PoreLab lecture on Robust Topological Characterization of Fluid Microstructure via Persistent Homology

PoreLab Lecture

When: Wednesday, February 17th, 2021, at 10:00 (Central Europe)

Where: Zoom. To join, please follow the link below the abstract or add this event to your calendar by opening the attached calendar ics-file.

Title: Robust Topological Characterization of Fluid Microstructure via Persistent Homology

Speaker: Dr. Anna Herring

Affiliation: Researcher / Postdoctoral fellow, Research School of Physics, ANU


Topological characterization of porous media microstructures and the multiphase fluid distributions within has become a promising and popular approach for linking structure to flow properties; including engineering-relevant properties such as fluid flow regime, capillary fluid trapping, fluid permeability, and medium wettability. Many experimental studies that characterize fluid distributions utilize 3D imaging techniques such as X-ray microcomputed tomography (X-ray microCT). Because topological measures are independent of length-scale, quantities such as the Betti numbers and the Euler characteristic are notoriously sensitive to length-scale dependent features; for imaged data, this indicates that issues like microporosity, speckle noise, partial-volume effects, and image resolution can have significant impacts on the measured results. Persistent homology is a branch of topological data analysis designed to measure topological features as a function of length scale, and can be used to discern and correct for these impacts. We present classical topological and persistent homology analysis of a high-resolution 3D x-ray microCT data set of quasi-static spontaneous imbibition into Bentheimer sandstone, a clean sandstone with multi-scale porosity. We observe the variation of topological metrics as a function of noise filter and resolution coarsening, and demonstrate how persistent homology can be used to measure and address these length-scale dependencies to provide robust topological characterization.

Connect to meeting:
Meeting link: https://uio.zoom.us/j/67356471922
Meeting ID: 673 5647 1922

Join by SIP:


Join by H.323:
Meeting ID: 673 5647 1922