A Search for MeV to TeV Neutrinos from Fast Radio Bursts with IceCube

The IceCube Neutrino Observatory instruments 1 km³ of Antarctic ice between depths of 1450 and 2450 m at the geographic South Pole and records particle interactions in the ice by capturing Cerenkov radiation, produced by secondary particles, in photomultiplier tubes that are housed in a glass pressure vessel known as digital optical modules (DOMs;Aartsen et al. 2017). IceCube observes a cosmic neutrino flux from 10 TeV to a few PeV (Aartsen et al. 2015a, 2015b) and recently reported evidence for neutrino emission from the blazar TXS 0506+056 (Aartsen et al. 2018b, 2018c).

Despite the evidence for neutrino emission from TXS 0506+056, the overwhelming majority of the diffuse astrophysical flux remains unexplained. Transient sources play a major role in high-energy astrophysics and potentially could account for a large fraction of the detected neutrino flux. Fast radio bursts (FRBs) are a class of nonperiodic, highly dispersed, millisecond-scale radio flashes (Keane 2018; Lorimer 2018). Although fewer than 60 unique sources have been detected, the expected rate of detectable FRBs each day is in the thousands (Bhandari et al. 2018). Many models for FRBs have been offered (Platts et al. 2018), but due to the scarcity of discoveries and multiwavelength follow-up detections, none is strongly favored. Some of these models allow for hadronic acceleration in the vicinity of the progenitors, such as super massive neutron stars, or supernova explosions, which would lead to the production of both high-energy cosmic rays and neutrinos (Li et al. 2014; Das Gupta & Saini 2017).

Previous analyses have set upper limits on neutrino emission from FRBs, using quality tracklike events in IceCube (Fahey et al. 2017; Aartsen et al. 2018a). A search for multiplets of tracklike neutrino events with minute-scale temporal coincidence limits the number density of a transient source class, under particular source evolution assumption, producing IceCube astrophysical flux to larger than ${10}^{-5}\,{\mathrm{Mpc}}^{-3}\,{\mathrm{yr}}^{-1}$ (Aartsen et al. 2019); the numerous and dim emission from FRBs is consistent with this lower bound (Callister et al. 2016).

This work includes two analyses that improve existing constraints on neutrino emission from FRBs and test a wider range of neutrino energies. IceCube has access to two different energy ranges, and we take advantage of that to search for coincidences with FRBs. In Section 2, we present a search using a tracklike event selection with improved effective area compared to previous IceCube searches. In Section 3, we present a search for temporal correlation of MeV neutrinos with FRBs. Section 4 summarizes analysis results and discusses the outlook for future searches for neutrinos from FRBs.

This search uses a procedure similar to that of IceCube's previous search for neutrino emission from FRBs (Aartsen et al. 2018a), which is hereafter referred to as the six years analysis. These analyses search for both temporal and spatial correlation of FRBs and muon neutrino events, in which a muon created from a charged-current interaction leaves a tracklike signature in the detector. All 28 nonrepeating FRBs are analyzed (Table 1) in a source-stacking search and in a search for the brightest source. The six years analysis used an event selection that was initially optimized for analyses of gamma-ray bursts. Because FRBs have much shorter durations, higher levels of background are tolerated, and we can use a looser event selection procedure to increase acceptance to astrophysical muon neutrinos and improve analysis sensitivity at emission timescales less than 10³ s.

2.1. Event Sample

The data used in this analysis consist of through-going muon neutrino candidate events from 2011 February 18 through 2018 March 13. In the six years analysis, to reduce the fraction of atmospheric muons in the sample, the data consisted largely of events with very high energy (E_ν > 10 TeV) or events that had penetrated many kilometers of ice prior to detection. Here, we instead use an event selection closer to IceCube's trigger level, resulting in a higher rate of atmospheric muons but increasing the acceptance of astrophysical neutrinos as well. This event selection focuses on removing low-energy events so that passing data can be transmitted via satellite within the bandwidth limit of roughly 75 gigabytes per day.

The sample of muon track events has an average all-sky rate of 35.7 Hz, due mainly to penetrating muons from cosmic-ray interactions in Earth's atmosphere. Roughly five events per day are caused by astrophysical muon neutrinos.⁵⁸ Compared to the six years analysis, the effective area of this event selection to muon neutrinos is an order of magnitude larger in the southern sky (Figure 1), with the largest improvements coming from energies less than 100 TeV. In the northern sky, where the Earth already attenuates the atmospheric muon background, the average improvement is roughly 50% in effective area. Correspondingly, the background rate for this event selection is much larger than that for the six years analysis. Figure 2 compares the distribution of background data versus the zenith angle for the six years analysis to this event selection; the peak-to-peak seasonal variation in this rate is about 25%, with a maximum rate occurring in January, the austral summer.⁵⁹

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

A total of 28 nonrepeating FRBs were analyzed in this search. Information about each FRB is presented in Table 2.

Table 2.  Analysis of 39 FRBs from 29 Unique Directions

FRB	SNDAQ	Tracks	Time (UTC)	Duration (ms)	R.A.	Decl.
FRB 110220	✓	✓	2011 Feb 20 01:55:48.957	5.6	22h 34'	−12° 24'
FRB 110523	✓	✓	2011 May 23 15:06:19.738	1.73	21h 45'	−00° 12'
FRB 110626	✓	✓	2011 Jun 26 21:33:17.474	<1.4	21h 03'	−44° 44'
FRB 110703	⋯	✓	2011 Jul 3 18:59:40.591	<4.3	23h 30'	−02° 52'
FRB 120127	✓	✓	2012 Jan 27 08:11:21.723	<1.1	23h 15'	−18° 25'
FRB 121002	✓	✓	2012 Oct 2 13:09:18.402	2.1; 3.7	18h 14'	−85° 11'
FRB 121102 b0	✓	⋯	2012 Nov 2 06:47:17.117	3.3	05h 32'	+33° 05'
FRB 130626	✓	✓	2013 Jun 26 14:56:00.06	<0.12	16h 27'	−07° 27'
FRB 130628	✓	✓	2013 Jun 28 03:58:00.02	<0.05	09h 03'	+03° 26'
FRB 130729	✓	✓	2013 Jul 29 09:01:52.64	<4	13h 41'	−05° 59'
FRB 131104	✓	✓	2013 Nov 4 18:04:01.2	<0.64	06h 44'	−51° 17'
FRB 140514	⋯	✓	2014 May 14 17:14:11.06	2.8	22h 34'	−12° 18'
FRB 150215	⋯	✓	2015 Feb 15 20:41:41.714	2.88	18h 17'	−4° 54'
FRB 150418	✓	✓	2015 Apr 18 04:29:05.370	0.8	07h 16'	−19° 00'
FRB 121102 b1	✓	⋯	2015 May 17 17:42:08.712	3.8	05h 32'	+33° 05'
FRB 121102 b2	✓	⋯	2015 May 17 17:51:40.921	3.3	05h 32'	+33° 05'
FRB 121102 b3	✓	⋯	2015 Jun 2 16:38:07.575	4.6	05h 32'	+33° 05'
FRB 121102 b4	✓	⋯	2015 Jun 2 16:47:36.484	8.7	05h 32'	+33° 05'
FRB 121102 b5	✓	⋯	2015 Jun 2 17:49:18.627	2.8	05h 32'	+33° 05'
FRB 121102 b6	✓	⋯	2015 Jun 2 17:49:41.319	6.1	05h 32'	+33° 05'
FRB 121102 b7	✓	⋯	2015 Jun 2 17:50:39.298	6.6	05h 32'	+33° 05'
FRB 121102 b8	✓	⋯	2015 Jun 2 17:53:45.528	6.0	05h 32'	+33° 05'
FRB 121102 b9	✓	⋯	2015 Jun 2 17:56:34.787	8.0	05h 32'	+33° 05'
FRB 121102 b10	✓	⋯	2015 Jun 2 17:57:32.020	3.1	05h 32'	+33° 05'
FRB 150610	⋯	✓	2015 Jun 10 05:26:59.396	2.00	10h 44'	−40° 05'
FRB 150807	⋯	✓	2015 Aug 7 17:53:55.83	0.35	22h 43'	−55° 05'
FRB 151206	⋯	✓	2015 Dec 6 06:17:52.778	3.00	19h 21'	−04° 08'
FRB 151230	⋯	✓	2015 Dec 30 16:15:46.525	4.40	09h 40'	−03° 27'
FRB 160102	⋯	✓	2016 Jan 2 08:28:39.374	3.40	22h 39'	−30° 11'
FRB 160317	⋯	✓	2016 Mar 17 09:00:36.53	21.00	07h 54'	−29° 37'
FRB 160410	⋯	✓	2016 Apr 10 08:33:39.68	4.00	08h 41'	+06° 05'
FRB 160608	⋯	✓	2016 Jun 8 03:53:01.088	9.00	07h 37'	−40° 48'
FRB 170107	⋯	✓	2017 Jan 7 20:05:45.139	2.60	11h 23'	−05° 00'
FRB 170827	⋯	✓	2017 Aug 27 16:20:18	0.40	00h 49'	−65° 33'
FRB 170922	⋯	✓	2017 Sep 22 11:22:23.40	26.00	21h 30'	−08° 00'
FRB 171209	⋯	✓	2017 Dec 9 20:34:23.50	2.50	15h 50'	−46° 10'
FRB 180301	⋯	✓	2018 Mar 1 07:34:19.76	3.00	06h 13'	+04° 34'
FRB 180309	⋯	✓	2018 Mar 9 02:49:32.99	0.58	21h 25'	−33° 59'
FRB 180311	⋯	✓	2018 Mar 11 04:11:54.80	12.00	21h 32'	−57° 44'

Note. Checkmarks (✓) indicate that an FRB is analyzed in the MeV (SNDAQ) and/or GeV–TeV (Tracks) stream. The latter included all nonrepeating FRBs. Repeated bursts from FRB 121102 are labeled with an additional "b0," "b1," and so on. Additional burst characteristics were taken from www.frbcat.org (Petroff et al. 2016): arrival time and duration corrected for signal dispersion, R.A., and decl. (J2000) are rounded to the nearest arcminute.

Download table as:  ASCIITypeset image

2.2. Analysis Method

The test statistic (TS) defined here is similar to that of the six years analysis (Aartsen et al. 2018a). For a search time window ΔT, temporal correlation of an event with an FRB is satisfied if the event triggers the detector in the interval $[{t}_{\mathrm{FRB}}-{\rm{\Delta }}T/2,\,\,{t}_{\mathrm{FRB}}+{\rm{\Delta }}T/2]$. Using a model-independent maximum likelihood method, events temporally coincident with a single FRB contribute to the maximum likelihood ratio TS, defined as

$$\begin{eqnarray}&&{\rm{TS}}=-{\hat{n}}_{s}+\sum _{i=1}^{N}\mathrm{ln}\left[1+\displaystyle \frac{{\hat{n}}_{s}S({x}_{i})}{{n}_{b}B({x}_{i})}\right],\end{eqnarray} \tag{ 1 }$$

where S(x_i) is the total spatial probability density distribution that considers the angular distance of an event direction x_i with respect to the coordinates of a given FRB and where B(x_i) combines separate spatial and temporal parameterizations of data to describe the background probability density distribution in that time and direction. Here, n_b is the expected number of background events in ΔT, and likelihood is maximized with respect to the best-fit number of observed signal events, ${\hat{n}}_{s}$. In the stacking search, the TS in Equation (1) will have an additional sum over the number of FRBs in the search.

As in the six years analysis, two tests are performed. The stacking test, which tests the hypothesis that the astrophysical class of FRBs emits neutrinos, evaluates the TS for all events in ΔT centered on all sources. The max-burst test, which tests the hypothesis that among a heterogeneous class of FRBs, one or a few bright sources emit neutrinos, evaluates a TS separately for each FRB and its respective events, returning only the largest TS as the observation at ΔT. In the stacking test, we consider a range of neutrino emission timescales by evaluating expanding iterations of ΔT from 0.03 10⁵ s: ${\rm{\Delta }}T=0.01\cdot {10}^{i/2}$ s, where i = 1, 2, ..., 14. Beyond ΔT > 10³ s, the sensitivity of the max-burst analysis exceeds upper limits set in the six years analysis for all tested spectra; therefore, we do not evaluate larger ΔT for the max-burst test.

2.3. Results

We find the results from both the stacking test and the max-burst test consistent with the background-only hypothesis (Table 3). After trials correcting each test for the number of time windows searched,⁶⁰ the p-values for the stacking and max-burst tests are 0.35 and 0.33, respectively.

Table 3.  TS Values for Stacking and Max-burst Tests in Each ΔT of the High-energy (Muon Tracks) Search

		Stacking TS			Max-source TS
ΔT (s)	Median	Result	p	Median	Result	p
3.16e–2	0	0	1	1.13	0.32	0.87
1.00e–1	0	0	1	1.21	1.10	0.54
3.16e–1	0	0	1	1.30	1.31	0.49
1.00e–0	0	0	1	1.40	0.12	0.98
3.16e–0	0	0	1	1.54	1.35	0.57
1.00e+1	0	0.29	0.124	1.76	2.38	0.32
3.16e+1	0	0.02	0.256	2.03	2.85	0.27
1.00e+2	0	0.07	0.274	2.44	4.86	0.07
3.16e+2	0	0	1	3.22	6.06	0.08
1.00e+3	0.024	2.32	0.042	5.05	10.57	0.05
3.16e+3	0.208	1.64	0.141	⋯	⋯	⋯
1.00e+4	0.779	0.79	0.492	⋯	⋯	⋯
3.16e+4	2.559	0	1	⋯	⋯	⋯
1.00e+5	8.023	0	1	⋯	⋯	⋯

Note. The median TS values from 10⁹ trials of background-only simulation are shown for comparison, along with pretrial p-values for results in their respective ΔT. For the max-burst analysis, we test only the ΔTs for which there is an improvement in sensitivity relative to the six years analysis. Both tests produce most significant results in ${\rm{\Delta }}T={10}^{3}\,{\rm{s}};$ posttrial significance for the stacking and max-source tests is p = 0.35 and p = 0.33, respectively.

Download table as:  ASCIITypeset image

Upper limits are calculated (90% confidence level) for the time-integrated flux per FRB at each ΔT (Figure 3). The results are listed in Table 4. In the stacking search (top panels), the limits we set for ΔT < 1 s are factors of 10 and 50 stronger on spectra of E⁻² and E⁻³, respectively. We compare to the southern sky results of the six years analysis because that search also excluded the repeater and tested the majority of single-burst FRBs available at the time. In the max-burst search (bottom panels), the same scale of improvement is made on the maximum flux among 28 sources at the smallest ΔT.

Table 4.  Upper Limits (90% C.L.) on the Time-integrated E⁻² Power-law Flux from 28 FRBs

FRB	IceCube U.L.	ANTARES U.L.
FRB 110220	0.0258	⋯
FRB 110523	0.0206	⋯
FRB 110626	0.112	⋯
FRB 110703	0.0204	⋯
FRB 120127	0.0353	⋯
FRB 121002	0.175	⋯
FRB 130626	0.0222	⋯
FRB 130628	0.0208	⋯
FRB 130729	0.0208	⋯
FRB 131104	0.129	1.1
FRB 140514	0.0262	1.9
FRB 150215	0.0204	2.3
FRB 150418	0.0390	1.7
FRB 150610	0.105	⋯
FRB 150807	0.134	1.6
FRB 151206	0.0212	1.3
FRB 151230	0.0203	1.6
FRB 160102	0.0755	2.0
FRB 160317	0.0689	1.6
FRB 160410	0.0213	1.5
FRB 160608	0.0995	2.1
FRB 170107	0.0213	1.1
FRB 170827	0.147	⋯
FRB 170922	0.0214	⋯
FRB 171209	0.124	⋯
FRB 180301	0.0210	⋯
FRB 180309	0.0832	⋯
FRB 180311	0.139	⋯

Note. This high-energy analysis sets upper limits on FRB neutrino emission on timescales similar to the radio emission (${\rm{\Delta }}T=30\text{ms}$), and we compare these to limits set by ANTARES with ΔT = 12 hr for 12 FRBs (Abbasi et al. 2017).

Download table as:  ASCIITypeset image

Zoom In Zoom Out Reset image size

IceCube is primarily designed to detect neutrinos with energies greater than 100 GeV, targeting sources with TeV neutrino emission. However, IceCube can measure a large burst of MeV neutrinos by detecting a collective rise in all photomultiplier rates on the top of the background noise. Although the increase in the counting rate in each light sensor is not statistically significant, the effect will be clearly seen once the rise is considered collectively over many sensors. This technique was originally developed for searches for ${ \mathcal O }$(10 MeV) neutrinos from supernovae. IceCube utilizes a real-time data stream called the Supernova Data Acquisition (SNDAQ) system to identify collective rises in the rates of photomultipliers across the detector (Abbasi et al. 2011). We use the SNDAQ data stream to search for MeV neutrinos from FRBs.

The signals from photomultiplier tubes, also called DOM hits, are counted in 2 ms bins by SNDAQ. Here, we use this data stream and search for neutrino signals from the 21 FRBs for which data were available. In this search, different bursts of the repeating FRB (FRB 121102) are considered as individual sources.

3.1. Analysis Method

To find an excess on top of the background noise rate in the detector, a one-dimensional Gaussian likelihood is used to determine the significance of a collective deviation (Δμ) of the noise across the detector.

$$\begin{eqnarray}&&{ \mathcal L }({\rm{\Delta }}\mu )=\prod _{i=1}^{{N}_{\mathrm{DOM}}}\,\displaystyle \frac{1}{\sqrt{2\pi }\,\langle {\sigma }_{i}\rangle }\,\exp \left(-\displaystyle \frac{{\left({n}_{i}-({\mu }_{i}+{\epsilon }_{i}{\rm{\Delta }}\mu )\right)}^{2}}{2\langle {\sigma }_{i}{\rangle }^{2}}\right),\end{eqnarray} \tag{ 2 }$$

where n_i is the per DOM_i rate in a chosen time bin, _i is a DOM-specific efficiency parameter that accounts for module- and depth-dependent detection probabilities, and μ_i and σ_i are the mean and standard deviation for individual DOMs. Maximizing the log-likelihood with respect to Δμ, one finds

$$\begin{eqnarray}&&{\rm{\Delta }}\mu ={\sigma }_{{\rm{\Delta }}\mu }^{2}\sum _{i=1}^{{N}_{\mathrm{DOM}}}\,\displaystyle \frac{{\epsilon }_{i}({n}_{i}-{\mu }_{i})}{\langle {\sigma }_{i}{\rangle }^{2}},\end{eqnarray} \tag{ 3 }$$

where

$$\begin{eqnarray}&&{\sigma }_{{\rm{\Delta }}\mu }^{2}={\left(\sum _{i=1}^{{N}_{\mathrm{DOM}}}\displaystyle \frac{{{\epsilon }_{i}}^{2}}{\langle {\sigma }_{i}{\rangle }^{2}}\right)}^{-1}.\end{eqnarray} \tag{ 4 }$$

If there is an excess in the rate across the detector, its significance, ξ, will be given by

$$\begin{eqnarray}&&\xi =\displaystyle \frac{{\rm{\Delta }}\mu }{{\sigma }_{{\rm{\Delta }}\mu }}.\end{eqnarray} \tag{ 5 }$$

To search for an increase in hits during the FRB period, we bin the data collected in the SNDAQ stream into bins of 10 ms. We search in eight different time windows from the lowest 10 ms and extending by powers of 2 up to a time window of 1280 ms. To estimate the background, we use data from a 10,240 ms background-only time window using the 8-hour runs before and after the actual FRB trigger. The background window excludes the signal time window, and its size does not change as we expand the signal window.

The distribution of the significance over the course of a run in IceCube is almost a Gaussian. We use the distribution of the significances, obtained from off-time windows before and after the run that includes each FRB, to obtain a threshold beyond which the significance would not arise from a random fluctuation of the background in the detector. We set a 3σ threshold (one-sided) in the significance to claim a discovery.

It has been shown that the rate of the hits in SNDAQ contains a contribution that is directly correlated with the seasonally changing rate of atmospheric muons traversing the detector. To remove this correlation, we subtract the muon dependency via linear regression as described in Baum et al. (2015). The 3σ threshold is reevaluated according to the corrected distribution. If the significance is found to be greater than the threshold set, we consider that a detection. Otherwise, we set upper limits for the absence of signal above the threshold.

3.2. Results

After obtaining the threshold for all time windows for the 21 FRBs considered in this analysis (see Table 2), we perform the likelihood analysis on the on-time window for each FRB.

No significance was found above the significance threshold in the data for the time FRBs happened. The three most significant searches are presented in Table 1. Figure 4 shows the observed significance before (left panel) and after (right panel) significance and the dependence of the significance on the atmospheric muon hit rates for the most significant FRB in the search. The distribution of the significances, along with the threshold and the observed significance for the most significant search, is shown in Figure 5. Given that no results were found beyond the threshold obtained from off-time periods, we set upper limits on the flux of antielectron neutrinos for each burst and time windows considered in this study.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The dependency of the signal hit rate on the flux of neutrinos is described in Abbasi et al. (2011). For the purpose of this analysis, in the absence of neutrino spectrum models for FRBs, we consider the neutrino emission from core collapse supernova as a fiducial model to obtain the upper limit on MeV neutrino emission. Here, the normalization was chosen such that it corresponds to a model describing the neutrino flux with average neutrino energy E_ν = 15.6 MeV and pinching parameter α = 3 (Totani et al. 1998), yielding $\langle {{E}_{\nu }}_{3}\rangle$ =7118 MeV³. To find the upper limit on the neutrino flux at MeV energies, we evaluate the required time-integrated flux of antielectron neutrinos that would produce an enhancement in the signal rate in the detector corresponding to a 90% one-sided confidence level for the Gaussian distribution of the significance in the off-time runs. The 90% one-sided confidence level for a Gaussian in ξ is given by

$$\begin{eqnarray}&&{\xi }^{90}={\xi }_{\mathrm{th}}+{z}_{90}\cdot {\sigma }_{\xi },\end{eqnarray} \tag{ 6 }$$

where ${\xi }_{\mathrm{th}}$ is the significance threshold, σ_ξ is the width of the significance distribution, and z₉₀ ≈ 1.282 is the fraction of the Gaussian width (1σ) representing the 90% confidence level interval. The significance for which one finds ξ < ξ_cut only in 10% of random draws is denoted by ξ⁹⁰. The corresponding additional hits per DOM will be

$$\begin{eqnarray}&&{\rm{\Delta }}{\mu }^{90}={\sigma }_{{\rm{\Delta }}\mu }\left({\xi }_{\mathrm{th}}+{z}_{90}\cdot {\sigma }_{\xi }\right).\end{eqnarray} \tag{ 7 }$$

The total increase in the signal hit rates in the detector is obtained by adding the signal of all DOMs. The upper limit on the time-integrated flux for FRBs considered in this analysis is shown in Figure 6.

Zoom In Zoom Out Reset image size

In two searches for neutrino emission from FRBs—one for tracklike events from muon neutrinos above 100 GeV and the other for MeV neutrino events—no significant association has been found.

We set upper limits (90% confidence level) on the time-integrated neutrino flux from FRBs that are the most constraining to date. For a power-law spectrum E⁻² (E⁻³), the limit set for GeV–TeV neutrinos is ${E}^{2}F\lt 2\,\times {10}^{-3}\,\mathrm{GeV}\,{\mathrm{cm}}^{-2}$ (${E}^{2}F\lt 2\times {10}^{-4}\,\mathrm{GeV}\,{\mathrm{cm}}^{-2}$ at 100 TeV) per burst for emission timescales less than 10 s. We also set the first upper limits on MeV neutrino emission from FRBs: $\langle {E}^{3}\rangle F\lt 8\times {10}^{11}\,{\mathrm{MeV}}^{3}\,{\mathrm{cm}}^{-2}$ at an emission timescale of 10 ms (Figure 6).

In the tracks search, IceCube is more sensitive in the northern sky than in the southern sky, where most FRB sources have been detected so far. In addition, the stacking sensitivity scales with roughly the number of sources for which detection of a spatially coincident background event is unlikely within the time window. Therefore, as more FRBs are detected, especially in the northern sky by observatories like the Canadian Hydrogen Intensity Mapping Experiment (CHIME; Amiri et al. 2018, 2019), the stacking analysis sensitivity can improve by orders of magnitude. With fields of view significantly larger than those of single-dish telescopes, new radio interferometers may detect several FRBs per day in the coming years (Newburgh et al. 2016; Amiri et al. 2018), accelerating this improvement.

The MeV neutrino search was made possible by the ability of IceCube to identify a burst of MeV neutrinos on top of the background noise in the detector. While this capability has been primarily incorporated for obtaining early alerts on supernova explosions, it offers a unique opportunity for temporal study of low-energy neutrino -emission from transients. With this opportunity, we have placed the first-ever limits on neutrino signals at MeV energies from FRBs. Prospects for observation of an excess of MeV neutrinos in IceCube depends on the distance to the source. While IceCube is highly sensitive for identification of MeV signals from Galactic distances, there is growing evidence that most FRBs are extragalactic now that the redshift has been measured for two sources (Bannister et al. 2019; Ravi et al. 2019). It is worth mentioning that, in addition, on request, all untriggered DOM hits in the detector may be stored for any particular time period for several days (Aartsen et al. 2017). IceCube has used this channel to search for neutrino emission from gravitational wave sources. For more details, see Abbasi et al. (2017).

The IceCube collaboration acknowledges the significant contributions to this manuscript from Sam Fahey, Ali Kheirandish, and Justin Vandenbroucke. The authors gratefully acknowledge the support from the following agencies and institutions: USA—U.S. National Science Foundation Office of Polar Programs, U.S. National Science Foundation Physics Division, Wisconsin Alumni Research Foundation, Center for High Throughput Computing (CHTC) at the University of Wisconsin–Madison, Open Science Grid (OSG), Extreme Science and Engineering Discovery Environment (XSEDE), U.S. Department of Energy National Energy Research Scientific Computing Center, particle astrophysics research computing center at the University of Maryland, Institute for Cyber-Enabled Research at Michigan State University, and astroparticle physics computational facility at Marquette University; Belgium—Funds for Scientific Research (FRS-FNRS and FWO), FWO Odysseus and Big Science programmes, and Belgian Federal Science Policy Office (Belspo); Germany—Bundesministerium für Bildung und Forschung (BMBF), Deutsche Forschungsgemeinschaft (DFG), Helmholtz Alliance for Astroparticle Physics (HAP), Initiative and Networking Fund of the Helmholtz Association, Deutsches Elektronen Synchrotron (DESY), and High Performance Computing cluster of the RWTH Aachen; Sweden—Swedish Research Council, Swedish Polar Research Secretariat, Swedish National Infrastructure for Computing (SNIC), and Knut and Alice Wallenberg Foundation; Australia—Australian Research Council; Canada—Natural Sciences and Engineering Research Council of Canada, Calcul Québec, Compute Ontario, Canada Foundation for Innovation, WestGrid, and Compute Canada; Denmark—Villum Fonden, Danish National Research Foundation (DNRF), Carlsberg Foundation; New Zealand—Marsden Fund; Japan—Japan Society for Promotion of Science (JSPS) and Institute for Global Prominent Research (IGPR) of Chiba University; Korea—National Research Foundation of Korea (NRF); Switzerland—Swiss National Science Foundation (SNSF); and United Kingdom—Department of Physics, University of Oxford.