Posted by
The world is awash in scientific data these days, but all that information is not being harnessed to its potential to help steer the planet to a more sustainable future.
Microsoft Research Podcast
AI Frontiers: Models and Systems with Ece Kamar
Ece Kamar explores short-term mitigation techniques to make these models viable components of the AI systems that give them purpose and shares the long-term research questions that will help maximize their value.
That’s the message Stephen Emmott (opens in new tab), head of the Computational Science Laboratory (opens in new tab) at Microsoft Research Cambridge (opens in new tab), will deliver Dec. 14 during the Eye on Earth Summit (opens in new tab), being held Dec. 12-15 at the Abu Dhabi National Exhibition Centre.
Emmott and colleague Lucas Joppa (opens in new tab), a scientist in the Computational Ecology and Environmental Science (opens in new tab) team within the Computational Science group, will be previewing prototype models and related tools, which they hope in the future will enable people to find, generate, contribute, and share environmental data, information, understanding, and predictions.
“The future state of the planet and its inhabitants rests on taking the right actions now,” Emmott said, citing the need to reduce and adapt to climate change, to preserve and protect ecosystems, to address risks to food and water security, and to avoid the devastating results of a global pandemic.
“There is an enormous gap between data and understanding and informed action,” Emmott said. “Bridging this gap is an urgent priority.”
Those are the sorts of priorities high-level government elites, including environmental ministers, will bring to the summit, organized by the United Nations Environment Programme and the Environment Agency—Abu Dhabi. Speakers from the areas of philanthropy, business, government, data engineering, and technology will address the conference, a preliminary event in the run-up to Rio+20, one of the most significant sustainable-development events in recent years, to be held in June 2012.
Attendees are expected to declare their commitment to the Eye on Earth (opens in new tab) Initiative, an agreement by stakeholders to make data broadly available for sharing.
Microsoft is working with the European Environmental Agency (opens in new tab) (EEA) and Esri (opens in new tab), a developer of geographic-information systems, to address critical environmental concerns that affect communities, economies, and businesses worldwide. On Dec. 1, the three organizations launched the Eye on Earth Network (opens in new tab), an online community to develop a cloud-based data-sharing network. During the summit, the EEA is expected to announce the expansion of the network as new partners and environmental stakeholders join the platform to store and visualize data sets.
One example of the models and tools to be previewed by Emmott and Joppa during the event is the Global Carbon Climate Model, which tracks publically available data on carbon stocks and flows from locations around the globe.
The model incorporates several tools created by the Computational Science team:
• FetchClimate (opens in new tab): a cloud-based service for automatic environmental-data retrieval.
• Filzbach (opens in new tab): a library for effective parameter estimation.
• Dmitrov (opens in new tab): a library for creating, storing, and manipulating large, multidimensional scientific data sets.
• DataSet Viewer (opens in new tab): a menu-driven tool for exploring time series, geographic distributions, and other data patterns.
“Transforming data into action will require new kinds of science,” Emmott said, “new kinds of tools, new kinds of data and information, and new kinds of predictive models.”