We were seeing light echoes — the original light from the supernova reflected off surrounding dust clouds in the LMC, like sound waves bouncing through the Alps. The echoed light’s pinball path meant it took extra time to make it to Earth, arriving about 15 years after the light that came straight from the supernova. These light echoes form rings tracing the surfaces of surrounding dust clouds that are located the same distances from the supernova. Individual rings with different radii represent reflections from sheets of dust at different distances from the explosion. It turned out that these light echoes had been discovered before by another project, so we had simply rediscovered them.
But as Rest continued to scrutinize the observations, he started finding additional light echoes with arcs that corresponded to locations other than that of SN 1987A. These new echoes were from supernovae that had exploded in the LMC centuries ago, yet we were still seeing the original explosions in delayed form. These light echoes revealed a unique way to study supernovae long after their initial blasts. And while our project ultimately did not find compelling evidence for MACHOs as an important source of dark matter, we still found something totally unexpected — and potentially quite powerful.
Following my work analyzing red supergiants in the LMC, my colleagues and I used the Hydra-CTIO spectrograph to measure velocities of other stars in the LMC. We concentrated on stars much older than the young red supergiants to see how the passage of billions of years would affect their dynamics, and eventually gathered thousands of spectra.
After dissecting the data, we found that most of these older stars had velocities that traced the same flat rotation curve as the red supergiants; however, they also exhibited a bit more random noise, which is expected given they’ve experienced billions of years of orbital evolution. But roughly 5 to 10 percent of the total had velocities that suggest they are orbiting in the opposite direction as the rest — very strange! It would be just as weird if one of the planets in our own solar system orbited the Sun in the wrong direction.
Because we could only measure the velocity of the stars along the line of sight (motion toward or away from us), another possibility was that the stars are orbiting in the same direction as the rest but are highly inclined, traveling well above and below the disk of the LMC. We concluded this was a more likely situation, and recently confirmed it by obtaining full 3D velocities of the stars using data from the Gaia satellite. But that still didn’t explain how this small subset of oddly orbiting stars got on their current track.
Tracing stars back through time
We have known for a long time that the LMC and SMC are a pair of gravitationally interacting galaxies. The most obvious result of this interaction is the Magellanic Stream and its leading arm, which form of a twisted ribbon of neutral hydrogen gas that astronomer David Nidever of Montana State University found extends over more than half the sky.
This ribbon includes streamers that connect to both the LMC and SMC, with two of these filaments converging on 30 Doradus. Viewed from the LMC, the streamers have the same velocities as those of the strangely moving stars we had previously found. Moreover, we measured the abundance of iron in some of our weird stars, finding they had surprisingly little iron compared to other stars in the LMC. However, they were a good match for SMC stars.
Based on the streamers and seemingly migrant stars, we concluded that the LMC had stolen stars from its smaller sibling, the SMC. These displaced stars are now orbiting in front of and behind the LMC’s disk. Plus, the arms connecting the LMC and SMC were apparently ripped out along with these stars and are now falling onto the LMC’s disk, right at the location of 30 Doradus. This scenario would explain why 30 Doradus is forming stars so aggressively. The infalling material provides external pressure on 30 Doradus and keeps the stellar winds and supernovae within from completely blowing away the nearby gas, which would halt further star formation. These ripped-out, massive stars might also be the source of the microlensing events that were seen in the 1990s, which would further rule out MACHOs as a significant source of dark matter scattered throughout the Milky Way.