NERSC Staff to Showcase Analytics Expertise at AGU Fall Meeting

December 9, 2018

This year NERSC’s presence at the annual fall meeting of the American Geophysical Union – being held Dec. 10-14 in Washington, D.C. - is larger than ever before, thanks in part to the center's strides in machine- and deep-learning-based data analytics for two fundamental challenges in climate science: pattern recognition for tracking and discovery in large climate datasets, and emulation of complex dynamical processes.

With a total of 15 talks and posters, the NERSC team will cover a wide range of analytics techniques, including machine learning, deep learning, applied topology, computational mechanics, and other novel statistical methods applied to problems in extreme weather and climate science. A significant portion of the work that will be presented at the meeting involves research output from the Big Data Center and NERSC-supported internships. All the graduate students funded by these programs have landed either a talk or a poster or both.

Below is a schedule of the talks and posters being presented by NERSC staff at the AGU meeting. The AGU is offering free live streaming of select sessions of the fall meeting, which runs through Friday, December 14. The required registration is free and grants access to recorded livestreams as well as additional recorded content.

Title	Venue	Lead Author
Towards Exascale Deep Learning for Climate Science (Invited)	IN11A-01, Mon, 12/10, 8:00AM	Prabhat et al.
Lessons learnt from applying Deep Learning to Scientific Problems at NERSC (Invited)	IN12A-07, Mon, 12/10, 11:35AM	Prabhat et al.
Using Deep Learning for probabilistic detection of extreme weather	IN13C-0682, Mon, 12/10, 01:40PM	Ankur Mahesh et al.
Probabilistic detection of extreme weather using deep learning methods (Invited)	U14B-12, Mon, 12/10, 4:34PM	Ankur Mahesh et al.
NEX-AI: A Cloud and HPC Agnostic Framework for Scaling Deep Learning and Machine Learning Applications for Earth Science	IN11A-04, Mon, 12/10, 08:42AM	Sangram Ganguly et al.
Tracking and Predicting Extreme Climate Events	IN13C-0683, Mon, 12/10, 01:40PM	Soo Kyung Kim et al.
Deep Learning recognizes weather and climate patterns (Invited)	IN14A-07, Mon, 12/10, 05:30PM	Karthik Kashinath et al.
Physics-informed Generative Learning to Emulate Unresolved Physics in Climate Models	IN21C-0722, Tue, 12/11, 08:00AM	Jinlong Wu et al.
Towards Generative Deep Learning Emulators for Fast Hydroclimate simulations	IN21C-0723, Tue, 12/11, 08:00AM	Adrian Albert et al.
Climate Science, Deep Learning, and Pattern Discovery: The Madden-Julian Oscillation as a Test Case	IN21D-0738, Tue, 12/11, 08:00AM	Benjamin Toms et al.
Project DisCo: Physics-based Discovery of Coherent Structures in Spatiotemporal Systems	IN24A-03, Tue, 12/11, 04:30PM	Adam Rupe et al.
Towards Topological Pattern Detection Methods in Climate Science: Application to Atmospheric Blocking Events	IN24A-07, Tue, 12/11, 05:30PM	Grzegorz Muszynski et al.
Deep Learning on the Sphere: Convolutional Neural Networks on Unstructured Meshes	IN33A-01, Wed, 12/12, 01:40PM	Chiyu Jiang et al.
Using a statistical tropical cyclone genesis model for assessing differences in climate scenarios and geographic basins	A43Q-3380, Thu, 12/13, 01:40PM	Arturo Fernandez et al.
ClimateNet: bringing the power of Deep Learning to the climate community via open datasets and architectures	ED53E-0758, Fri, 12/14, 01:40PM	Prabhat et al.

These presentations highlight how NERSC is pushing the boundaries of advanced data analytics and bringing cutting-edge climate informatics solutions to the most pressing problems in climate science today.

About NERSC and Berkeley Lab
The National Energy Research Scientific Computing Center (NERSC) is a U.S. Department of Energy Office of Science User Facility that serves as the primary high performance computing center for scientific research sponsored by the Office of Science. Located at Lawrence Berkeley National Laboratory, NERSC serves almost 10,000 scientists at national laboratories and universities researching a wide range of problems in climate, fusion energy, materials science, physics, chemistry, computational biology, and other disciplines. Berkeley Lab is a DOE national laboratory located in Berkeley, California. It conducts unclassified scientific research and is managed by the University of California for the U.S. Department of Energy. »Learn more about computing sciences at Berkeley Lab.