OS019
Innovation and Exploration with Machine Learning in Ocean and Atmospheric Sciences: Global and Regional Applications

Session ID#: 85463

Session Description:
As machine learning methods mature, many possibilities open in Ocean and Atmospheric Sciences for describing and understanding large data sets. Increasing volumes of data are becoming available that allow novel exploration of complex phenomena. From characterizing global dynamical regimes, improving subgridscale parameterizations and leveraging robotic technology such as drones and remote sensing, machine learning promises innovation. As a tool, machine learning can play a pivotal role as the link between theory, modelling and observational efforts. This session invites submissions that demonstrate progress in understanding both the uses and misuses of machine learning in terms of supervised, unsupervised, and active learning, as well as visual analytics. Submissions are welcomed from regional and global applications to ocean and atmospheric sciences.
Co-Organized:
  • A - Atmospheric Sciences
Index Terms:

0555 Neural networks, fuzzy logic, machine learning [COMPUTATIONAL GEOPHYSICS]
1640 Remote sensing [GLOBAL CHANGE]
3337 Global climate models [ATMOSPHERIC PROCESSES]
4299 General or miscellaneous [OCEANOGRAPHY: GENERAL]

Primary Convener:  Maike Sonnewald, Harvard University, Earth and Planetary Sciences, Cambridge, MA, United States
Conveners:  Redouane Lguensat, Institut des Géosciences de l'Environnement, CNRS/Univ. Grenoble Alpes/G-INP/IRD, Grenoble, France, Pierre Gentine, Department of Earth and Environmental Engineering, Columbia University, New York, NY, United States and Patrick Clifton Gray, Duke University, Duke University Marine Lab, Beaufort, NC, United States
Primary Liaison:  Maike Sonnewald, Harvard University, Earth and Planetary Sciences, Cambridge, MA, United States
 
Machine learning submesoscale dynamics of Lagrangian ocean drifters (488929)
Nikolas Olson Aksamit1, Themistoklis Sapsis2 and George Haller1, (1)ETH Swiss Federal Institute of Technology Zürich, Institute for Mechanical Systems, Zürich, Switzerland, (2)Massachusetts Institute of Technology, Department of Mechanical Engineering, Cambridge, MA, United States
 
Discovering Novel Eddy Parameterisations with Machine Learning (Invited) (489286)
Laure Zanna, University of Oxford, Dept of Physics, Oxford, United Kingdom and Thomas Bolton, University of Oxford, Department of Physics, Oxford, United Kingdom
 
Short-term predictions of GPM satellite rainfall data using a deep learning neural network model (493987)
Carlos H Ribeiro Lima, UNB University of Brasilia, Asa Norte, Brazil, Hyun-Han Kwon, Sejong University, Soul, South Korea and Upmanu Lall, Columbia University, New York, NY, United States
 
Object-based processing of high-resolution precipitation forecasts : from detection to practical applications (505757)
Laure Raynaud, Meteo-France CNRM/GMAP, Toulouse, Occitanie, France
 
Deep Learning Approaches for Visibility Classification Using Traffic Cameras (514731)
Giuliano Andrea Pagani1, Wiel Wauben2 and Jan Willem Noteboom2, (1)Royal Netherlands Meteorological Institute, De Bilt, 3730, Netherlands, (2)Royal Netherlands Meteorological Institute, De Bilt, Netherlands
 
Convolutional Neural Networks for Estimating Atmospherically-forced Sea Level Variations (532865)
Redouane Lguensat, Mickael Lalande, Sally E Close and Thierry Penduff, Institut des Géosciences de l'Environnement, CNRS/Univ. Grenoble Alpes/G-INP/IRD, Grenoble, France
 
Development of a Deep Learning Super-Resolution Generative Adversarial Network (SGAN) for Improving the Resolution of Global Precipitation Measurement (GPM) Dual-Frequency Precipitation Radar (DPR) Data (552769)
John Malone Beck1, Todd Berendes1, Charles Collins1, Sean M. Pagani2 and Matthew James Daigle2, (1)University of Alabama in Huntsville, Huntsville, AL, United States, (2)University of Alabama in Huntsville, ITSC, Huntsville, AL, United States
 
Supervised and Unsupervised Machine Learning tools for uncovering climatological feedback patterns in black carbon emissions. (612563)
Mads Jacob Kalisz Hedegaard, Independent Scholar, Bremen, Germany and Anna Beata Kalisz Hedegaard, University of Bremen, Bremen, Germany; German Aerospace Center DLR Oberpfaffenhofen, Oberpfaffenhofen, Germany
 
Beyond Aquaplanet: Is Deep Learning Emulation of Global Superparameterization Viable in Realistic Settings? (619144)
Griffin Mooers, University of California Irvine, Department of Earth System Science, Irvine, CA, United States, Mike S Pritchard, University California Irvine, Department of Earth System Science, Irvine, CA, United States, Tom Beucler, Massachusetts Institute of Technology, Cambridge, MA, United States and Stephan Rasp, Technical University of Munich, Munich, Germany
 
Machine Learning to Improve Satellite Observations of Orographic Winds (631281)
Thomas Kilpatrick1, William Chapman1, Bruce D Cornuelle2 and Shang-Ping Xie1, (1)Scripps Institution of Oceanography, La Jolla, CA, United States, (2)University of California San Diego, La Jolla, CA, United States
 
Estimation of nighttime cloud properties using machine learning (632540)
Andrew M Snauffer, Northeastern University, Departments of Civil and Environmental Engineering and Marine and Environmental Sciences, Boston, MA, United States, Loren Oh, University of British Columbia, Earth, Ocean and Atmospheric Sciences, Vancouver, BC, Canada and Philip H Austin, Univ of British Columbia, Vancouver, BC, Canada
See more of: Ocean Sciences
<< Previous Session Proposal | Next Session Proposal >>