🔭 astrophysics

AMPM I. A Targeted Search for Asteroid Mass Primordial Black Hole Microlenses

The AMPM survey introduces a high-cadence microlensing search in the Large Magellanic Cloud to constrain asteroid-to-planetary-mass primordial black hole dark matter, utilizing five nights of data and refined efficiency modeling to detect a single candidate and establish limits of up to 30% of the Galactic dark matter distribution at the 95% confidence level.

Original authors: Renee Key, Edward N. Taylor, Ken C. Freeman, Jeremy Mould, Abhijit Saha, Anais Möller, Timothy M. C. Abbott, Alan R. Duffy

Published 2026-05-20

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Renee Key, Edward N. Taylor, Ken C. Freeman, Jeremy Mould, Abhijit Saha, Anais Möller, Timothy M. C. Abbott, Alan R. Duffy

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Hunting for Invisible Ghosts

Imagine the universe is filled with a mysterious, invisible substance called Dark Matter. We know it's there because it holds galaxies together, but we can't see it. For decades, scientists have wondered: What is it made of?

One popular theory suggests that Dark Matter might be made of Primordial Black Holes (PBHs). These aren't the giant black holes formed by dying stars; they are tiny, ancient black holes that formed in the very first split-second after the Big Bang. Some might be as heavy as a mountain, others as heavy as a planet, and some as light as an asteroid.

The authors of this paper wanted to hunt for these tiny, asteroid-sized black holes. They call their project AMPM (Asteroid-Mass Primordial black hole Microlensing).

The Detective Tool: Gravitational Microlensing

How do you find something you can't see? You look for how it bends the light of things you can see.

Think of a gravitational lens like a magnifying glass. If a massive object (like a black hole) passes between Earth and a distant star, its gravity bends the star's light, making the star look temporarily brighter. This is called microlensing.

The Analogy: Imagine you are in a dark room looking at a single lightbulb in the distance. If a clear glass marble (a small black hole) floats between you and the bulb, it might briefly focus the light, making the bulb sparkle for a moment.
The Challenge: The smaller the black hole, the shorter the "sparkle" lasts. An asteroid-sized black hole might only make a star twinkle for a few minutes. To catch this, you need to take pictures of the sky very, very fast—like a high-speed camera taking a photo every 50 seconds.

The Mission: The AMPM Survey

The team used a powerful camera called DECam (on the Blanco telescope in Chile) to stare at a specific patch of the Large Magellanic Cloud (LMC), a small galaxy next to our own.

The Strategy: They took 5 nights of photos, snapping images every 50 seconds. They were looking for stars that suddenly got brighter and then faded back down in a matter of hours.
The Obstacle (The "Fuzzy" Problem): The paper highlights a major hurdle called the Finite Source Effect.
- The Metaphor: Imagine trying to see a tiny pebble cast a shadow on a giant beach ball. If the pebble is too small, the shadow is so faint and spread out that you can't tell it's there. Similarly, if the background star is huge (a giant star) and the black hole is tiny, the "brightening" effect gets washed out and becomes very hard to detect.
- Because many stars in the LMC are giant stars, the team had to build a very sophisticated computer model to account for this "fuzziness."

The Hunt: Filtering the Noise

The team collected data on millions of stars. However, the sky is noisy. Clouds, atmospheric turbulence, and other types of variable stars (like flaring stars) can look like a black hole passing by.

To find the real signal, they built a "digital sieve" (a detection pipeline) with several filters:

Quality Check: They threw out data from stars that were too faint or had bad measurements.
Atmosphere Check: They looked for patterns caused by the Earth's atmosphere (seeing) and removed those.
Pattern Check: They checked if the brightness change was a one-time event (like a black hole) or a repeating pattern (like a pulsing star).

The Results: One Candidate and New Limits

After running their complex filters through 5 nights of data (covering over 2 million stars), they found one single candidate.

The Candidate: On the fourth night, they saw a star brighten for about 3.5 hours. This timescale fits perfectly with the idea of an asteroid-to-moon-sized black hole passing by.
The Constraint (The "No-Go" Zone): Even though they only found one candidate, the fact that they didn't find hundreds of them tells us something important.
- The Analogy: Imagine you are fishing in a lake. If you cast your net 100 times and catch only one fish, you know the lake isn't completely full of fish. You can put a limit on how many fish could possibly be there.
- The Finding: The AMPM survey can rule out the idea that asteroid-mass black holes make up more than 30% of all Dark Matter in our galaxy. If they made up 100%, the team would have seen many more events.

Why This Matters

This paper is a "pilot study" (a test run). It proves that we can use fast-cadence cameras to hunt for these tiny, ancient black holes.

The Twist: The one candidate they found is interesting. It's similar to a few other candidates found by other telescopes (like Subaru-HSC). If this one star is indeed a black hole, it suggests that these tiny black holes might be more common than the most strict previous surveys (like OGLE) thought.
The Future: The authors say that to find even smaller black holes (lighter than asteroids), we need to look at even smaller, fainter stars (like white dwarfs) or look at more distant galaxies, because the "fuzziness" of giant stars hides the smallest black holes.

Summary

The AMPM team took a high-speed photo marathon of a nearby galaxy to hunt for invisible, asteroid-sized black holes. They found one promising "blink" in the starlight. While they didn't find enough to say these black holes are the main ingredient of Dark Matter, they proved that these tiny objects could make up a significant chunk (up to 30%) of the universe's missing mass. They also showed that finding them requires very fast cameras and smart math to cut through the "noise" of the sky.

Technical Summary: AMPM I. — A Targeted Search for Asteroid Mass Primordial Black Hole Microlenses

Problem Statement
Gravitational microlensing is a primary tool for constraining the abundance of dark matter in the form of compact objects, specifically Primordial Black Holes (PBHs), across a wide mass range ( $10^{-11} \lesssim \log M/M_\odot \lesssim 5$ ). While previous surveys (e.g., OGLE, Subaru-HSC, Kepler) have placed stringent limits on PBH dark matter, the "asteroid-to-planetary-mass" regime ( $10^{-11} - 10^{-6} M_\odot$ ) remains challenging to probe. Detecting microlensing events in this mass range requires high-cadence observations (minutes or less) because the Einstein crossing time ( $t_E$ ) scales with the square root of the lens mass. Furthermore, for low-mass lenses, the physical size of the source star becomes comparable to the Einstein ring radius, leading to "finite-source" effects that dampen the peak amplification and broaden the event timescale. Existing surveys have struggled to balance the need for rapid cadence with the ability to resolve these damped signals in crowded stellar fields like the Large Magellanic Cloud (LMC).

Methodology
The authors present AMPM (Asteroid-Mass Primordial black hole Microlensing), a pilot survey utilizing the Dark Energy Camera (DECam) to image a single field in the LMC (Field 51) over five nights.

Observations: The survey employed a rapid cadence of 50 seconds (20-second exposures plus overhead) using a single broad-band VR filter to minimize overheads and maximize sensitivity to fast transients. The field was chosen in a peripheral region of the LMC to minimize stellar crowding and photometric blending.
Photometry: Data were processed using a variant of DoPHOT for point-spread function (PSF) fitting. Light curves were constructed for millions of stars, with quality control applied to remove sources with poor PSF fits or those near CCD chip gaps.
Simulation and Efficiency: To determine detection efficiency, the authors injected simulated microlensing events into the observed light curves. These simulations utilized a Finite-Source Point-Lens (FS-PL) model, incorporating realistic stellar radii derived from MIST synthetic photometry matched to SMASH DR2 data. The simulations assumed a Navarro-Frenk-White (NFW) profile for the Milky Way dark halo.
Detection Pipeline: A multi-stage, cut-based pipeline was developed to identify fast, non-repeating variability:
1. Quality Control: Filtering for well-fitted stars with complete light curves.
2. Statistical Screening: Utilizing the Von Neumann ratio ( $\eta$ ) to detect trends, a "Seeing-related Flux Blending" parameter ( $B$ ) to identify atmospheric artifacts, and a Periodicity Time Comparison ( $T$ ) to reject periodic variables.
3. Thresholding: A joint threshold in $\eta$ , $B$ , and $T$ spaces was calibrated using injected simulations to maximize efficiency while minimizing false positives.
4. Fitting: Candidates underwent a fast Point-Source Point-Lens (PS-PL) fitting routine to evaluate $\chi^2$ and signal significance.
5. Rejection: Candidates were cross-matched with the Gaia DR3 variability catalogue and checked for repeated variability to remove known variable stars.

Key Contributions

High-Cadence Survey Design: The paper demonstrates a targeted approach using DECam to achieve sub-minute cadence over a wide field, specifically optimized for the asteroid-mass PBH regime.
Finite-Source Analysis: The authors provide a detailed analysis of how finite-source effects impact detection efficiency. They show that for the LMC stellar distribution (which includes a significant population of giant stars), the maximum amplification is significantly damped for low-mass lenses ( $<10^{-8} M_\odot$ ), effectively shifting the survey's peak sensitivity toward the lunar-mass regime ( $10^{-8} - 10^{-6} M_\odot$ ).
Detection Pipeline: The development of a flexible statistical pipeline combining Von Neumann ratios, seeing-correlation metrics, and periodicity checks to isolate short-duration microlensing signals from atmospheric noise and stellar variability.
Wave Optics Limit: The paper establishes an absolute lower mass limit for the survey ( $9 \times 10^{-13} M_\odot$ ) based on wave optics diffraction effects, below which classical microlensing amplification is no longer valid.

Results

Candidate Detection: From five nights of data covering approximately two million quality light curves, the pipeline identified a single microlensing candidate on the fourth night (December 18). This candidate exhibited a full-width half-maximum timescale of $t_{FWHM} = 3.57$ hours.
Dark Matter Constraints:
- Under a null detection hypothesis (assuming the single candidate is not a PBH), AMPM constrains PBHs to constitute less than 30% of the Galactic dark matter density at the 95% confidence level, with the strongest constraint occurring at $M \approx 10^{-8} M_\odot$ .
- If the single candidate is treated as a genuine PBH detection, the constraint weakens to 55% of the dark matter density.
Efficiency: The survey achieved a maximum detection efficiency of $\epsilon \approx 0.6$ for PBHs around $10^{-5} M_\odot$ under optimal seeing conditions. Efficiency drops significantly for masses below $10^{-8} M_\odot$ due to finite-source damping and for masses above $10^{-3} M_\odot$ due to longer timescales exceeding the survey duration.

Significance and Claims
The paper claims that AMPM successfully demonstrates the feasibility of using ground-based, wide-field, high-cadence surveys to probe the asteroid-to-lunar mass range of PBH dark matter. The authors note that their single candidate event has a timescale consistent with the 12 candidates recently reported by the Subaru-HSC survey (Sugiyama et al. 2026), suggesting a potential population of short-timescale microlenses in the Galactic halo.

However, the authors are modest regarding the interpretation of this single event. They highlight an apparent tension with the OGLE-IV survey, which observed the LMC for 465 days with 16-minute cadence and found no such events. The paper suggests three possible resolutions to this discrepancy:

AMPM may have overestimated its efficiency.
OGLE-IV may have underestimated its efficiency at low masses due to its longer cadence.
Both AMPM and Subaru-HSC may be detecting a class of contaminating objects (e.g., flaring stars or free-floating planets) rather than PBHs.

The paper concludes that while finite-source effects severely limit the detection of sub-asteroid mass PBHs in the LMC, the detection of hour-long events supports the existence of a population of compact objects in the Galactic halo. The authors advocate for further high-cadence microlensing surveys to build population statistics and distinguish between baryonic compact objects and PBHs. They explicitly state that detecting PBHs below $10^{-12} M_\odot$ via microlensing in the LMC is likely prohibited by finite-source damping, suggesting that future progress in this mass regime may require space-based observations or targets with smaller source stars (e.g., white dwarfs).