Nat has a phd in applied mathematics from University of Colorado Boulder, where they did research in collaboration with the National Center for Atmospheric Research High Altitude Observatory. Now they are a postdoctoral fellow at Goddard working on machine learning methods to model the coronal magnetic field.
Nat Mathews
(NASA Postdoctoral Program Fellow)
| Email: | n.h.mathews@nasa.gov |
| Org Code: | 671 |
| Address: |
NASA/GSFC Mail Code 671 Greenbelt, MD 20771 |
| Employer: | NPP POST-DOC CONTRACT |
Brief Bio
Current Projects
Data-Optimized Coronal Field Model
Remote Sensing
An inversion framework by which the full magnetic field in an active region of the sun may be ascertained
Coronal Physics-Informed Neural Net
Theory & Modeling
Training a machine learning model on MHS coronal simulations
Positions/Employment
Postdoctoral Fellow
NASA - Goddard Space Flight Center
January 2022 - Present
Teaching Experience
2015-2021, University of Colorado Boulder
Laboratory Instructor: Calculus 3, Differential Equations
Teaching Assistant: Calculus 2, Calculus 3, Differential Equations
2013-2014, Rochester Institute of Technology
Teaching Assistant: Calculus 2, Linear Algebra
Education
2021, PhD: Applied Mathematics, University of Colorado Boulder
Advisor: Natasha Flyer. Dissertation: Computational Modeling for 3D Data Reconstruction of Solar Coronal Magnetic Fields
2015, BS: Computational Mathematics, Rochester Institute of Technology
Magna cum Laude, Astronomy minor.
Awards
2021, NASA Postdoctoral Program Fellowship, NASA Goddard Space Flight Center
2018, Newkirk Fellowship, National Center for Atmospheric Research High Altitude Observatory
2014, Summer Undergraduate Research Fellowship, Rochester Institute of Technology
2011, National Merit Scholarship
Publications
Refereed
2022. "Solving 3D magnetohydrostatics with RBF-FD: Applications to the solar corona." Journal of Computational Physics 462 111214 [10.1016/j.jcp.2022.111214] [Journal Article/Letter]
2020. "Reconstructing the Coronal Magnetic Field: The Role of Cross-field Currents in Solution Uniqueness." The Astrophysical Journal 898 (1): 70 [10.3847/1538-4357/ab9dfd] [Journal Article/Letter]
2019. "Data-optimized Coronal Field Model. I. Proof of Concept." The Astrophysical Journal 877 (2): 111 [10.3847/1538-4357/ab1907] [Journal Article/Letter]
Talks, Presentations and Posters
Invited
Machine Learning as an Emulation Tool for Inverse Problems in Space Weather
February 19, 2023
American Meteorological Society
Computational Modeling for 3D Data Reconstruction of Solar Coronal Magnetic Fields
February 26, 2022
High Altitude Observatory Colloquium
Reconstructing the Coronal Magnetic Field: The Role of Cross-Field Currents in Solution Uniqueness
February 14, 2021
Boulder Space Weather and Machine Intelligence seminar
Other
Emulating Coronal Magnetic Fields with Physics-Informed Neural Networks
11, 2022
American Geophysical Union
A 3D Mesh-Free Solver for Magnetohydrostatic Simulations in the Corona
September 8, 2022
Triennial Earth-Sun Summit
Emulating Magnetohydrostatic Models with Physics-Informed Neural Nets
September 7, 2022
Triennial Earth-Sun Summit
Emulating Coronal Fields with Physics-Informed Neural Nets
July 28, 2022
SHINE
A 3D Mesh-Free Solver for Magnetohydrostatic Simulations in the Corona
13, 2021
New Capabilities for Adaptive Mesh Simulation Use Within FORWARD
14, 2016
American Geophysical Union