Climate change, air pollution, natural disasters and availability of resources are just a sampling of issues that humanity faces, and geoscientists are working to understand and model. The field of big data and machine learning offers many opportunities for tackling problems like these. However, geosciences – unlike some commercial sectors, which have seen success with machine learning applications – may introduce challenges that the field of machine learning has yet to encounter.
New research published in IEEE Transactions on Knowledge and Data Engineering from geoscientists including Sai Ravela, principal research scientist in the MIT Department of Earth, Atmospheric and Planetary Sciences (EAPS), enumerates “the common categories of geoscience problems where machine learning can play a role, discussing the challenges faced by existing machine learning methods and opportunities for novel machine learning research.” The authors conclude with a discussion of how advances in the fields of machine learning and geosciences can inform and improve each other.
Advances in artificial intelligence are needed to collect data where and when it matters, to integrate isolated observations into broader studies to create models in the absence of comprehensive data, and to synthesize models from multiple disciplines and scales.
Intelligent systems need to incorporate extensie knowledge about the physical, geological, chemical biological, ecological, and anthropomorphic factors that affect the Earth system while leveraging recent advances in data-driven research.
A new generation of knowledge-rich intelligent systems have the potential to significantly transform geosciences research practices.