H094
Scientific Machine Learning for Flow, Transport, and Coupled Processes
Session ID#: 160373
Session Description:
Machine/deep learning (ML/DL) methods have been greatly advanced to efficiently extract information, discover knowledge, and identify correlations and patterns. Recent efforts to develop and apply ML/DL driven models for problems in flow, transport, and coupled processes including chemical reaction, thermal response, and mechanical properties in natural and engineered environments demonstrated very promising outcomes. Numerical simulations, governing equations, physical constraints, laboratory observations, and field measurement data all can be utilized to develop fast and reliable data-driven models. The purpose of this session is to bring the latest theoretical, algorithmic, and practical advancements for flow, transport, and the coupled processes across spatial and/or temporal scales in order to advance various ML/DL methods with our domain knowledge in terms of various data types and amounts, modeling and simulations, and physical principles and constraints.
Index Terms:
1805 Computational hydrology [HYDROLOGY]
1832 Groundwater transport [HYDROLOGY]
1846 Model calibration [HYDROLOGY]
1849 Numerical approximations and analysis [HYDROLOGY]
Primary Convener: Hongkyu Yoon, Sandia National Laboratories, Geomechanics Department, Albuquerque, NM, United States
Conveners: Jonghyun Harry Lee, University of Hawai‘i at Mānoa, Civil and Environmental Engineering, Honolulu, HI, United States and Teeratorn Kadeethum, Sandia National Laboratories, Albuquerque, United States
Student/Early Career Convener: Teeratorn Kadeethum, Sandia National Laboratories, Albuquerque, United States
Prediction of Concentration Fields in Porous Media Using a Multiscale Neural Network (1165730)
Javier E. Santos1, Agnese Marcato2, Hari Selvi Viswanathan3, Gianluca Boccardo2, Daniele Marchisio2 and Masa Prodanovic4, (1)Los Alamos National Laboratory, Los Alamos, United States, (2)Politecnico di Torino, Torino, Italy, (3)Los Alamos National Laboratory, Los Alamos, NM, United States, (4)The University of Texas at Austin, Hildebrand Department of Petroleum and Geosystems Engineering, Austin, TX, United States
Training Deep Operator Networks for Coupled, Multi-Physics Applications in Coastal Hydrodynamics (1183056)
Peter Rivera Casillas1, Sourav Dutta2, Orie Cecil2 and Matthew Farthing2, (1)US Army Corps of Engineers, Information Technology Laboratory, Vicksburg, United States, (2)US Army Corps of Engineers, Coastal and Hydraulics Laboratory, Vicksburg, United States
State-Parameter mappings in unsaturated flow models in dikes using Generative Adversarial Networks (1197352)
Yusuke Homma, NARO National Agriculture and Food Research Organization, Ibaraki, Japan, Seiichiro Kuroda, National Agricultural Research Organization, Tsukuba, Japan and Nobuo Makino, National Agricultural Research Organization, Tsukuba Shi, Japan
Local and Global Mass Balance of Physics-Informed Neural Network (PINN) (1197573)
Md Lal Mamud, Los Alamos National Laboratory, Los Alamos, United States; University of Mississippi Main Campus, Department of Geology and Geological Engineering, University, United States, Bulbul Ahmmed, Los Alamos National Laboratory, Los Alamos, NM, United States, Satish Karra, Pacific Northwest National Laboratory, Richland, WA, United States and Maruti K. Mudunuru, Pacific Northwest National Laboratory, Richland, United States
Physics-Informed Machine Learning with Differentiable Programming for Heterogeneous Underground Reservoir Pressure Management (1048690)
Aleksandra Pachalieva1, Daniel O'Malley2, Dylan R Harp3 and Hari Selvi Viswanathan2, (1)Los Alamos National Laboratory, Computational Earth Science, Los Alamos, NM, United States, (2)Los Alamos National Laboratory, Los Alamos, NM, United States, (3)Los Alamos National Laboratory, Earth and Environmental Sciences Division, Los Alamos, NM, United States
Towards developing a generic method for predicting Flow Resistance in Alluvial Rivers using Machine Learning (1051825)
Amin Shakya1, Sanjay Giri2, Toshiki Iwasaki3, Mohamed Nabi2, Biswa Bhattacharya4 and Dimitri P Solomatine4, (1)Indian Institute of Science, Interdisciplinary Centre for Water Research, Bengaluru, India, (2)Deltares, Delft, Netherlands, (3)Hokkaido University, Sapporo, Japan, (4)IHE Delft Institute for Water Education, Delft, Netherlands
Machine Learning Applications for Estimation of Greenhouse Gas Emissions Using Multiple Satellite Images (1063470)
John Ringer, Sandia National Laboratories, Department of Geomechanics, Albuquerque, NM, United States, Teeratorn Kadeethum, Sandia National Laboratories, Albuquerque, United States, Trevor Harris, Texas A&M, Department of Statistics, College Station, United States and Hongkyu Yoon, Sandia National Laboratories, Geomechanics Department, Albuquerque, NM, United States
Embedding Hard Physical Constraints in Convolutional Neural Networks for Improved Electric Current Prediction in Porous Media (1075711)
Bernard Chang, The University of Texas at Austin, Hildebrand Department of Petroleum and Geosystems Engineering, Austin, MD, United States, Javier E. Santos, Los Alamos National Laboratory, Los Alamos, United States, Rodolfo Victor, PETROBRAS, Rio De Janeiro, Brazil, Hari Selvi Viswanathan, Los Alamos National Laboratory, Los Alamos, NM, United States and Masa Prodanovic, The University of Texas at Austin, Hildebrand Department of Petroleum and Geosystems Engineering, Austin, TX, United States
On incorporating first principles based physical conservation laws into global climate emulators (1076769)
Peetak Mitra1, Dipti Swapnil Hingmire2, Haruki Hirasawa3, Salva Ruhling Cachay1, Subhashis Hazarika1, Soo Kyung Kim4, Phil Rasch5, Hansi Alice Singh6 and Kalai Ramea7, (1)PARC, Palo Alto, United States, (2)Indian Institute of Tropical Meteorology, Pune, Pune, India, (3)University of Toronto, Toronto, ON, Canada, (4)LLNL, Livermore, United States, (5)Pacific Northwest National Laboratory, Richland, United States, (6)University of Washington Seattle Campus, Seattle, United States, (7)PARC, Palo Alto, CA, United States
Forward and inverse modeling of nonisothermal multiphase poromechanics using physics-informed neural networks (1077073)
Ehsan Haghighat, Massachusetts Institute of Technology, Civil and Environmental Engineering, Cambridge, United States, Danial Amini Mr., Sharif University of Technology, Civil Engineering, Tehran, Iran and Ruben Juanes, Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering and Earth Resources Laboratory, Dept. of Earth, Atmospheric and Planetary Sciences, Cambridge, MA, United States
Learning the Representation of Lee Vortices over Complex Topography in Taiwan (1084833)
Min-Ken Hsieh1, Chien-ming Wu1, YI-Chun Chen2 and Wei-Ting Chen1, (1)National Taiwan University, Department of Atmospheric Sciences, Taipei, Taiwan, (2)Academia Sinica, Research Center for Environmental Changes, Taipei, Taiwan
Subsurface Characterization using Deep Generative Adversarial Networks in a Bayesian Inverse Modeling (1096388)
Jichao Bao, University of Hawaii at Manoa, Honolulu, HI, United States, Hongkyu Yoon, Sandia National Laboratories, Geomechanics Department, Albuquerque, NM, United States and Jonghyun Harry Lee, University of Hawai‘i at Mānoa, Civil and Environmental Engineering, Honolulu, HI, United States
Rapid Modelling of Reactive Transport in Porous Media using Machine Learning: Limitations and Solutions (1106351)
Vinicius Silva1, Geraldine S. G. Regnier1, Pablo Salinas2, Claire Heaney1, Matthew Jackson1 and Christopher Pain2, (1)Imperial College London, London, SW7, United Kingdom, (2)Imperial College London, London, United Kingdom
Machine Learning-assisted Identification of Structural Properties Controlling Transport in Fracture Networks (1113077)
Seonkyoo Yoon, University of Minnesota Twin Cities, Minneapolis, MN, United States, Jeffrey Hyman, Los Alamos National Laboratory, Los Alamos, NM, United States, Weon Shik Han, Yonsei University, Department of Earth System Science, Seoul, South Korea and Peter K. Kang, University of Minnesota, Department of Earth and Environmental Sciences, Minneapolis, United States
Physics-Informed Machine Learning of Geothermal, Geomechanical, Geochemical Process (1125282)
Velimir monty V Vesselinov, EnviTrace LLC, CTO, Santa Fe, United States and Tracy L Kliphuis, EnviTrace LLC, CEO, Santa Fe, United States
