Title: A New Look at Climate Diagnosis and Modeling in the Era of Climate Informatics
Update (6-16-2015): Abstract submission is NOW OPEN for this session!
Submission deadline: August 5, 2015.
Official weblink at AGU: https://agu.confex.com/agu/fm15/preliminaryview.cgi/Session7924
Yi Deng - Earth and Atmospheric Sciences, Georgia Tech
Imme Ebert-Uphoff - Electrical and Computer Engineering, Colorado State University
The size and complexity of observational and model-simulated climate data have seen accelerated growth since the late 1970s. This increasing amount of data and our growing computational capacity create unprecedented opportunities for bringing innovative approaches of machine learning and data mining to climate data for interdisciplinary knowledge discovery, thus the birth of a new area “climate informatics”. This session seeks contributions from all application areas with the goal of improving process-level understanding and modeling of the Earth’s coupled climate system through advanced data mining and machine learning methods. These include but are not restricted to the development and implementation of new data mining methods for climate diagnosis and atmospheric process study, new ideas of data assimilation, stochastic climate and environment modeling, use of causal discovery and structure learning methods to understand large-scale dynamical processes, uncertainty quantification in climate simulation and projection, and data-driven approaches in weather forecasting and climate prediction.