🔭 astrophysics

ASAS-SN Rates IV: Constraints on the Kilonova Rate

Using 11 years of ASAS-SN survey data and injection-recovery simulations, this study establishes a competitive 95% upper limit on the local kilonova rate of 4400 yr⁻¹ Gpc⁻³, which remains significantly higher than the binary neutron star merger rate inferred from gravitational wave observations.

Original authors: Dhvanil D. Desai, Benjamin J. Shappee, Christopher S. Kochanek, Krzysztof Z. Stanek, Katie Auchettl, John F. Beacom, Jeff Cooke, Subo Dong, Willem B. Hoogendam, Jose L. Prieto, Todd A. Thompson, Micha

Published 2026-03-31

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Dhvanil D. Desai, Benjamin J. Shappee, Christopher S. Kochanek, Krzysztof Z. Stanek, Katie Auchettl, John F. Beacom, Jeff Cooke, Subo Dong, Willem B. Hoogendam, Jose L. Prieto, Todd A. Thompson, Michael A. Tucker, Natasha Van Bemmel

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 Great Cosmic Hunt: Why We Haven't Seen a "Kilonova" Yet

Imagine the universe as a giant, dark ocean. Most of the time, it's quiet. But occasionally, two massive ships—neutron stars (the super-dense corpses of dead stars)—crash into each other. When they collide, they don't just make a splash; they create a spectacular, short-lived fireworks display called a Kilonova.

These explosions are the universe's "gold factories." They are the primary reason we have heavy elements like gold, platinum, and uranium in our jewelry and electronics. But here's the problem: We don't know how often these fireworks happen.

This paper is like a report from a very dedicated, very patient lighthouse keeper who has been scanning the entire ocean for 11 years, trying to count these fireworks.

1. The Detective and the Flashlight

The team behind this study uses a project called ASAS-SN. Think of ASAS-SN as a giant, automated flashlight that sweeps across the entire night sky every single night. It's not the most powerful flashlight in the world (it can't see very faint, distant objects), but it is relentless. It never sleeps, it never stops, and it covers the whole sky.

Other telescopes are like high-powered spotlights: they can see very far and very clearly, but they only look at a tiny patch of the sky at a time. ASAS-SN is the "wide-angle lens" that watches everything, hoping to catch something bright and nearby.

2. The Search Strategy: "If It Happened, We'd See It"

The scientists asked a simple question: "If a Kilonova happened near us in the last 11 years, would our flashlight have seen it?"

To answer this, they didn't just wait and hope. They ran a massive computer simulation. Imagine they took a digital version of their telescope and "injected" thousands of fake Kilonova explosions into their data. They used a specific recipe for what a Kilonova looks like, based on the only real one we've ever confirmed (a famous event in 2017 called SSS17a).

They checked:

Did the weather block the view?
Was the moon too bright?
Did the telescope miss it because it was moving too fast?

They found that if a "bright, blue" Kilonova (like the 2017 one) happened within about 60 million light-years of Earth, ASAS-SN would have spotted it almost every time.

3. The Result: The Great Silence

After checking 11 years of data and running millions of simulations, the team found zero Kilonovae.

This doesn't mean they failed. In science, finding nothing is actually a very powerful result. It's like a fisherman casting a net in a lake for 11 years and catching zero fish. He can't say, "There are no fish in the lake." But he can say, "If there are fish, they are either very rare, or they are hiding in the deep, dark parts of the lake where my net can't reach."

Because they found nothing, they calculated an upper limit. They can now say with 95% confidence:

"There are fewer than 4,400 of these explosions happening in the entire observable universe every year."

4. The Big Picture: Why Does This Matter?

The scientists compared their result to two other ways we try to count these events:

Gravitational Waves (The "Rumble"): When neutron stars crash, they send out ripples in space-time called gravitational waves. The LIGO/Virgo detectors "hear" these rumbles. They estimate there are about 250 to 1,700 crashes per year.
Gamma-Ray Bursts (The "Flash"): Sometimes, these crashes are followed by a burst of high-energy light. Astronomers count those to guess the rate.

The Conflict:
The ASAS-SN team's "no sightings" limit is about 18 times higher than the number of crashes LIGO thinks are happening.

Analogy: Imagine LIGO is a microphone that hears 100 car crashes a year. ASAS-SN is a security camera that sees the wreckage. If the camera sees zero wrecks, but the microphone hears 100 crashes, it means either:
- The crashes are happening in a place the camera can't see (too far away or too dim).
- The crashes are happening, but they aren't making the "fireworks" we expect (maybe they are "red" and dim instead of "blue" and bright).
- Or, the microphone is overestimating the number of crashes.

5. The Takeaway

This paper is a crucial piece of the puzzle. It tells us that while we are getting better at hearing the "rumble" of neutron star crashes, we still haven't caught enough of the "fireworks" to be sure how often they happen.

The ASAS-SN team proved that even with a "modest" telescope that isn't the most powerful in the world, consistency wins. By watching the whole sky, every night, for over a decade, they set a strict rule: If bright Kilonovae were common nearby, we would have seen them by now.

Since we didn't, the universe is either more mysterious (hiding its fireworks) or the "rumble" detectors need to be recalibrated. Either way, we are one step closer to understanding how the universe makes the gold in our rings.

1. Problem Statement

Kilonovae (KNe) are the electromagnetic counterparts to binary neutron star (BNS) mergers and are believed to be the primary site of cosmic r-process nucleosynthesis. Despite the confirmed detection of GW170817/SSS17a (AT 2017gfo), the intrinsic volumetric rate of KNe remains poorly constrained.

Current Limitations: Estimates derived from short Gamma-Ray Bursts (GRBs) and Gravitational Wave (GW) detections vary by more than an order of magnitude.
Survey Gaps: Previous optical surveys (ATLAS, DES, PTF, ZTF, KNTraP) have failed to confirm any KNe, providing only loose upper limits (ranging from $\sim 800$ to $180,000 \text{ yr}^{-1} \text{ Gpc}^{-3}$ ). These surveys often lack the continuous, high-cadence monitoring required to catch fast-evolving transients or are limited by their operational triggers.
Objective: To place a robust, independent constraint on the local volumetric rate of bright, nearby KNe using the All-Sky Automated Survey for Supernovae (ASAS-SN) over an 11-year baseline (2014–2024).

2. Methodology

The authors employed a rigorous injection-recovery simulation framework to determine the survey's completeness and derive rate limits.

A. Light Curve Modeling

Template Selection: The study utilized the shock-cooling cocoon model by Piro & Kollmeier (2018). This model was chosen because it best explains the early, blue optical emission of SSS17a (AT 2017gfo), the only well-sampled KN associated with a GW event.
Relevance: ASAS-SN operates in the $V$ and $g$ bands with a limiting magnitude of $g \sim 18$ mag. It is primarily sensitive to the "blue" component of KNe (lanthanide-poor ejecta) which peaks in the UV/optical on timescales of hours to days.
Templates: Normalized light curves for $V$ and $g$ bands were generated, shifted so the peak occurs at $t=0$ .

B. Completeness Analysis (Injection-Recovery)

Baseline Data: Light curves were extracted at 105 random sky coordinates to empirically capture real observing conditions (noise, cadence, weather gaps, seasonal gaps).
Injection: Simulated KN templates were injected into these baselines.
Recovery Criteria: Instead of a simple signal-to-noise ( $>5\sigma$ $> 5 σ$ ) threshold, the authors used empirically derived efficiency curves ( $p(\text{SNR})$ $p (SNR)$ ) specific to ASAS-SN's pipeline.
- V-band Era (2014–2017): Peak recovery efficiency $p \approx 0.65$ .
- g-band Era (2018–2024): Peak recovery efficiency $p \approx 0.48$ (averaged), accounting for operational changes (e.g., pandemic impacts).
Simulation Parameters:
- $10^6$ simulated sources uniformly distributed in comoving volume ( $0.001 < z < 0.08$ ).
- Peak absolute magnitudes drawn from $M_{\text{peak}} \in [-16.5, -15.5]$ mag (centered on SSS17a).
- Galactic latitude cut: $|b| > 15^\circ$ to avoid high extinction.
- The 11-year baseline was divided into eight epochs (one for V-band, seven yearly for g-band).

C. Candidate Verification

The authors systematically re-examined all unclassified transients flagged by the ASAS-SN pipeline.
Filtering: Spectroscopically confirmed objects, high-confidence supernovae, and sources with stellar counterparts (cross-matched with Gaia, SDSS, WISE, etc.) were removed.
Host Association: Remaining candidates were cross-matched with the REGALADE catalog of local galaxies within 60 Mpc.
Result: No viable KN candidates were found. Six specific unclassified transients were analyzed and rejected due to slow evolution, single-epoch artifacts, or inconsistent light curve morphology (see Appendix A).

3. Key Results

A. Upper Limit on Volumetric Rate

Based on zero detections over the 11-year baseline, the authors calculated the total effective volume-time ($VT$) and derived the following upper limits for "blue," SSS17a-like events:

2 $\sigma$ (95% Confidence): $R_{\text{KN}} < 4400 \text{ yr}^{-1} \text{ Gpc}^{-3}$ .
1 $\sigma$ (68% Confidence): $R_{\text{KN}} < 1700 \text{ yr}^{-1} \text{ Gpc}^{-3}$ .
3 $\sigma$ (99.7% Confidence): $R_{\text{KN}} < 8500 \text{ yr}^{-1} \text{ Gpc}^{-3}$ .

B. Contextual Comparisons

Vs. GW Estimates: The 2 $\sigma$ limit is a factor of 18 higher than the latest LIGO-Virgo-KAGRA (GWTC-4) estimate of the BNS merger rate ( $7.6 - 250 \text{ yr}^{-1} \text{ Gpc}^{-3}$ ).
Vs. Other Optical Surveys: The ASAS-SN limit is competitive with the strongest constraints from ZTF, ATLAS, and KNTraP, falling within an order of magnitude of their limits despite ASAS-SN's shallower depth ( $g \sim 18$ vs. $g \sim 21-22$ for others).
Vs. Supernovae: The KN rate is constrained to be at least 6 times lower than Type Ia SNe rates and 16 times lower than core-collapse SNe rates in the local universe.
Per Galaxy Rate: Translated to Milky Way-like ( $L^*$ ) galaxies, the rate is $< 0.13 \text{ century}^{-1} (L^* \text{ galaxy})^{-1}$ .

C. Horizon Distance

For an event with $M_{\text{peak}} = -16$ mag, ASAS-SN's effective horizon is approximately 60 Mpc (roughly 20 Mpc further than the distance to SSS17a).

4. Significance and Discussion

Unique Niche: This study highlights the unique value of ASAS-SN. While other surveys go deeper, ASAS-SN provides continuous, high-cadence, all-sky monitoring over a decade. This is critical for detecting fast-evolving transients that might be missed by surveys with lower cadence or gaps in coverage.
Model Dependence: The authors acknowledge that the limit is model-dependent. It assumes the KN population resembles the "blue" component of SSS17a. If a significant fraction of KNe are "red" (lanthanide-rich, fading quickly in optical bands) or intrinsically fainter, the actual rate could be higher, and the ASAS-SN efficiency would be lower.
Operational Lesson: The paper notes that SSS17a itself was not detected by ASAS-SN not due to sensitivity, but due to operational constraints (limited sites operational and poor visibility windows at the time of the merger), validating the need for the robust injection-recovery simulations used in this study.
Future Outlook: While ASAS-SN provides a competitive constraint, the authors note that next-generation facilities like the Vera Rubin Observatory will be required to detect tens of un-triggered KNe annually to fully characterize the luminosity function and color distribution of the population.

Conclusion

The paper establishes a robust, model-based upper limit on the local kilonova rate using a decade of ASAS-SN data. By finding zero events, it constrains the rate to $< 4400 \text{ yr}^{-1} \text{ Gpc}^{-3}$ (95% confidence). This result is consistent with gravitational wave estimates but remains significantly higher, suggesting that either the population of KNe is diverse (with many "red" or faint events undetectable by ASAS-SN) or that the GW merger rate estimates may still be subject to large uncertainties. The study underscores the importance of long-baseline, high-cadence surveys in constraining transient rates independent of gravitational wave triggers.