond their vicinity. In the event of a nuclear accident, for instance, one monitoring station might report high levels of radioactivity but another, 30 kilometres away, might report significantly lower levels. In the case of a major environmental crisis, authorities would need to decide if an evacuation is necessary and, if so, to select the most appropriate place to which the population should be taken. It is these types of scenarios and the need to make better use of existing environmental data that keep European researchers on the lookout for new ways of piecing together point measurement sets to generate information that is representative of a larger expanse of terrain. Under the INTAMAP project, Dr Edzer Pebesma of the University of Münster in Germany and his colleagues spent three years developing new statistical methods and software for interpolating environmental data. The term 'interpolation' is used by statisticians to define the process of locating the value of an environmental variable at a point on a map where there is no monitoring instrument. The challenge is to create a contour map based on a fixed number of measurements taken from specific areas and to rate the accuracy of these contours. The approach taken by the INTAMAP researchers was to make the technology easy to use and to focus on real-world issues. For example, if an issue requires urgent action, the software knows to use less sophisticated but faster models to produce a calculation. If time is not as critical, the software can then use the most accurate interpolation techniques to generate information. The technology also accepts raw data published on the Internet using standards developed by the Open Geospatial Consortium (OGC). Once this data has been entered, services that conform to OGC standards can generate and update maps as required. In order to rate accuracy and convey this information via the web, the team also created UncertML, a new dialect of the XML document markup language. The INTAMAP system is already being used in areas where the need to generate predictions are essential to decision-making. For instance, the German radiation protection authority Bundesamt für Strahlenschutz is using the technology to produce hourly readings of gamma radiation provided by the European Radiological Data Exchange Platform (EURDEP). Dr Pebesma added that the tools could be used in other industries, such as medicine, where a two-dimensional image needs to be generated from a series of point readings. In fact, the UK's Aston University (an INTAMAP partner), has already developed an application for mobile phones that maps temperature based on information from home weather stations. Research for INTAMAP was undertaken by stakeholders from Belgium, Germany, Greece, the Netherlands, Austria and the UK.