IceCube Second Track Data Release IceTracks-DR2: Data from 2008-2022 for Neutrino Source Searches

arXiv:2605.19040 · astro-ph.HE, hep-ex · Submitted 2026-05-18 · 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: We're looking at the paper: "IceCube Second Track Data Release IceTracks-DR2".

Title and authors: Vera: We're talking about the IceCube Second Track Data Release IceTracks-DR2: Data from two thousand eight-two thousand twenty-two for Neutrino Source Searches <ref:2605.19040#pg1,IceCube Second Track Data Release IceTracks-DR2: Data from 2008-2022 for>. This paper is essentially laying out the new data we get from the IceCube detector over a much longer stretch of time.

Jocelyn: It's important to note that the authors include a lot of people, like R. Abbasi, M. Ackermann, J. Adams, and many others on page one of this work <ref:2605.19040#pg1>. They are all contributing to compiling this data and the analysis methods described in the paper itself.

Subrahmanyan: The sheer volume of data accumulated over those fourteen years is what gives us the statistical power to make these kinds of high significance claims for point sources.

Vera: And they're focusing on how this extended observation period allows for a better understanding of steady and transient neutrino sources than before.

Jocelyn: It’s about seeing if we can use this extended time baseline to find signals that might be too faint or too short-lived to catch with shorter observations.

Subrahmanyan: The implications are that we gain better constraints on the physics of these neutrino sources, whether they are steady emitters or something that flares up suddenly.

Vera: So, when you look at this paper, you’re seeing how the authors have put together a massive observational effort to build a more complete picture of neutrino emission history.

Jocelyn: It’s about leveraging that continuous data stream to find those elusive signals that might be changing in nature.

The paper's summary: Vera: The summary of the IceTracks-DR2 paper emphasizes that this release features an updated event selection and improved detector calibration for data recorded after June one two thousand ten <ref:2605.19040#pg1>.

Jocelyn: Beyond the technical updates, the authors highlight that they've included binned instrument response functions and effective areas which enable the community to perform sensitive searches for steady and transient neutrino sources.

Subrahmanyan: They frame this as IceCube’s most sensitive and comprehensive publicly available all-sky muon track dataset to date, which is a big statement about the quality of the data provided.

Vera: That means they explicitly state that this release should be preferred over previous releases because it offers better tools for the search community.

Jocelyn: They also stress that the accompanying detector response information is relevant for accurate interpretation and modeling of those observed event properties.

Subrahmanyan: So, in essence, the summary says: this is a high-quality data package with updated methods and models designed to maximize our ability to find neutrino sources.

Vera: It’s not just more data; it’s a better way to use the existing data we have gathered so far.

Jocelyn: They're making it clear that the methodology is as important as the raw counts when you are trying to interpret these results for astrophysical understanding.

The paper's improvements: Vera: Looking at what they suggest as improvements, one key thing is how their AI system can perform sensitive searches for steady and transient neutrino sources by utilizing those binned instrument response functions and effective areas included in IceTracks-DR2.

Jocelyn: That’s really smart because it means the AI isn't just doing a simple count; it’s using detailed modeling to search for things that vary over time.

Subrahmanyan: This capability is what lets the system provide reproducible research by offering benchmark results obtained using this data release and publicly available software, which aids in the development of new methods for astrophysical neutrino observations.

Vera: That reproducibility is vital; it means other researchers can trust the results and use them to develop their own techniques without starting from scratch.

Jocelyn: Then there's the search capability for time dependent searches, where you maximize a likelihood function that includes an additional temporal probability density function for both signal and background, allowing it to compute a test statistic maximized over n s, gamma, T zero and delta t.

Subrahmanyan: That maximization across those four parameters tells us exactly how the source might be changing over time based on the data we have.

Vera: And they can also perform diffuse Galactic plane emission hypothesis testing by constructing spatial probability density functions using models like Fermi LAT pi zero, and comparing those against templates such as KRA5 gamma and KRA50 gamma.

Jocelyn: That comparison helps us figure out if the diffuse signal we see in our galaxy matches what we expect from other astrophysical processes.

Subrahmanyan: The paper also allows the AI to validate detector response information by reproducing benchmark results using the SkyLLH package, which lets you construct signal and background probability density functions directly from released IRFs.

Vera: That validation step is crucial because it confirms that the models we are using to describe the detector response are actually accurate representations of what we see.

Conclusion: Jocelyn: So, to wrap up, this IceTracks-DR2 release provides a much larger dataset with enhanced selection and calibration methods for neutrino source searches over a fourteen year period.

Vera: The main implication is that the inclusion of binned instrument response functions allows the AI system to perform sensitive searches for both steady and transient neutrino sources effectively.

Subrahmanyan: This leads to more powerful time dependent searches, where the AI can maximize a likelihood function over multiple parameters including n s, gamma, T zero and delta t.

Jocelyn: Plus, we have tools for testing diffuse emission hypotheses against models like KRA5 gamma and KRA50 gamma.

Vera: And they provide validation tools to check the detector response using packages like SkyLLH to reproduce benchmark results from the IRFs.

Subrahmanyan: Overall, this IceTracks-DR2 release is a major step forward in providing a comprehensive resource for studying neutrino sources across various time scales and spatial scales.

Jocelyn: It’s about giving the community better tools to interpret these long-term observations and plan the next steps for discovery.

Vera: This paper really sets a high bar for what we expect from future data releases in this field, building on this solid foundation.

Subrahmanyan: I think the IceTracks-DR2 release is a foundational piece that will be used to push these searches into deeper cosmological and astrophysical questions about neutrino emission.

IceCube Collaboration

astro-ph.HE, hep-ex

Submitted: 2026-05-18

Updated: 2026-11-17

Code: https://github.com/icecube/skyllh

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

Importance score: 78/100

The gist: IceTracks-DR2 presents IceCube’s latest release of muon track data for neutrino point-source searches…

Key concepts

IceTracks-DR2
This is the latest release of muon track data from IceCube, covering 14 years of observations (April 6, 2008 – May 23, 2022). It represents the most comprehensive all-sky muon track dataset available for neutrino point-source searches.
Instrument Response Function (IRF) Matrix
This is a detailed mathematical tool used to model how IceCube detects events. The IRF matrix describes the detector's response across different parameters like neutrino energy, direction, and reconstructed track properties, allowing scientists to accurately interpret observed event characteristics.
Boosted Decision Tree (BDT)
A machine learning algorithm used in the data selection process. It is trained to distinguish between true neutrino-induced track events and background events like cascade-like events by evaluating various event qualities.

Terminology

Summary

IceTracks-DR2 presents IceCube’s latest release of muon track data for neutrino point-source searches, extending the previously published 10-year dataset to cover 14 years of observations (April 6, 2008 – May 23, 2022) IceTracks-DR2 presents IceCube’s latest release of muon track data for neutrino point-source searches, extending the previously published 10-year dataset to cover 14 years of observations (April 6, 2008 – May 23, 2022)

Data Release Scope and Updates

This release features an updated event selection and improved detector calibration for data recorded after June 1, 2010 The release also includes binned instrument response functions and effective areas, enabling the community to perform sensitive searches for steady and transient neutrino sources This release represents IceCube’s most sensitive and comprehensive publicly available all-sky muon track dataset to date and should be preferred over previous releases The dataset is accompanied by relevant detector response information to enable accurate interpretation and modeling

Event Selection Methodology

The event selection begins with a set of filters that selects for events with muon tracks from all directions that could be induced by neutrinos Events in the Northern hemisphere (declination ≥ -5 degrees) are selected based on event quality properties including the number of DOMs illuminated and the length of the reconstructed track The southern-sky selection is tuned to produce a similar event rate as the northern sky event selection, resulting in a final all-sky event rate of approximately 4 mHz Events from both hemispheres are then evaluated by a boosted decision tree (BDT) trained to select for track-like events while rejecting cascade-like events

Detector Response Characterization

Detailed simulation is used to evaluate the response of IceCube to the various types of events seen by the detector The relevant simulations can be characterized by the detector effective area (Aeff) and a set of functions describing the detector response, characterized as a 5-dimensional response matrix (the “Instrument Response Function or “IRF matrix) spanned by neutrino energy, neutrino declination, reconstructed event (muon) energy, point spread function, and estimated angular uncertainty The Instrument Response Function matrix allows for modeling of observed event properties for simulated events with known true energy and direction

Search Results and Comparisons

The results of the time-integrated all-sky point-source search are from the analysis described in (R. Abbasi et al. p. 0 p. 0 2025a) The hottest spot in the northern sky is located at δ = −0. p. 830◦ and R.A. = 40.8◦ The fitted values for the all-sky, time-integrated search are nˆs = 69 and γˆ = 3. p. 73 The highest-significance source in the catalog is NGC 1068, with a pre-trial p-value of 2. p. 907 × 10−6 (4.6σ) The results from R. p. 9 Abbasi et al.

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 AI system can perform sensitive searches for steady and transient neutrino sources by utilizing binned instrument response functions and effective areas included in IceTracks-DR2, as these are enabling the community to perform sensitive searches for steady and transient neutrino sources.

  2. The system can provide reproducible research by providing benchmark results obtained using this data release and publicly available software, which aids in the development of new methods for astrophysical neutrino observations.

  3. The AI system can conduct time-integrated point-source searches across the all-sky map, calculating a Test Statistic (TS) as defined by the ratio of the likelihood under the null hypothesis (ns = 0) to the likelihood at the best-fit signal parameters.

  4. The system can perform time-dependent searches by maximizing a likelihood function that includes an additional temporal PDF for both signal and background, allowing it to compute a TS that is maximized over ns, γ, T0 and ∆T.

  5. The system can perform diffuse Galactic plane emission hypothesis testing by constructing spatial PDFs using models like Fermi-LAT pi 0 and comparing the resulting test statistics against templates such as KRA5 gamma and KRA50 gamma.

  6. The AI system can validate detector response information by reproducing benchmark results using the SkyLLH package, which allows for the construction of signal and background probability density functions directly from released IRFs.

  7. The system can compare its sensitivity to published results by performing likelihood scans around best-fit parameters and comparing them to existing analyses, showing that the two overlap closely and exhibit the same shape as a function of the two parameters.

Sources

Related papers