Artificial Intelligence in Geosciences

With its successful application in diverse disciplines, Artificial Intelligence (AI), especially Machine Learning, has recently also become more and more popular in geosciences, particularly in the advent of increasingly affordable high performance computational power and the open source mentality of the AI community. The applications range from spatial classification and regression tasks, time-series forecasting, image/signal processing, anomaly detection, event classification, bias correction, and many others, and include for example artificial neural networks, support vector machines, self-organizing maps, decision trees, random forests, or genetic programming. Particular topics of interest of this session include but are not limited to: 1) Novel application examples of AI models in geosciences, e.g. intelligent reservoir characterization, seismic and remote sensing data processing, well logging, rainfall-runoff modelling, forecasts of environmental time series etc., 2) novel theory or advanced model types for geoscience applications, including integrating physical constraints, 3) comparison between AI/ML algorithms and that of physically-based or statistically-based models, 4) explainable AI, and 5) machine learning-assisted data analytics.