Hi! I’m Even — a physicist by training, currently working on weather modeling with data-driven approaches at the Norwegian Meteorological Institute (MET Norway). I’m passionate about physics, modeling, visualization, and machine learning, and I often find myself solving problems through code. If you’re ever up for a chat about any of those topics, don’t hesitate to reach out!
In my spare time, I enjoy staying active and social — you’ll often find me doing CrossFit, gravel cycling, or playing chess.
My background
I grew up in a small town just outside Oslo, Norway, and moved to the city in 2014 to study physics at the University of Oslo (UiO). I earned my bachelor’s degree in 2017, including an exchange semester at Nanyang Technological University (NTU) in Singapore.
For my master’s degree, I explored how to solve the Schrödinger equation using machine learning. The variational principle in quantum mechanics states that any trial wave function will yield an energy equal to or higher than the true ground state energy. By representing the wave function with a neural network, we optimized variational parameters to minimize the local energy — effectively treating the problem as a machine learning task. This work resulted in a publication.
Alongside my studies, I had what I’d consider the ideal part-time job: teaching assistant at UiO. I taught a range of physics courses including classical mechanics, electromagnetism, and computational physics (full list below). During the summer of 2018, I interned at Andøya Space Center (ASC), where I developed an automated framework for cloud detection. Rocket launches depend on precise atmospheric measurements using LIDARs — which require clear skies — so we created software using an all-sky camera to replace the manual, error-prone cloud assessments. As a side note, ASC just became the first site in Europe to launch satellites!
Here in 2018, holding the lens system of the all-sky camera used for cloud detection at Andøya Space Center.
After completing my master’s, I pursued a PhD focused on friction at the nanoscale using molecular dynamics (MD) simulations. These simulations offer space-time resolution far beyond what’s possible in experiments, allowing deeper insights. I discovered that frictional silicon carbide interfaces strengthen over time due to surface diffusion — explaining why friction decreases at higher sliding velocities. This work culminated in a publication in Tribology Letters. During my PhD, I also delved into generative machine learning, which led me to address the inverse frictional design problem. I defended my thesis in August 2024.
In May 2024, I joined MET Norway as a researcher, working on developing machine learning models for high-resolution weather forecasting in the Nordics. Machine learning has recently shown great promise in global weather forecasting due to its accuracy and efficiency, and we’re building on this momentum with a regional model. (You can check out our preprint here: arXiv:2409.02891).
Looking back, I’ve noticed a trend in my academic journey: the scale of the systems I’ve modeled keeps growing. I started with single atoms, moved on to nanoscale systems with millions of atoms, and now I’m working on atmospheric-scale weather systems. Maybe the next step is cosmological modeling?
Summary
Education
- 2019-2024: PhD supervised by Anders Malthe-Sørenssen and Henrik Andersen Sveinsson at University of Oslo. Thesis: “Molecular Dynamics Simulations of Dry Friction on the Nanoscale”
- 2017-2019: Masters in Computational Physics supervised by Morten Hjorth-Jensen at University of Oslo. Thesis: “Studies of Quantum Dots using Machine Learning”
- 2017: Exchange student, Department of Physical and Mathematical Sciences, Nanyang Technological University, Singapore
- 2014-2017: Bachelor of Physics, Astronomy and Meteorology, University of Oslo
Work experience
- 2024-today: Researcher, Norwegian Meteorological Institute
- 2018: Summer internship, Andøya Space Center
- 2018-2019: Teaching Assistant, University of Oslo
Courses taught
- FYS4460, Percolation and Disordered Systems, 2024
- NAT3000, Scientific Programming in School, 2023
- FYS3150/4150, Computational Physics, 2021, 2022, 2023
- FYS1100, Mechanics and Modeling, 2023
- FYS-MEK1110, Classical Mechanics, 2018, 2019, 2020, 2021, 2022
- FYS1120, Electromagnetism, 2018, 2019
Skills
- Programming languages: Python, C++
- Software and Tools: Linux, HPC, Data Analysis, Machine Learning
Languages
- Norwegian
- English
Presentations
- EuroHPC User Day 2024, Amsterdam, Netherlands
- European Geological Union (EGU) General Assembly 2023, Vienna, Austria