🔭 astrophysics

A Two-month, Galaxy-targeted NIR Follow-up of the Sub-solar-mass Gravitational Wave Candidate S251112cm: Probing Electromagnetic Counterpart Scenarios

This paper presents a two-month, galaxy-targeted near-infrared follow-up of the sub-solar-mass gravitational wave candidate S251112cm, which found no convincing electromagnetic counterparts and thereby constrains bright kilonova scenarios while establishing a host-weighted framework for future searches.

Original authors: Gregory S. H. Paek, Felipe Olivares E., Willem B. Hoogendam

Published 2026-08-25

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Gregory S. H. Paek, Felipe Olivares E., Willem B. Hoogendam

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

Gravity, in the way Albert Einstein described it, is not a force pulling objects together, but a warping of space and time itself. When massive objects like neutron stars or black holes spiral into each other and collide, they create violent ripples in this fabric, known as gravitational waves. For decades, these ripples were only a theoretical prediction, but in 2015, scientists finally heard them. Since then, the detection of these waves has opened a new window on the universe, allowing astronomers to listen to cosmic collisions that were previously invisible. However, listening is only half the story. To truly understand what happens during these collisions, scientists want to see the light they produce. This is the goal of multi-messenger astronomy: combining the sound of gravitational waves with the light of telescopes to get a complete picture of the event.

Most of the collisions detected so far involve objects with masses similar to our Sun or larger. But theory suggests there might be a hidden population of objects much lighter than the Sun, perhaps formed in the very early universe or through exotic processes. These "sub-solar-mass" objects would be a revolutionary discovery, proving the existence of new types of matter or even primordial black holes. The challenge is that we do not know what light these lighter collisions would produce. They might be faint, they might be delayed, or they might not produce light at all. Finding them requires a different kind of search than the one used for heavier objects, one that looks for faint glows over a longer period of time.

In late 2025, a gravitational wave detector network picked up a signal from a collision named S251112cm. The signal was strong and came from a relatively close distance of about 93 million light-years. What made this event special was that the data suggested at least one of the colliding objects was lighter than the Sun. This was the most significant candidate for such a light object ever found. The scientific community immediately began searching for the flash of light that might have accompanied the crash. While many teams looked for a bright, quick flash in visible light, a team led by Gregory Paek and colleagues decided to look for something different. They hypothesized that if this collision involved exotic, light objects, the light it produced might be faint, red, and slow to appear, lingering for weeks or even months.

To test this idea, the team turned to the United Kingdom Infrared Telescope, or UKIRT, located on a mountain in Hawaii. They chose to look in the near-infrared part of the spectrum, a range of light that is invisible to the human eye but is excellent for seeing through dust and spotting cool, fading objects. Instead of scanning the entire vast area of the sky where the collision might have happened, which would be too slow and shallow, they focused their efforts on specific galaxies. They created a list of the most likely host galaxies, ranking them based on how probable it was that the collision happened there and how much starlight the galaxy contained. They then pointed the telescope at 59 of these top candidates over a period of two months.

The team took deep images of these galaxies repeatedly, building up a clear picture of what was there before the collision and what was there after. They were looking for a new point of light that appeared and then slowly faded away. This was a patient game. They knew that if a bright explosion like a supernova had occurred, or if the collision produced a "kilonova"—a flash of light from heavy elements being forged—it would have been visible. Their observations were deep enough to see objects that were billions of times fainter than what the naked eye can see. They checked every single image, comparing the new photos with old ones to spot any changes.

After two months of careful watching, the result was clear: there was no new light. No bright flash, no fading glow, and no mysterious red spot appeared in any of the 59 galaxies they monitored. This absence of light is a significant finding in its own right. By not seeing anything, the team was able to rule out several specific scenarios. They found that if the collision had produced a bright explosion similar to a Type Ic supernova, or a standard kilonova like the one seen in 2017, it would have been easily visible. The fact that it was not seen means that such bright, standard events did not happen in the galaxies they watched.

The researchers also tested more complex ideas. Some theories suggested that if a light object merged inside a collapsing star, it might produce a "kilonova within a supernova," a fainter event that glows for a long time. They checked if their observations could have seen this kind of delayed, long-lasting glow. While they did not find it, their data showed that the brightest versions of these long-lasting events were also unlikely. The only scenarios that remained possible were those where the light was extremely faint, or where the event happened in a galaxy they did not happen to watch.

The team was careful to note the limits of their search. The sky area where the collision occurred is very large, and they only looked at a small fraction of the possible host galaxies. It is entirely possible that the collision happened in a galaxy they did not observe, or that the light was so faint that even their powerful telescope could not see it. They also acknowledged that the light might have appeared after their two-month campaign ended, or that the collision involved objects that simply do not produce light at all, such as primordial black holes.

Despite the lack of a discovery, this work provides a crucial blueprint for the future. It demonstrates that looking for these elusive, light-weight collisions requires a different strategy than looking for heavy ones. It requires patience, deep infrared vision, and a willingness to watch for months rather than days. By proving that bright, standard explosions are not happening in these specific locations, the team has narrowed down the possibilities for what these mysterious light objects might be. They have shown that if such objects exist and collide, their light is either very weak, very delayed, or hidden in a way that requires even more sensitive tools to find. This quiet, persistent search has not found the answer yet, but it has cleared the path for the next generation of astronomers to look deeper and longer, bringing us closer to understanding the true nature of the universe's lightest heavyweights.

Technical Summary: A Two-month, Galaxy-targeted NIR Follow-up of the Sub-solar-mass Gravitational Wave Candidate S251112cm

Problem and Motivation
The detection of the gravitational wave (GW) candidate S251112cm, characterized by a high probability ( $P_{\text{HasSSM}} = 100\%$ ) that at least one component is sub-solar mass ( $m \lesssim 1 M_\odot$ ), presents a unique challenge for electromagnetic (EM) follow-up. Unlike standard binary neutron star (BNS) mergers, the EM counterparts of sub-solar-mass (SSM) mergers are theoretically predicted but unconfirmed. Specific formation channels, such as collapsar disk fragmentation, predict distinct emission signatures: a "kilonova-within-a-supernova" (KN-in-SN) scenario producing a late-time Near-Infrared (NIR) excess over weeks to months, or a "super-kilonova" (super-KN) with a redder, longer-lived peak. Previous follow-up efforts for SSM candidates (e.g., S250818k) and S251112cm itself were predominantly optical and confined to short baselines (days post-trigger), leaving the predicted late-time NIR emission phase unconstrained. This work addresses the need for deep, long-baseline NIR monitoring to test these specific theoretical scenarios.

Methodology
The authors conducted a targeted J-band ( $\lambda_{\text{eff}} \approx 1.25 \, \mu\text{m}$ ) monitoring campaign using the UKIRT Wide-Field Camera (WFCAM) over approximately two months (2025 November 13 – 2026 January 17).

Target Selection: Due to the broad GW localization volume ( $\sim 1681 \, \text{deg}^2$ ), a blind search was deemed inefficient. Instead, the team prioritized 59 host galaxy candidates drawn from the NED gravitational-wave galaxy catalog. Prioritization was based on a ranking metric combining the 3D localization probability ( $P_{3D}$ ) and the WISE W1 luminosity ( $L_{W1}$ ) as a proxy for stellar mass, assuming SSM merger rates trace stellar mass similarly to BNS/NSBH systems.
Observations: The campaign achieved a typical 5 $\sigma$ depth of $J \sim 22\text{--}23$ mag (AB). Observations were performed at a cadence of several days, with higher frequency in the early phase.
Data Processing: Standard reduction pipelines were used, followed by image subtraction using HOTPANTS. To ensure robustness against transient contamination in reference images, a brute-force multi-reference cross-check was performed.
Analysis Framework: The study employed a "host-weighted" framework. Constraints were not claimed as event-wide exclusions but were quantified as conditional probabilities: "if the source lies within the monitored subset, what fraction of the parameter space is excluded?" This involved comparing non-detections against:
1. A canonical kilonova model grid (R. T. Wollaeger et al. 2021) spanning various ejecta masses and velocities.
2. Empirical templates: a Type Ic-BL supernova (SN 2007I) and the AT 2017gfo kilonova light curve (including time-shifted variants up to $\Delta t \sim 60$ days).

Key Results

Non-detection: Difference imaging revealed no convincing transient counterpart associated with the 59 monitored hosts.
Coverage: The 59 monitored hosts represent 7.7% of the cumulative weighted host prior ( $P_{3D} \times L_{W1}$ ). Consequently, all constraints are conditional on the GW source residing within this specific subset.
Model Constraints:
- Canonical Kilonovae: Bright configurations with high dynamical and wind-ejecta masses are strongly disfavored within the monitored subset. However, less luminous configurations remain permissible.
- Template Comparison: At the distance of S251112cm (93 Mpc), both a Type Ic-BL SN template and an AT 2017gfo-like KN template (including delayed variants up to 60 days) would have remained above the median detection depth for multiple epochs. Their non-detection suggests that if a counterpart exists in the monitored hosts, it is either intrinsically fainter than these benchmarks or delayed beyond the campaign's baseline.
- Epoch Dependence: The constraining power is heavily time-dependent. The strongest constraints on canonical KN models were provided by hosts observed within $\sim 0.25\text{--}7$ days post-trigger. Late-time observations contributed less to excluding canonical models but were necessary to probe the specific late-time NIR excess predicted by KN-in-SN scenarios.

Significance and Claims
The paper positions this work as the first empirical case study of long-baseline, deep NIR follow-up for a sub-solar-mass merger GW candidate.

Framework Validation: It demonstrates a host-weighted framework applicable to future galaxy-targeted campaigns, allowing for the quantification of excluded parameter space even when full sky coverage is impossible.
Physical Implications: The results do not rule out all SSM scenarios. The non-detection is consistent with several interpretations:
1. The GW source resides outside the monitored 7.7% of the prior.
2. The counterpart is intrinsically fainter than the benchmark models (e.g., a low-mass ejecta KN).
3. The emission is delayed beyond the two-month baseline.
4. The progenitor is a non-EM channel, such as a primordial black hole (PBH) merger.
5. The counterpart is a KN-in-SN where the SN ejecta reprocesses the NIR emission, or dust extinction suppresses the flux (though the latter requires extreme extinction values).
Strategic Recommendations: The authors conclude that future SSM follow-up requires a hybrid strategy: early deep NIR observations to constrain canonical KN emission, followed by longer-baseline monitoring for super-KN and KN-in-SN scenarios. They also suggest that archival revisitation of continuous all-sky survey data (e.g., Pan-STARRS, ATLAS, Rubin/LSST, SPHEREx) is a critical, immediately accessible complement to targeted follow-up to extend the search window for slowly evolving transients.

The paper maintains a modest tone, explicitly stating that the constraints are host-limited and that the assumption of stellar-mass tracing for SSM progenitors remains theoretically unverified. It does not claim to have identified the nature of S251112cm but rather establishes the observational limits and methodological approach for testing specific EM counterpart hypotheses.

More like this