T31E:
Geospatial Artificial Intelligence: Machine Learning in Earth Sciences Posters


Submit an Abstract to this Session


Session ID#: 50531

Session Description:
Amount and variety of data created in the process of measuring and modeling Earth's sub-systems is drastically increasing with ever-improving computational and data acquisition capabilities. Gaining insight into a wide array of Earth processes from diverse data sources that are vast in size and variety requires new methods. Machine learning (ML) methods have enabled the analysis of big data in geosciences that would not have been possible otherwise, and have allowed new insights from data used in traditional analysis. The ability to integrate big data into scientific workflows and emerging new ML methods has opened new avenues of research and may fill in the gaps where traditional approaches fall short.  

This session seeks submissions that:

  1. Demonstrate application of ML methods, including deep learning, to understand Earth'€™s subsystems in ways that were not possible with traditional approaches
  2. Present novel machine learning methods that can replace or complement traditional methods
Primary Convener:  Orhun Aydin, ESRI, Redlands, CA, United States
Conveners:  Emily Law, CalTech JPL, Pasadena, CA, United States, Warren T Wood, US Naval Research Laboratory, Ocean Sciences Division, Washington, DC, United States and Jens F Klump, CSIRO Mineral Resources, Perth, Australia
Primary Liaison:  Warren T Wood, US Naval Research Laboratory, Ocean Sciences Division, Washington, DC, United States
Chairs:  Orhun Aydin, ESRI, Redlands, CA, United States and Emily Law, CalTech JPL, Pasadena, CA, United States
OSPA Liaison:  Warren T Wood, US Naval Research Laboratory, Ocean Sciences Division, Washington, DC, United States
Co-Organized with:
Tectonophysics, Earth and Space Science Informatics, Global Environmental Change, and Ocean Sciences

Cross-Listed:
  • GC - Global Environmental Change
  • IN - Earth and Space Science Informatics
  • OS - Ocean Sciences

Proposed Co-Organized Session with:
  • GC - Global Environmental Change
  • IN - Earth and Space Science Informatics
  • OS - Ocean Sciences
Index Terms:

1699 General or miscellaneous [GLOBAL CHANGE]
3099 General or miscellaneous [MARINE GEOLOGY AND GEOPHYSICS]
3399 General or miscellaneous [ATMOSPHERIC PROCESSES]
8099 General or miscellaneous [STRUCTURAL GEOLOGY]

Abstracts Submitted to this Session:

Using Spatially Constrained Unsupervised Machine Learning to Objectively Derive Climate Zones of the Contiguous United States (Invited) (427469)
Kevin A. Butler, Environmental Systems Research Institute, Redlands, CA, United States
Long-lead Forecast of Heatwaves in the Eastern United States using Artificial Intelligence (466413)
Negin Sobhani1, Davide Del Vento1 and Alessandro Fanfarillo2, (1)National Center for Atmospheric Research, Boulder, CO, United States, (2)University Corporation for Atmospheric Research, Boulder, CO, United States
Machine Learning for Global Precipitation Measurement (GPM) Dual-Frequency Precipitation Radar (DPR) Resolution Improvement (452696)
John Malone Beck1, Sara J Graves2, Charles Collins1, Susan Bridges1 and Stephanie Mullins Wingo3, (1)University of Alabama in Huntsville, Huntsville, AL, United States, (2)Univ. of Alabama/Huntsville, Huntsville, United States, (3)NASA IMPACT / UAH, Huntsville, United States
Prediction of CO2 flux using Long Short Term Memory (LSTM) Recurrent Neural Networks with data from Flux towers and OCO-2 remote sensing (403008)
Phuong Nguyen and Milton Halem, University of Maryland Baltimore County, Computer Science, Baltimore, MD, United States
Cloud Removal in Hyperspectral Satellite Images Using Deep Learning (430176)
Yi-Lin Tsai1, Laura Domine2, Abhijeet Phatak3, Christopher B Field4, Katharine J Mach5 and Peter K Kitanidis1, (1)Stanford University, Department of Civil and Environmental Engineering, Stanford, CA, United States, (2)Stanford University, Department of Physics, Stanford, CA, United States, (3)Stanford University, Department of Electrical Engineering, Stanford, CA, United States, (4)Stanford University, Stanford Woods Institute for the Environment, Stanford, United States, (5)Stanford University, Stanford Environment Assessment Facility, Stanford, CA, United States
Development of a Global Predictive Seabed Model (GPSM) (343823)
Warren T Wood1, Benjamin J Phrampus2, Taylor Runyan Lee2 and Jeffrey Obelcz3, (1)US Naval Research Laboratory, Geology and Geophysics, Washington, DC, United States, (2)US Naval Research Laboratory, Ocean Sciences Division, Washington, DC, United States, (3)US Naval Research Laboratory Stennis Space Center, Marine Geology and Geophysics, Stennis Space Center, United States
Machine-learning – converting geoscience data into predictive geochemical and 3D surface models. (459966)
John Wilford1, Karol Czarnota1, Sudipta Basak1, Lachlan Lachlan.Mccalman2, Daniel Steinberg2, Niket Chhajed1 and Rakib Hassan1, (1)Geoscience Australia, Canberra, ACT, Australia, (2)CSIRO, Canberra, ACT, Australia
Using machine learning to classify landforms for minerals exploration (427843)
Jens F Klump1, Thomas Albrecht2, Ignacio González-Álvarez2 and Gregory Smith3, (1)CSIRO Mineral Resources, Perth, Australia, (2)CSIRO Mineral Resources, Perth, WA, Australia, (3)CSIRO Data61, Hobart, TAS, Australia
Continuous monitoring of volcano-seismic signals with Recurrent Neural Networks (373201)
Manuel Titos Luzón1, Angel Bueno Rodriguez1, Luz Garcia Martinez1, Isaac Álvarez Sr.2, Carmen Benitez1 and Jesús M Ibáñez3, (1)University of Granada, Signal Theory, Telematic and Communications, Granada, Spain, (2)University of Granada, Signal Theory, Telematics and Communications, Granada, Spain, (3)University of Granada, Instituto Andaluz de Geofísica, Granada, Spain
Deep Learning of Geological Structures in 3-D Seismic Data (358981)
Thilo Wrona1, Indranil Pan2, Rebecca E Bell3, Haakon Fossen1 and Robert Gawthorpe1, (1)University of Bergen, Department of Earth Science, Bergen, Norway, (2)Imperial College London, Centre for Process Systems Engineering & Centre for Environmental Policy, London, United Kingdom, (3)Imperial College London, Department of Earth Science and Engineering, London, United Kingdom
Assign Points To Existing Clusters: An algorithm for labeling clustered point cloud datasets for predicting physical variables (448479)
Sean M Hendryx, University of Arizona, School of Information, Tucson, AZ, United States

See more of: Tectonophysics