Multiscale, multifidelety, and hybrid machine-learning methods for flow and transport in hydrologic systems

Session ID#: 24628

Session Description:
Again and again we find that effective flow and transport equations fail to accurately describe observations, across hydrologic systems. This includes multiphase processes and reactive transport in complex flows, including porous media, streams, rivers and beyond. To deal with shortcomings novel approaches based on multiscale methods, meso-scale models, and hybrid data-driven methods have emerged as alternatives to standard models. This session is devoted to advances in such methods. General topics of interest include, but are not limited to :

- hybrid methods describing processes on different scales in different parts of computational domains;

- multiscale methods using innovative computational closures;

- mesoscale methods, including generalized random walks;

- multiscale finite-volume and finite element methods;

- Gaussian Processes, Deep Neural Nets;

- other methods bridging gaps between data and models at different scales. 

We seek applications to subsurface contaminant migration, bioremediation, multiphase flows, hyporheic exchange and any related topics.

Primary Convener:  Diogo Bolster, University of Notre Dame, Department of Civil and Environmental Engineering and Earth Sciences, Notre Dame, IN, United States
Convener:  Alexandre M Tartakovsky, Pacific Northwest National Laboratory, Richland, WA, United States

Abstracts Submitted to this Session:

Zhi Dou, Hohai University, Nanjing, China
Thomas Sherman1, Diogo Bolster1, Abbas Fakhari2, Savannah Miller3 and Kamini Singha4, (1)University of Notre Dame, Notre Dame, IN, United States, (2)Edmonton, AB, Canada, (3)Colorado School of Mines, Golden, CO, United States, (4)Colorado School of Mines, Department of Geology and Geological Engineering, Golden, CO, United States
David A Barajas-Solano and Alexandre M Tartakovsky, Pacific Northwest National Laboratory, Richland, WA, United States

See more of: Hydrology