Machine learning for data-driven discovery in solid Earth geoscience

• Published in: Science (American Association for the Advancement of Science) Vol. 363; no. 6433; p. 1299
• Main Authors: Bergen, Karianne J., Johnson, Paul A., de Hoop, Maarten V., Beroza, Gregory C.
• Format: Journal Article
• Language: English
• Published: United States American Association for the Advancement of Science 22.03.2019; The American Association for the Advancement of Science; American Association for the Advancement of Science (AAAS)
• Subjects: Acoustic emission; Algorithms; Artificial intelligence; Atmospheric models; Automation; Benchmarks; Biological activity; Classification; Communities; Computer memory; Computer simulation; Computers; Data processing; Data sources; Datasets; Driving ability; Earth; Earth science; Earthquakes; Fiber optics; Fresh water; Freshwater environments; Geologic mapping; Geological hazards; Groundwater; Groundwater reservoirs; Inverse problems; Learning algorithms; Lidar; Machine learning; Mathematical models; Mineral resources; Oceans; Open data; Optical fibers; Organic chemistry; Predictions; Remote sensing; REVIEW SUMMARY; Rheology; Seismic activity; Teaching Methods; Topology; Urbanization; Viscoelasticity; Volcanic ash
• ISSN: 0036-8075; 1095-9203; 1095-9203
• DOI: 10.1126/science.aau0323

Solid Earth geoscience is a field that has very large set of observations, which are ideal for analysis with machine-learning methods. Bergen et al. review how these methods can be applied to solid Earth datasets. Adopting machine-learning techniques is important for extracting information and for understanding the increasing amount of complex data collected in the geosciences. Science , this issue p. eaau0323 Understanding the behavior of Earth through the diverse fields of the solid Earth geosciences is an increasingly important task. It is made challenging by the complex, interacting, and multiscale processes needed to understand Earth’s behavior and by the inaccessibility of nearly all of Earth’s subsurface to direct observation. Substantial increases in data availability and in the increasingly realistic character of computer simulations hold promise for accelerating progress, but developing a deeper understanding based on these capabilities is itself challenging. Machine learning will play a key role in this effort. We review the state of the field and make recommendations for how progress might be broadened and accelerated.