From Curiosity to Neutrinos
By Karim Hassinin, University of Houston
The first science movie I ever watched had no subtitles I could read. Planets slid across the screen, and I sat there sounding out English words I did not yet know. I kept rewinding the same scene, trying to match the narration to the pictures. That habit — replay, examine, name what you are seeing — never left. It became the way I approach science.
Years later, in my first semester as a teaching assistant in my Ph.D., I discovered how satisfying it is to make an idea land in someone else’s head. A student and I were aligning a simple setup of lenses and light beams — an optics bench used to study how light focuses. We nudged the lens a few millimeters; the blurry spot on the screen snapped into focus; the student smiled; I did too. The equation hadn’t changed, but our understanding had. That small click — “so that’s what the result means” — is why I like research.
I work in neutrino physics. Neutrinos are tiny, nearly massless particles that rarely interact with anything. About 60 billion of them from the Sun pass through each square centimeter every second — trillions stream through your body right now, unnoticed. Once in a while, one hits the nucleus of an atom, like a cue ball striking a cluster of billiard balls, knocking out protons or neutrons and leaving behind a faint trail of energy. Our detectors capture those traces, and from them we try to reconstruct what happened.
I study how to model those collisions precisely enough that experiments can interpret their data with confidence — because every neutrino experiment depends on these models to turn what the detector sees (light and charge) into what it really means (neutrino energy, type, and interaction). If the models are wrong, our conclusions about fundamental properties — like the neutrino mass ordering or whether neutrinos and antineutrinos behave differently — could be biased.
In neutrino research, we can’t control or aim the particles individually, so we depend on detailed computer simulations to understand what our detectors are seeing.
These simulations are built using event generators — software that encodes the known physics of how neutrinos interact with different materials. In an event generator, you specify the neutrino flux (its types, or “flavors,” and energy distribution), the target material (such as argon or carbon), and the interaction and nuclear models, including how particles behave inside the nucleus after the collision (called final-state interactions). The generator then produces “events” — lists of outgoing particles with their energies and momenta — that mimic what a detector would record.
By comparing these simulated predictions with real data, we can identify where our models fall short. When the disagreement is larger than what can be explained by statistical uncertainty — that is, beyond what random chance would allow — it often signals missing physics or an incorrect model assumption. We then adjust the model parameters (a process called reweighting or tuning), account for the resulting systematic uncertainties, and iterate until the predictions and data align more faithfully.
Alongside this modeling work, I also contribute to ICARUS, a neutrino detector at Fermilab. I help maintain its slow controls — software interfaces and graphical displays that continuously monitor detector conditions and alert the team if anything drifts out of range, helping ensure the detector operates safely and reliably.
I came to neutrinos because they live where the interesting questions are. Neutrinos can change “flavor” as they travel — a behavior called oscillation — which means they must have a tiny but nonzero mass. That discovery surprised physicists because the original Standard Model of particle physics assumed neutrinos were massless. It’s one of the clearest signs that our current theory is incomplete.
To uncover what’s missing, we need to understand how neutrinos interact with the building blocks of matter — the nuclei inside atoms. Each nucleus is a dense bundle of protons and neutrons held tightly together by the strong nuclear force. It’s not a neat pile of billiard balls; it’s a dynamic, many-body system where every piece influences the others. When a neutrino delivers even a tiny kick, the whole system can respond in chorus. Sometimes that response involves more than one nucleon at once, creating complex signals that are challenging — and fascinating — to model.
Every time I see a model line up with data after days of tuning and checking, it feels like catching a rare alignment — proof that we have understood one corner of nature. To compare the models that describe neutrino interactions, I had to become good in programming. My physics training was solid, but I realized that understanding equations wasn’t enough — I needed to teach those equations to a computer. I began learning algorithms: step-by-step instructions that tell the computer how to solve a problem. I discovered that there is not only one “right” way to do it. Each approach trades speed, precision, and elegance in different ways, and the challenge is finding the balance that works.
Debugging became my favorite part — the detective work of tracing a mistake through hundreds of lines of code until it reveals itself. Sometimes the fix is logical; sometimes it is creative, almost like finding light in darkness. Those moments trained me to think differently about physics itself: to see equations, software, and experimental data not as separate languages but as essential chapters of the same story.
I come alive in those moments of clarity — when an idea clicks, when a code finally runs without error, when the pieces of a problem suddenly fit together. It feels like pieces of the puzzle come together. I want to keep learning to connect ideas so deeply and precisely that new possibilities start emerging from developing my intuition.
Foundational research tends to pay off in ways we do not predict.
Quantum mechanics — the idea that particles can exist in multiple states until observed — once sounded like philosophy. Now it runs the technology inside every smartphone and computer chip. Neutrino science might feel just as distant today, but it plants the same kind of seeds. The tools we develop to sense and analyze these elusive particles often spill over into everyday technologies: ultra-sensitive detectors become the basis for medical imaging or environmental monitoring; data-analysis techniques built for particle collisions power advances in artificial intelligence and climate modeling; and large-scale computing methods used to simulate neutrino interactions help optimize energy grids and spacecraft navigation.
Even if the headline discoveries take years, the side effects arrive early: better sensors, smarter algorithms, and more powerful ways to handle massive amounts of data. At its core, neutrino research teaches us a discipline that extends far beyond physics — how to separate signal from noise, to measure the world honestly, and to find patterns that were invisible before.
Two puzzles frame the field. First, how do we extend our framework to explain neutrino masses and mixing — and do any hints of sterile behavior point to new particles? Sterile neutrinos are a hypothetical type of neutrino that would not interact through any of the known forces except gravity, making them even harder to detect than ordinary neutrinos. Oscillations already tell us we need neutrino mass terms and mixing parameters beyond the original Standard Model. Second, how do we model neutrino–nucleus interactions inside complex materials well enough that experimental conclusions aren’t biased by modeling errors?
I study cross-section models — ways of quantifying how likely different kinds of neutrino interactions are and how uncertain those predictions remain. Sometimes a neutrino strikes a single nucleon inside a nucleus, sometimes it bumps into a correlated pair, sometimes it excites the nucleus into a short-lived resonance, and at higher energies it can break the nucleus apart entirely. I compare predictions from different simulation tools, test them against real data, and identify where the models need refinement.
I like the feeling of a problem decreasing.
Early on, I was terrified of making a wrong assumption and biasing a result. A senior physicist told me, “You will catch it sooner next time.” That was freeing. Since then, I have kept notebooks of mistakes: a mislabeled axis that flipped a stray factor of two in a normalization, a plot that looked “too good” until we realized we had double normalized. Each error tightened the work.
Neutrinos carry information from places we cannot otherwise see. Understanding them is not only about building a perfect theory; it is about extending what humans can know. When a new model lines up with a stubborn dataset, or a detector stays stable through a long run because of a small software fix, I feel the same thing I felt as a kid rewinding a scene: the world coming into focus. If you have ever stared at a photo until the shapes made sense, you know the feeling that drives this field. My research is one part of that: make the models honest, keep the detector steady, compare prediction to reality, repeat. The universe doesn’t explain itself; we learn its language one clear sentence at a time.
Acknowledgements: This essay was written for the Science Policy & Advocacy for Research Competition (SPARC) at the Universities Research Association.
