Earth science

IBM and NASA will use artificial intelligence to investigate climate change

Earth science

The technology giant IBM and the US space agency, NASA, teamed up to investigate climate change through the use of artificial intelligence (AI). With the technology of foundational models they will seek to collect and analyze data regarding this topic.

To understand this alliance, in a joint communication from IBM and NASA it is explained that foundational models are “types of AI models that are trained on a large set of unlabeled data, that can be used for different tasks and allow information to be applied about one situation to another.

As explained, in recent years the volume of information resulting from Earth observation that has been collected has increased.

Faced with this vastness of data, IBM and NASA created this alliance to “provide researchers with an easier way to analyze and obtain insights from these large data sets. IBM’s foundational model technology has the potential to accelerate the discovery and analysis of this data to rapidly advance scientific understanding of Earth.”

Raghu Ganti, Principal Investigator at IBM, assured that “the application of foundational models will allow very valuable knowledge and information to be available to a much broader group of researchers, companies and citizens.”

The results expected by IBM and NASA

Both organizations explain that the expected outcome of the alliance is the compilation of an “easily searchable corpus of Earth Science literature.”

For this, IBM developed a natural language processing model trained on nearly 300,000 Earth science journal articles to organize the literature and facilitate the discovery of new knowledge.

“The beauty of foundational models is that they can potentially be used for many downstream applications,” explained Rahul Ramachandran, a senior researcher at NASA’s Marshall Space Flight Center in Huntsville, Alabama. With this possibility, IBM and NASA also put on the table the possibility of creating a foundational model for weather and climate prediction.