Dr. Megan Masterson

Tidal Disruption Event Dust Echoes

When an unfortunate star wanders just a little too close to a supermassive black hole, the tidal forces of the black hole can exceed the star's self-gravity, ripping the star apart and temporarily awakening a previously dormant black hole. As some of the stellar debris accretes onto the black hole, a temporary flare of emission is released across a variety of wavelengths, allowing us to study these previously invisible supermassive black holes. The recent advances in optical time-domain surveys have propelled the field of TDEs into the population regime, but there is growing evidence that these surveys may be biased against finding TDEs in dusty, gas-rich environments, which are expected to be common in the nuclei of star-forming galaxies. Obtaining a full census of TDEs and understanding observational biases is crucial for using TDEs as probes of supermassive black hole demographics.

To this end, I led the first systematic search for TDEs in the mid-infrared (MIR) band using data from Wide-field Infrared Survey Explorer (WISE), which scanned the sky in two filters every 6 months from 2013-2024. The MIR is the ideal wavelength regime to search for TDEs in dusty nuclei, as the UV/optical emission from the TDE is absorbed by surrounding dust and re-radiated at longer wavelengths, producing a "dust echo" that peaks in the MIR. In Masterson et al. (2024), we identified 12 TDEs within 200 Mpc, the majority of which were missed by optical surveys. This sample includes the closest TDE discovered to date, first published in Panagiotou et al. (2023). These MIR-selected TDEs occurs at a comparable rate to optically selected TDEs, no strong preference towards post-starburst galaxies (as is commonly seen in optical samples), and help solve the TDE Missing Energy Problem by revealing the extreme UV light that is impossible for us to measure directly. This work demonstrates that MIR time-domain surveys are a powerful tool for obtaining a complete census of TDEs and probing the dusty, circumnuclear environments of supermassive black holes.

I presented these results at the 2024 UCSB KITP TDE workshop, and you can watch the recording below.

Building on this foundation, in Masterson et al. (2025b) we obtained the first JWST/MIRI spectra of TDEs, providing an unprecedented view of their circumnuclear environments. All four TDEs in our sample display compact, accretion-driven emission lines from highly ionized gas, alongside strong silicate dust emission features at 10 and 18 µm. These spectra demonstrate that dust heated by TDEs is largely optically thin and that these features are much stronger than those seen in normal AGN with dusty tori. We modeled the expected time-dependent MIR spectrum, finding that most sources show a short-wavelength (< 8 µm) excess that may represent a late-time plateau in the flare emission, similar to the UV behavior of unobscured TDEs. Together, these results establish JWST as a transformative tool for studying black holes in dusty nuclei, and they pave the way for detailed modeling of how stellar disruptions interact with surrounding gas and dust.

I am also involved in many follow-up efforts to find more IR-selected TDEs and characterize their multi-wavelength properties. For example, Golay et al. (2026) completed a radio study of WTP14adeqka, finding that it shows a delayed radio outflow, similar to what is seen in optically selected TDEs. With VLBA observations, we showed that this is the first resolved radio outflow from a TDE, allowing us to directly measure the expansion speed of the outflow and estimate its kinetic energy. Our student, Prajna Nair, also recently published a study of high-luminosity, IR-selected TDEs in Nair et al. (2026), finding a break in the luminosity function that we expect for TDEs due to the tidal radius being within event horizon for more massive supermassive black holes. This paper is an important finding that helps confirm that these IR-selected TDEs are indeed TDEs, and not some other type of nuclear transient. Stay tuned for more exciting results from IR-selected TDEs in the near future!

The Dynamic X-ray Corona

Throughout my PhD work, I've shown that nuclear transients, including TDEs and changing-look AGN, can be powerful probes of this elusive object we call the corona. The corona is a compact, hot region of plasma that produces the hard X-ray emission from accreting supermassive black holes, but its origin and geometry are still fiercely debated. In X-ray binaries (i.e., accreting stellar mass black holes), there is evidence that the corona is related to the base of a relativistic jet, particularly because of the connected radio and X-ray variability during their state transitions (see e.g., Kara et al. 2019, Wang et al. 2020). However, recent X-ray polarization results from the IXPE mission suggest that the corona has a different geometry, parallel to the accretion disk. Study the dynamic nature of the corona in supermassive black holes can be much more difficult, primarily because the timescales are much longer compared to the evolution we regularly see in X-ray binaries (simply because supermassive black holes are just that much bigger). Supermassive black hole transients, though, are one key way to study the dynamic corona.

In Masterson et al. (2022a), we used a changing-look AGN, 1ES 1927+654, to show that the corona can be destroyed and reformed over the course of just a few years. This AGN underwent a dramatic optical/UV outburst in early 2018, and then within just a few months, the X-ray corona was destroyed -- the first time we ever saw this in an AGN (see Ricci et al. 2020 for more details of this discovery). My work showed that the corona eventually reformed, reaching its pre-outburst state within just a few years. I also modeled the X-ray spectra of this AGN during the outburst, showing that a peculiar 1 keV emission line could be modeled as relativistic reflection off of a super-Eddington, geometrically thick accretion disk. To model this feature, we created a new flavor of the xillver reflection models that assumes a thermal blackbody irradiating spectrum, appropriate for super-soft X-ray sources like TDEs and hence called xillverTDE. Motivated by theoretical models of super-Eddington reflection features, we blurred this model with a simple Gaussian kernel and allowed the observed line to be blueshifted. Together, this model provided a good fit to the observed 1 keV feature and suggested an ultrafast outflow launched from the innermost regions of the accretion flow with a velocity of roughly a third the speed of light. This feature eventually disappeared as the corona reformed.

Despite returning to its pre-outburst state, the story of 1ES 1927+654 became even more puzzling after the corona reformed. In Masterson et al. (2025a), I discovered an X-ray quasi-periodic oscillation (QPO), in which the X-ray flux varied (quasi-)periodically as can be seen even by eye in the below light curves. The most remarkable aspect of this QPO is that its frequency changed over time, increasing in frequency from 0.9 mHz in 2022 to 2.3 mHz by 2024. At its maximum, this frequency is associated with the orbital timescale within just a few gravitational radii of the black hole, meaning that whatever is driving these oscillations is occuring very close to the black hole. This is just the second high-confidence QPO ever discovered in an AGN and the first to show any frequency evolution. These oscillations persist through (at least) mid-2026, and in Masterson et al. (2026), we show that the QPO frequency has now plateaued around 2.5 mHz (and is detected with two separate X-ray telescopes -- XMM-Newton and NuSTAR!).

What drives the QPO in 1ES 1927+654? Two important observations are that the QPO is dominant in the hard X-ray band (2-10 keV) and that a radio flare was observed within 6 months of the QPO turning on (see Meyer et al. 2025). Together, these observations suggest that the QPO is related to the X-ray corona, but the exact mechanism is still unclear. We proposed two potential solutions: either the QPO is associated with magneto-acoustic oscillations in the X-ray corona, or it arises from a degenerate companion like a white dwarf orbitting the supermassive black hole. In the latter case, the source would emit gravitational waves that would be detectable by the upcoming LISA mission, making this one of the first known extragalatic multi-messenger sources in the LISA era. Regardless, of interpretation, the QPO is like nothing we've seen before and presents a new window into extreme accretion physics. Continuing to monitor this enigmatic source will be crucial to unlocking the origin of the QPO, its connection to the earlier outburst, and the corona-jet connection. I'm also particularly interested in finding other sources that exhibit similar behavior to 1ES 1927+654, as this will help us to understand how this source fits into the broader population of extreme AGN variability.

High Resolution X-ray Spectroscopy

During my Masters work at the University of Cambridge, I worked with data from the High Energy Transmission Grating on the Chandra X-ray Observatory to probe the narrow Fe Kα line around nearby AGN. Despite this line being nearly ubiquitous in AGN X-ray spectra, its origin is still debated. In Masterson et al. (2022b), we used a novel spectroscopy technique that leveraged the fact that in a dispersive spectrograph, the spatial extent of the emitting gas will impart an additional broadening to the line that is different across different spectral orders. By combining three different spectral orders, we were able to disentangle the effect of velocity broadening (which is constant across different spectral orders) and spatial extent. This new technique provides unique constraints on the location of the emitting material, which is nearly impossible to do with standard imaging techniques, even with the best resolution provided by Chandra. This analysis revealed that the Fe Kα line in our sample of nearby AGN is typically produced in the dusty torus, consistent with other recent work, but with a larger radius than is typically inferred from dynamical broadening alone. In the future, I'm interested in applying similar techniques with the recently launched XRISM mission.

Cosmic Evolution of AGN Feedback

I am also broadly interested in how accretion can impact a black hole's surroundings, and have worked on this avenue by exploring AGN feedback in galaxy clusters as a senior undergrad and during the first year of my PhD. In Masterson et al. (2023a), I worked with Prof. Mike McDonald on a relaxed galaxy cluster at z ~ 1.4 (high redshift for galaxy clusters!). This system looked surprisingly like many of the low-redshift galaxy clusters that have been studied for the last few decades -- namely, it showed a cool core in the X-rays with a strong radio source associated with the brightest cluster galaxy (BCG) that had enough power to prevent the gas from cooling. Essentially, this high-redshift cluster showed well-regulated AGN feedback, where the central supermassive black hole in the BCG is able to prevent gas in the galaxy cluster from cooling! This is the most distant cluster we know of that has these properties, and thus, it helps set limits on when AGN feedback became a well-regulated and normal process in the universe!