Open AccessA new international initiative for facilitating data-driven Earth science transformation
Author(s) -
Qiuming Cheng,
Roland Oberhänsli,
Molei Zhao
Publication year - 2020
Publication title -
special publication - geological society of london/geological society, london, special publications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.673
H-Index - 132
eISSN - 2041-4927
pISSN - 0305-8719
DOI - 10.1144/sp499-2019-158
Subject(s) - transformation (genetics) , earth (classical element) , earth system science , earth observation , earth science , political science , engineering ethics , geology , engineering , aerospace engineering , mathematics , oceanography , chemistry , satellite , biochemistry , mathematical physics , gene
Data-driven techniques including machine-learning (ML) algorithms withbig data are re-activating and re-empowering research in traditionaldisciplines for solving new problems. For geoscientists, however, whatmatters is what we do with the data rather than the amount of it. Whilerecent monitoring data will help risk and resource assessment, thelong-earth record is fundamental for understanding processes. Thus, how bigdata technologies can facilitate geoscience research is a fundamentalquestion for most organizations and geoscientists. A quick answer is thatbig data technology may fundamentally change the direction of geoscienceresearch. In view of the challenges faced by governments and professionalorganizations in contributing to the transformation of Earth science in thebig data era, the International Union of Geological Sciences has establisheda new initiative: the IUGS-recognized Big Science Program. This paperelaborates on the main opportunities and benefits of utilizing data-drivenapproaches in geosciences and the challenges in facilitating data-drivenearth science transformation. The main benefits may include transformationfrom human learning alone to integration of human learning and AI, includingML, as well as from known questions seeking answers to formulating as-yetunknown questions with unknown answers. The key challenges may be associatedwith intelligent acquisition of massive, heterogeneous data and automatedcomprehensive data discovery for complex Earth problem solving.