IceCube High Energy Starting Events at 7.5 Years -- New Measurements of Flux and Flavor

TL;DR

This paper presents new measurements of the flux and flavor composition of high-energy astrophysical neutrinos using 7.5 years of IceCube data, providing insights into their sources and production mechanisms.

Contribution

It introduces a new flavor composition measurement of astrophysical neutrinos using the IceCube High-Energy Starting Event sample over 7.5 years, enhancing understanding of neutrino sources.

Findings

Measured flavor ratio on Earth constrains source models.

Detected tau neutrino interactions above 100 TeV.

Provides improved flux estimates of astrophysical neutrinos.

Abstract

The IceCube Neutrino Observatory at the South Pole, which detects Cherenkov light from charged particles produced in neutrino interactions, firmly established the existence of an astrophysical high-energy neutrino component. Here I present IceCube's High-Energy Starting Event sample and the new results obtained with a livetime of about 7.5 years. I will focus on the new measurement of the flavor composition performed using this sample. IceCube is directly sensitive to each neutrino flavor via the single cascade, track and double cascade event topologies, the latter being the topology produced in tau-neutrino interactions above an energy threshold of ~100 TeV. A measurement of the flavor ratio on Earth can provide important constraints on sources and production mechanisms within the standard model, and also constrain various beyond-standard-model processes.

Tables

Table 1. Table 1: Observables of the two Double Cascades

	Event#1	Event#2
Energy of 1st cascade	1.2 PeV	9 TeV
Energy of 2nd cascade	0.6 PeV	80 TeV
Energy Asymmetry	0.29	-0.80
Length	16 m	17 m

Equations

$$\Phi =2.19_{-0.55}^{+1.10}\cdot 10^{-18}\cdot (\mathrm{E}/100\mathrm{TeV}{)}^{-2.91_{-0.22}^{+0.33}}{\mathrm{GeV}\mathrm{cm}}^{-2}{\mathrm{s}}^{-1}{\mathrm{sr}}^{-1}$$

Full text

1 The High Energy Starting Events With 7.5 Years of Livetime

The IceCube collaboration IC_det has reported the observation of a diffuse flux of astrophysical neutrinos using the High-Energy Starting Events (HESE) selection Science . The HESE selection has been applied to more data collected since then, with the last update having a livetime of 6 years HESE6 . In the meantime, the understanding of the detector, as well as the atmospheric backgrounds to the astrophysical neutrino searches have improved. This has triggered a full re-analysis of IceCube data using the HESE event selection with 7.5 years of data and improved techniques. Older data has been reprocessed using a new and improved detector calibration, and 1.5 years of new data have been added. The latest knowledge of the ice optical properties has been incorporated into the simulation chain and the reconstruction updated, and a newly developed likelihood taking limited Monte Carlo statistics into account is being used. A new calculation of the atmospheric self-veto APRSWY is used, describing the probability of atmospheric neutrinos to be rejected due to accompanying muons from the same cosmic ray induced shower. Further, a designated algorithm to determine the topology of the events Marcel has been incorporated. Not only is this superior to the previously used by-eye classification, it can also distinguish three topologies: the well known single cascades created by all-flavor Neutral Current (NC) interactions, Charged Current (CC) $\nu_{e}$ and CC $\nu_{\tau}$ interactions; the tracks created by CC $\nu_{\mu}$ interactions as well as atmospheric muons, and CC $\nu_{\tau}$ interactions where the tau lepton decays muonically; and the double cascades. The latter topology opens up at high energies, when a CC $\nu_{\tau}$ interaction can produce two cascades that can be individually resolved in space and time Learned . The 7.5 year HESE sample has 60 events with reconstructed energies above 60 TeV, with 12 (5) new events observed in the 2016 (2017) data taking season. The best fit of the HESE events with a reconstructed deposited energy above 60 TeV is a single power-law with

[TABLE]

per neutrino flavor. The best fit yields a vanishing prompt component, with the 90% upper limit being 12.3 times the theoretical prediction in the BERSS model BERSS . Our results are compatible with the previous results using in the 6-year HESE update HESE6 as well as with the fit performed on the through-going muon TGM sample.

2 Astrophysical Neutrino Flavor Composition

Most previous flavor composition measurements did not distinguish between single and double cascades, leading to a large degeneracy between the $\nu_{e}$ and $\nu_{\tau}$ flavors. In this analysis we identify the first two double cascades, on an expectation of $\sim 2$ (1.4 signal and 0.7 background) events. To classify the events with deposited energies above 60 TeV we use an algorithm first applied to the 6 year HESE sample Marcel , developed with the goal of achieving a high $\nu_{\tau}$ purity in the final double cascade topology sub-sample. Observables used for classification are the double cascade length, the asymmetry between the two cascades’ energies (called energy asymmetry herafter), and the fraction of the total energy deposited close to the cascade vertices (called energy confinement herafter). Events passing the quality, energy asymmetry and energy confinement cuts are classified as double cascades. Above 60 TeV reconstructed deposited energy we classify the 60 events into 42 single cascades, 16 tracks, and 2 double cascades. A multi-component maximum-likelihood fit is performed on the three topology samples using two-dimensional PDFs obtained from Monte Carlo simulations. For single cascades and tracks, the observables Deposited Energy and Cosine $\theta_{z}$ are used, where $\theta_{z}$ is the zenith angle. For double cascades, Total Deposited Energy and Double Cascade Length are used. For $\nu_{\tau}$ induced double cascades, we expect a correlation between the energy of the event and the tau decay length. Events stemming from other flavors typically cluster at the thresholds, both in energy due to the falling spectrum and in double cascade length due to the very small mean reconstructed double cascade length for true single cascades. The PDFs for signal and background are shown in Figure 1 with the data events overlaid as white circles. The result of the flavor composition measurement and the sensitivity are shown in Figure 2. Note that not all systematic uncertainties have been taken into account yet in this work in progress. This analysis yields $\nu_{e}:\nu_{\mu}:\nu_{\tau}=0.29:0.50:0.21$ as the best fit flavor composition. This is consistent with previously published results by IceCube APJ15 ; Spencer ; Marcel , as well as with the expectation of $\sim$ 1 : 1 : 1 flavor composition on Earth coming from a pion decay production mechanism.

2.1 A Closer Look at the Double Cascades

The two observed double cascades are shown in Figure 3. The grey circles mark the reconstructed cascade positions, with the arrow indicating the reconstructed direction. For several Digital Optical Modules (DOMs), the photon counts over time are displayed alongside with the predicted photon count PDFs for a single cascade and double cascade hypothesis. Note that the DOMs labeled as "bright"111Bright DOMs have collected 10 times more light than the average DOM for an event and can bias the reconstruction at the highest measured energies. were excluded from the reconstruction. The left panel shows the event observed in 2012. Several DOMs near the interaction vertex are saturated and excluded from the reconstruction. Two further bright DOMs have been excluded as well. The photon count rates do not show a clear preference for a single vs. a double cascade hypothesis. The right panel on Figure 3 shows the event observed in 2014. Also this event has several bright DOMs which have been excluded from the reconstruction. For this event, the predicted photon count PDFs differ remarkably between the single and double cascade hypothesis, with the single cascade hypothesis disfavored according to the observed counts.

The observables are shown in Table 1. Event#1 has a large positive energy asymmetry, very close to the cut value of 0.3 where the background contribution from single cascades is significant. In the case of Event#2, the observables are in a signal-dominated region. To firmly conclude how compatible each of the double cascades is with a background hypothesis, i.e. with not being due to a $\nu_{\tau}$-CC interaction, an a posteriori analysis is ongoing.

3 Discussion

7.5 years of HESE events were (re-)analyzed with new analysis tools. The older, previously shown data set with 6 years of livetime was reprocessed using an improved detector calibration. We have performed a flavor composition measurement using a ternary topology classification directly sensitive to the tau flavor. The analysis found the first two double cascades, indicative of $\nu_{\tau}$-CC interactions, with an expectation of 1.4 $\nu_{\tau}$-induced signal events and 0.7 $\nu_{e,\mu}$-induced background events. The best fit flavor composition is $\nu_{e}:\nu_{\mu}:\nu_{\tau}=0.29:0.50:0.21$, consistent with all previously published results by IceCube, as well as with the expectation of $\sim$ 1 : 1 : 1 for astrophysical neutrinos. The first event has an energy asymmetry very close to the cut value at the boundary of the signal region. The photon arrival pattern does not show a clear preference between a single and double cascade hypothesis. The second event lies in a signal dominated region and the photon arrival pattern is well described with a double cascade hypothesis, but not with a single cascade hypothesis. An a posteriori analysis is ongoing to determine the compatibility of each of the events with a background hypothesis.