A more sober reanalysis
In December 2019, as an undergraduate student at the University of Central Florida, I was taking the graduate-level course, AST5765C, taught by Dr. Joseph Harrington. I submitted a final project analyzing a spectroscopic observation of the spiral galaxy UGC 9039. The goal was to derive the galaxy's rotation curve and use it to demonstrate the presence of dark matter — a well-understood result, from a dataset chosen specifically because it was expected to work. My analysis did not work. In fact, I seemingly proved the existence of superluminal velocities and negative mass. I am still waiting on my two Nobel prizes. The paper I submitted acknowledged this honestly and declined to draw any conclusions it couldn't support, which was, in retrospect, the only thing it did right. (That and saving the data in a Git repository, which is how I was able to find it again seven years later.)
Three specific errors in my code were responsible. They were not subtle. I knew, at the time, that something was wrong. I ran out of time to fix it, submitted what I had, got a B-minus in the class, and moved on. Or tried to. To say I "ran out of time", though, is generous. I waited until the last minute, 48 hours before the project was due, to even start working on the project, despite the prescient advise *NOT TO DO THAT* from Dr. Harrington. The project has sat in the back of my head for the better part of seven years, not constantly, but with a regularity that is, on reflection, difficult to justify for an undergraduate final assignment.
This past spring I sat down, found the original data still committed to a Git repository from 2019, fixed the three errors, and ran the analysis correctly. The rotation curve came out clean. The dark matter is there, at a detection significance of 20σ, which is the kind of number that means "unambiguously yes." It was always there. The data was fine. Only the analysis was lacking, and now, it isn't! The only thing I regret is not doing it correctly the first time, and that I'll never get my two Nobel prizes.
To also be fair, part of my motivation for revisiting this project was as a letter of appreciation to Dr. Harrington, who was a fantastic professor and mentor, and whose letter of recommendation I solely credit for getting me my job at the Mikulski Archive for Space Telescopes. I wanted to show him that I had, eventually, done the project correctly despite the fact that I, as he humorously put it, had not "achieved a standout grade".
The reanalysis is written up in the style of an Astronomy & Astrophysics paper, because that is the format the original used and it seemed only fair to meet it on its own terms. Both documents are below. The code for the project is available on GitHub.
P.S. Yes, UGC 9039 has dark matter. No, it does not have superluminal velocities or negative mass.