We had an interview with Prof. Wolfgang E. Kerzendorf on 22 January 2026, beginning with an overview of his academic career and the milestones that shaped his research trajectory.
Prof. Kerzendorf obtained his Vordiplom[1] in Physics from Universität Heidelberg. This was followed by a summer research placement at Mount Stromlo Observatory, where he later completed his Ph.D. in 2011 with the dissertation Type Ia Supernovae: Progenitors and Explosions.
Type Ia Supernovae: Progenitors and Explosions
His career continued with a series of increasingly influential positions. He was a Postdoctoral Researcher at the University of Toronto, followed by an ESO Fellowship at the European Southern Observatory between 2014 and 2018. He subsequently worked at New York University and the Flatiron Institute until the middle of 2019.
He currently serves as an Assistant Professor at Michigan State University, with appointments in the Department of Physics & Astronomy and the Department of Computational Mathematics, Science & Engineering.
Discussing his doctoral work, Prof. Kerzendorf explained that his early research focused on Type Ia supernovae and the search for possible companion stars of their progenitors. Later observational evidence suggested that such companions are much rarer than previously expected, reshaping the understanding of these stellar explosions. His Ph.D. years were strongly influenced by observational astronomy, close mentoring, and participation in large scientific collaborations.
One of his central messages was that a Ph.D. is not primarily a demonstration of intelligence, but rather of sustained motivation, curiosity, and the ability to remain focused on difficult problems over long periods. During the interview he remarked that "quitting is underrated," emphasizing that successful doctoral research ultimately depends on genuine passion for the subject rather than perseverance alone.
Following his doctorate, his career developed in a distinctly nonlinear manner, often shaped by unexpected opportunities and collaborations. Important stages included his postdoctoral work in Toronto, major contributions to the Astropy Project, and subsequent positions across Europe and the United States. He considers compatibility with collaborators, open-source scientific culture, and interdisciplinary environments to have been decisive factors throughout his career.
A major methodological contribution from his research group is the development of emulation techniques that dramatically accelerate computationally expensive astrophysical simulations. By constructing neural-network-based surrogate models, his group achieved speedups of up to one hundred million times for the radiative transfer code TARDIS, making sophisticated astrophysical inference substantially more efficient. He also noted that his group adopted machine learning techniques well before they became mainstream within astrophysics.
Beyond astrophysics, Prof. Kerzendorf has developed a parallel research direction in computational meta-science—the "science of science." This work investigates scientific activity at scale by tracking publications, uniquely identifying researchers, mapping expertise, and improving peer review and collaboration systems. He argues that exponential growth in scientific output has exceeded the capacity of traditional research structures, requiring algorithmic approaches and modern infrastructure to organize scientific knowledge more effectively.
Interestingly, this research direction originated from a casual dinner conversation during which he mentioned it as a personal interest outside his primary astrophysical research. His colleague immediately recognized its scientific significance and encouraged him to elevate it into a formal research programme. This experience reinforced the value of interdisciplinary thinking and remaining open to unexpected opportunities.
He also emphasized an important methodological caution regarding artificial intelligence. While modern large language models are valuable tools, they are not universally optimal for problems such as expert matching or scientific knowledge organization. In many cases, simpler statistical approaches can outperform significantly more complex machine learning systems.
Within astrophysics, his current research focuses on connecting high-fidelity numerical simulations with observations of explosive transients, including supernovae and neutron star mergers. As an example, he discussed the interpretation of events such as GW170817, which combined gravitational-wave observations, kilonova emission, and heavy-element nucleosynthesis, providing direct insights into the production of elements such as gold and uranium. His group investigates both the physical mechanisms behind these events and their frequency throughout the Universe.
Overall, Prof. Kerzendorf's career illustrates that modern scientific paths are rarely linear. They are shaped by interdisciplinary exposure, collaborative environments, unexpected opportunities, and a willingness to pursue new ideas. His experience suggests that impactful research increasingly emerges at the intersection of physics, computation, and large-scale data science, while long-term success remains fundamentally driven by intrinsic motivation, curiosity, and openness to unconventional opportunities.