KMi has begun its research activity on ClimatePrediction.Net, a massive distributed computing project designed to allow up to 2 million individuals each to run part of a 100-year-projection climate model on their personal computers. ClimatePredicton.Net, headed by a team from the Atmospheric Physics Department at the University of Oxford, has been awarded £400K by the Natural Environment Research Council. Prof. Bob Spicer in the OUs Department of Earth Sciences will be in charge of the production of associated teaching materials. KMi was invited to join the project to help build a strong and sustainable web-based community of users.
The climate models require a personal commitment by users to allow the programs to run intermittently in the background on their PCs over a period of several months. During this period, it will be necessary to keep individual users motivated and informed about the progress of the computer models, which is where KMi comes in. A mixture of existing and newly-emerging KMi technologies will be deployed, including integrated discussion forums based on D3E, web-based news based on KMi Planet and Rostra, conferencing services based on KMi Stadium and WebSymposia, semantic web services based on KMis ontology research, and presence awareness/messaging technologies based on KMis BuddySpace.
The semantic web services will constitute a semantic portal, whereby users can view any web page through an intelligent filter that highlights terms of interest to climate modellers (for instance). Every highlighted term then becomes a jumping off point not merely to another web page, but instead to value-added services that depend entirely on the meaning of the highlighted terms and the context in which they are deployed. This capability is being piloted in KMis semantic tool called Magpie, which will form part of the growing semantic climate portal.
KMi’s full-time Research Fellow on ClimatePrediction.Net is Martin Dzbor, working with Principal Investigator and grantholder Enrico Motta, along with KMi’s John Domingue and Marc Eisenstadt.