Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos: a case for IC230724A with Subaru/HSC and prospects with Rubin/LSST

arXiv:2606.01158 · astro-ph.HE · Submitted 2026-05-31 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: I'm Vera.

Jocelyn: And I'm Jocelyn, here with Subrahmanyan.

Vera: We're looking at the paper: "Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos".

Jocelyn: Here's the gist of the paper. Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos: a case for IC230724A with Subaru/HSC and prospects with Rubin/LSST is a study that uses deep optical follow-up observations to investigate whether Tidal Disruption Events (TDEs) could be the source of cosmic high-energy neutrinos, finding no TDE candidate in the region of interest for a specific event.

Vera: Search Strategy and Methodology The researchers adopted a blind analysis policy to identify or disfavor tidal disruption events (TDEs) as cosmic neutrino sources The study first needed to set search criteria and estimate the number of unrelated transients and the true positive rate (TPR) of the target source using data outside the region of interest They followed a three-step selection process: (I) Imagequality cut where they remove fake detections caused by image subtraction, (II) Galactic star cut where they remove Galactic objects, and (III) photometric classification where they remove non-TDE variable objects For the photometric classification cut, seven criteria were introduced to distinguish TDEs from other transients/variables using their lightcurve features Data Reduction and Source Modeling The observations involved Target-of-Opportunity (ToO) imaging toward the localization region of IC230724A with Subaru/HSC on 2023 August 10 and 23 UT, resulting in Epoch 1 and Epoch 2 Data reduction involved standard processing using the HSC reduction pipeline, followed by image subtraction using an algorithm developed by Alard & Lupton (1998) A population of TDEs was tested as the neutrino source, constructed based on ZTF-detected TDE samples and modeled by a Gaussian rise and power-law decay lightcurve Candidate Selection Results The analysis applied the selection criteria to the background region (1.44 deg2) and estimated the TPR of the source TDE using both data-driven and simulation-driven methods The results showed that after applying all cuts, no TDE candidate is found in our source region after the final photometric classification cut The expected number of background objects in the source region is estimated to be 0.030±0.024 Future Constraints with Rubin/LSST The paper evaluates future prospects with Rubin/LSST, suggesting that if the survey covers the error regions of 10 neutrino events, it will enable constraints on the fractional contribution of TDEs to the cosmic high-energy neutrino background For non-detection, this fraction could be constrained as ≲ 60% and for detection as ≳ 30% The analysis indicates that current follow-up data are insufficient to constrain the hypothesis, but multiple follow-up campaigns near future will be able to put a meaningful constraint The study concludes that the current follow-up data are insufficient to constrain or confirm the TDE hypothesis by follow-up observations to a single neutrino event, even if TPR ∼ 0.5 was satisfied Robust constraints by the blind analysis can be obtained without performing ToO observations, as ToO observations might hinder to put robust constraint with the blind analysis The fractional contribution to cosmic neutrino background will be constrained either λ ≳ 0.6 or λ ≲ 0.3 for non-detection or detection, respectively, if 10 neutrino error regions are covered by Rubin/LSST survey The analysis found no TDE candidate in the region of interest The fractional contribution to cosmic high-energy neutrino background will be constrained either λ ≳ 0.6 or λ ≲ 0.3 for non-detection and detection, respectively, if Rubin covers the error regions of 10 neutrino events The results have significant implications for the follow-up strategy in the Rubin era; Robust constraints by the blind analysis can be obtained without performing ToO observations Table Summary Table 7 summarizes the number of objects during (III) photometric classification cut, showing that after all selection process, no objects are left in the source region Table 9 shows the values of TDE candidates in our source region during photometric classification cut, with zero remaining after selection The probability of realizing nT = 0 as a function of λ is shown in Figure 6, indicating that all the range, 0 ≤ λ ≤ 1, is allowed Table 7 also shows the TPR obtained by our simulated observations of (I) image-quality cut and (III) photometric classification cut The cumulative TPR for the entire selection process is estimated to be 0.11 ± 0.

Paper summary: Vera: So we're looking at this paper now titled "Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos: a case for IC230724A with Subaru/HSC and prospects with Rubin/LSST". The main idea here is using deep optical follow-up observations to see if tidal disruption events, or TDEs, could be the source of those cosmic high-energy neutrinos we detect.

Jocelyn: And what the paper claims is that they found no TDE candidate in the specific region they were looking at for this event. It's a test of a hypothesis about where these neutrinos come from.

Subrahmanyan: That’s interesting because it ties this observational search directly into the theoretical models for cosmic neutrino sources, which often involve things like TDEs or supernovae, as mentioned in their work on that paper <ref:2606.01158#pg2>.

Vera: Exactly. The whole point of this study is to use deep optical follow-up data to check if a TDE can be linked to the neutrino alert from IceCube, specifically for the event IC230724A <ref:2606.01158#pg3>.

Jocelyn: They set up a blind analysis policy, which means they didn't know beforehand if they were looking for a TDE or not when they started analyzing the data. That’s important for minimizing bias.

Subrahmanyan: Blind analysis is key because it lets them see what the data actually says without pre-determining what kind of object is interesting, which helps avoid confirmation bias <ref:2606.01158#pg2>.

Vera: They used a three step selection process to filter out noise and fake detections. First they had an image quality cut to remove things like artifacts from the subtraction process. Then a Galactic star cut to get rid of objects in our own galaxy. And finally, they had a photometric classification cut with seven criteria to distinguish TDEs from other transients or variables <ref:2606.01158#pg2>.

Jocelyn: They then used Target-of-Opportunity imaging with Subaru/HSC on August 10th and twenty-three UT for Epoch one and Epoch two data, which is how they gathered the actual optical images <ref:2606.01158#pg3>.

Subrahmanyan: The way they modeled the TDE population as a Gaussian rise and power-law decay lightcurve based on ZTF-detected samples gives them a concrete way to test their hypothesis <ref:2606.01158#pg2>.

Vera: After applying all those cuts, they found zero TDE candidates in their source region. They estimated that there were about zero point zero three zero plus or minus zero point zero two four background objects expected in that area <ref:2606.01158#pg3>.

Jocelyn: That means, based on the data they gathered, no TDE was found in the region of interest for this specific neutrino event <ref:2606.01158#pg3>.

Subrahmanyan: This result is significant because it directly constrains how much contribution TDEs might have to the cosmic high-energy neutrino background <ref:2606.01158#pg2>.

Vera: They then looked at what future surveys like Rubin/LSST could do. If that survey covers the error regions of ten neutrino events, it will give us constraints on how much TDEs contribute to the background <ref:2606.01158#pg2>.

Jocelyn: The paper suggests that if they don't find a TDE with future data, the fractional contribution could be constrained to be less than or equal to sixty percent <ref:2606.01158#pg2>. If they do find something, it could constrain it to be greater than or equal to thirty percent <ref:2606.01158#pg2>.

Subrahmanyan: It sounds like current follow-up data aren't enough on their own to really nail down this hypothesis, which is a bit disappointing for now, but it points toward future surveys being crucial <ref:2606.01158#pg2>.

Vera: They also found that robust constraints can be gotten just by the blind analysis without needing those specific Target-of-Opportunity observations <ref:2606.01158#pg3>.

Jocelyn: So, the implication is that we don't necessarily need to spend huge telescope time on every neutrino alert immediately to constrain this hypothesis <ref:2606.01158#pg3>.

Subrahmanyan: It means the constraint on the fractional contribution to the cosmic neutrino background will be either greater than or equal to zero point six if they find nothing, or less than or equal to zero point three if they find something <ref:2606.01158#pg2>.

Vera: That's a very clear way to put the future constraints down based on what Rubin/LSST could do <ref:2606.01158#pg2>.

Jocelyn: So, to wrap up this part, the whole point of this paper is that for this specific event, no TDE was found in the search area.

Subrahmanyan: And for the bigger picture, it sets up a clear roadmap for how future large surveys can constrain whether TDEs are a significant source of cosmic neutrinos <ref:2606.01158#pg2>.

Conclusion: Vera: So, this paper is all about using deep optical follow-up data to check if tidal disruption events could be the source of those cosmic high-energy neutrinos from IceCube, specifically for that event IC230724A.

Jocelyn: And what they found in that specific region was no TDE candidate, which is a pretty direct result of their search strategy.

Subrahmanyan: That lack of a TDE candidate in the target area means it doesn't strongly support that particular neutrino source hypothesis for this event right now.

Vera: Exactly. They ran a blind analysis policy to make sure they weren't just looking for something they expected to find, and after all their cuts—image quality, Galactic stars, photometric classification—nothing showed up in the area where the neutrino alert pointed.

Jocelyn: So it’s not a smoking gun for TDEs in this case then. But they’re not done with the story because they look ahead to future surveys like Rubin/LSST.

Subrahmanyan: That’s where it gets interesting, because even if we don't find one now, these large surveys could still tell us something important about how much TDEs contribute to the overall cosmic high-energy neutrino background.

Vera: Right. They’re suggesting that if Rubin/LSST covers the error regions of ten neutrino events, it will give us constraints on whether TDEs are a significant part of that background at all.

Jocelyn: They’re saying if they find nothing with future data, the fraction of TDEs could be less than or equal to sixty percent. If they *do* find one, it could constrain that fraction to be greater than thirty percent.

Subrahmanyan: It sounds like the current follow-up data isn't enough on its own to settle this hypothesis firmly. But their roadmap for future surveys is pretty clear about what we need from them.

Vera: And they also found that you can get robust constraints just by doing the blind analysis without needing those specific Target-of-Opportunity observations, which is a bit optimistic for a follow-up study, but it shows the power of that method.

Jocelyn: So, in short, this paper doesn't find the TDE we were hoping for in this search area and points to Rubin as the next big step for settling this whole question.

Subrahmanyan: The implication is that we need bigger surveys to really constrain whether TDEs are a major player in the cosmic neutrino background at all.

Frontier Research Institute for Interdisciplinary Sciences, Tohoku University · Astronomical Institute, Graduate School of Science, Tohoku University · Division for the Establishment of Frontier Sciences, Organization for Advanced Studies, Tohoku University · Astronomy Research Research Center, Chiba Institute of Technology · International Center for Hadron Astrophysics, Chiba University · National Astronomical Observatory of Japan National Institutes of Natural Sciences · Department of Astronomical Science School of Physical Sciences The Graduate University of Advanced Studies SOKENDAI Department of Physics Faculty of Science and Engineering Konan University

astro-ph.HE

Submitted: 2026-05-31

Updated: 2026-10-31

Comments: 18 pages, 7 figures, 9 tables, accepted for publication in ApJ

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 72/100

The gist: Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos: a case for IC230724A with Subaru/HSC and prospects with Rubin/LSST is a study that uses deep optical follow-up observations to…

Key concepts

Tidal Disruption Events (TDEs)
These are rare astronomical events where a star gets pulled apart by a supermassive black hole. The resulting material forms an accretion disk that can emit high-energy radiation, potentially producing cosmic neutrinos. Researchers modeled these events using lightcurves based on ZTF samples.
Blind Analysis Policy
The study used a blind analysis approach to avoid bias when searching for TDEs. This means the researchers didn't know which specific region was the source beforehand, allowing them to objectively test if any transients matched the expected TDE characteristics across the entire search area.
Photometric Classification Cut
This is a detailed filtering step used to distinguish true TDEs from other variable stars or transients. Seven specific criteria based on lightcurve features were introduced to ensure that only objects matching the expected behavior of a TDE were considered candidates, minimizing false positives.

Terminology

Summary

Deep Optical Follow-up Observations to IceCube Cosmic Neutrinos: a case for IC230724A with Subaru/HSC and prospects with Rubin/LSST is a study that uses deep optical follow-up observations to investigate whether Tidal Disruption Events (TDEs) could be the source of cosmic high-energy neutrinos, finding no TDE candidate in the region of interest for a specific event. p. 0

Search Strategy and Methodology

The researchers adopted a blind analysis policy to identify or disfavor tidal disruption events (TDEs) as cosmic neutrino sources The study first needed to set search criteria and estimate the number of unrelated transients and the true positive rate (TPR) of the target source using data outside the region of interest They followed a three-step selection process: (I) Imagequality cut where they remove fake detections caused by image subtraction, (II) Galactic star cut where they remove Galactic objects, and (III) photometric classification where they remove non-TDE variable objects For the photometric classification cut, seven criteria were introduced to distinguish TDEs from other transients/variables using their lightcurve features

Data Reduction and Source Modeling

The observations involved Target-of-Opportunity (ToO) imaging toward the localization region of IC230724A with Subaru/HSC on 2023 August 10 and 23 UT, resulting in Epoch 1 and Epoch 2 Data reduction involved standard processing using the HSC reduction pipeline, followed by image subtraction using an algorithm developed by Alard & Lupton (1998) A population of TDEs was tested as the neutrino source, constructed based on ZTF-detected TDE samples and modeled by a Gaussian rise and power-law decay lightcurve

Candidate Selection Results

The analysis applied the selection criteria to the background region (1.44 deg2) and estimated the TPR of the source TDE using both data-driven and simulation-driven methods The results showed that after applying all cuts, no TDE candidate is found in our source region after the final photometric classification cut The expected number of background objects in the source region is estimated to be 0. p. 7030±0.024

Future Constraints with Rubin/LSST

The paper evaluates future prospects with Rubin/LSST, suggesting that if the survey covers the error regions of 10 neutrino events, it will enable constraints on the fractional contribution of TDEs to the cosmic high-energy neutrino background For non-detection, this fraction could be constrained as ≲ 60% and for detection as ≳ 30% The analysis indicates that current follow-up data are insufficient to constrain the hypothesis, but multiple follow-up campaigns near future will be able to put a meaningful constraint

Conclusion and Implications

The study concludes that the current follow-up data are insufficient to constrain or confirm the TDE hypothesis by follow-up observations to a single neutrino event, even if TPR ∼ 0.5 was satisfied Robust constraints by the blind analysis can be obtained without performing ToO observations, as ToO observations might hinder to put robust constraint with the blind analysis The fractional contribution to cosmic neutrino background will be constrained either λ ≳ 0. p. 96 or λ ≲ 0.3 for non-detection or detection, respectively, if 10 neutrino error regions are covered by Rubin/LSST survey

Summary of Key Findings

The analysis found no TDE candidate in the region of interest The fractional contribution to cosmic high-energy neutrino background will be constrained either λ ≳ 0.6 or λ ≲ 0.3 for non-detection and detection, respectively, if Rubin covers the error regions of 10 neutrino events The results have significant implications for the follow-up strategy in the Rubin era; Robust constraints by the blind analysis can be obtained without performing ToO observations

Table Summary

Table 7 summarizes the number of objects during (III) photometric classification cut, showing that after all selection process, no objects are left in the source region Table 9 shows the values of TDE candidates in our source region during photometric classification cut, with zero remaining after selection The probability of realizing nT = 0 as a function of λ is shown in Figure 6, indicating that all the range, 0 ≤ λ ≤ 1, is allowed Table 7 also shows the TPR obtained by our simulated observations of (I) image-quality cut and (III) photometric classification cut The cumulative TPR for the entire selection process is estimated to be 0.11 ± 0.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved transient classification module that uses a set of seven criteria to distinguish TDEs from other transients/variables, including 5σ detection at t1, Long duration, and Monotonically declining lightcurve features. This allows the AI to classify detected optical events with higher precision based on photometric analysis.

  2. Improved background rate estimation by applying a blind analysis policy, as described in the paper: Under this policy, we first need to set search criteria and estimate the number of unrelated transients and the true positive rate (TPR) of the target source using the data outside the region of interest. This prevents bias in identifying potential neutrino sources.

  3. Improved statistical hypothesis testing for TDE origin by implementing a likelihood analysis comparing signal and alternative hypotheses, specifically using the log-likelihood ratio, or our test statistic, is given by Λ = −2 ln Lsig / Lalt to determine the fractional contribution of TDEs to the cosmic neutrino background.

  4. Improved future survey constraint prediction by using mock observations based on a given value of lambda and testing the probability of realizing nT = 0, as shown in Figure 6: If the probability is less than 0.1, lambda could be constrained. This allows for robust constraints on TDE fraction using projected data from Rubin/LSST.

  5. Improved ToO observation strategy recommendation by analyzing the trade-offs between current follow-up and future surveys: ToO observations might hinder to put robust constraint with the blind analysis. The AI can advise on whether single event follow-ups or wide, deep surveys are more effective for source population constraints.

Sources

Related papers