Observation of an Unusual Upward-going Cosmic-ray-like Event in the Third Flight of ANITA
P. W. Gorham, B. Rotter, P. Allison, O. Banerjee, L. Batten, J. J. Beatty, K. Bechtol, K. Belov, D. Z. Besson, W. R. Binns, V. Bugaev, P. Cao, C. C. Chen, C. H. Chen, P. Chen, J. M. Clem, A. Connolly, L. Cremonesi, B. Dailey, C. Deaconu, P. F. Dowkontt, B. D. Fox, J. W. H. Gordon, C. Hast, B. Hill, K. Hughes, J. J. Huang, R. Hupe, M. H. Israel, A. Javaid, J. Lam, K. M. Liewer, S. Y. Lin, T.C. Liu, A. Ludwig, L. Macchiarulo, S. Matsuno, C. Miki, K. Mulrey, J. Nam, C. J. Naudet, R. J. Nichol, A. Novikov, E. Oberla, M. Olmedo, R. Prechelt, S. Prohira, B. F. Rauch, J. M. Roberts, A. Romero-Wolf, J. W. Russell, D. Saltzberg, D. Seckel, H. Schoorlemmer, J. Shiao, S. Stafford, J. Stockham, M. Stockham, B. Strutt, G. S. Varner, A. G. Vieregg, S. H. Wang, S. A. Wissel
OObservation of an Unusual Upward-going Cosmic-ray-like Event in the Third Flight of ANITA
P. W. Gorham, B. Rotter, P. Allison, O. Banerjee, L. Batten, J. J. Beatty, K. Bechtol, K. Belov, D. Z. Besson,
6, 7
W. R. Binns, V. Bugaev, P. Cao, C. C. Chen, C. H. Chen, P. Chen, J. M. Clem, A. Connolly, L. Cremonesi, B. Dailey, C. Deaconu, P. F. Dowkontt, B. D. Fox, J. W. H. Gordon, C. Hast, B. Hill, K. Hughes, J. J. Huang, R. Hupe, M. H. Israel, A. Javaid, J. Lam, K. M. Liewer, S. Y. Lin, T.C. Liu, A. Ludwig, L. Macchiarulo, S. Matsuno, C. Miki, K. Mulrey, J. Nam, C. J. Naudet, R. J. Nichol, A. Novikov, E. Oberla, M. Olmedo, R. Prechelt, S. Prohira, B. F. Rauch, J. M. Roberts, A. Romero-Wolf, J. W. Russell, D. Saltzberg, D. Seckel, H. Schoorlemmer, J. Shiao, S. Stafford, J. Stockham, M. Stockham, B. Strutt, G. S. Varner, A. G. Vieregg, S. H. Wang, and S. A. Wissel Dept. of Physics and Astronomy, Univ. of Hawaii, Manoa, HI 96822. Dept. of Physics, Center for Cosmology and AstroParticle Physics, Ohio State Univ., Columbus, OH 43210. Dept. of Physics and Astronomy, University College London, London, United Kingdom. Dept. of Physics, Enrico Fermi Institute, Kavli Institute for Cosmological Physics, Univ. of Chicago , Chicago IL 60637. Jet Propulsion Laboratory, Pasadena, CA 91109. Dept. of Physics and Astronomy, Univ. of Kansas, Lawrence, KS 66045. National Research Nuclear Univ., Moscow Engineering Physics Inst., Moscow, Russia. Dept of Physics & McDonnell Center for the Space Sciences, Washington Univ in St Louis, MO Dept. of Physics, Univ. of Delaware, Newark, DE 19716. Dept. of Physics, Grad. Inst. of Astrophys.,& Leung Center for Cosmologyand Particle Astrophysics, National Taiwan University, Taipei, Taiwan. SLAC National Accelerator Laboratory, Menlo Park, CA, 94025. Dept. of Physics and Astronomy, Univ. of California, Los Angeles, Los Angeles, CA 90095. Physics Dept., California Polytechnic State Univ., San Luis Obispo, CA 93407.
We report on an upward traveling, radio-detected cosmic-ray-like impulsive event with characteristics closelymatching an extensive air shower. This event, observed in the third flight of the Antarctic Impulsive TransientAntenna (ANITA), a NASA-sponsored long-duration balloon payload, is consistent with a similar event reportedin a previous flight. These events may be produced by the atmospheric decay of an upward-propagating τ -leptonproduced by a ν τ interaction, although their relatively steep arrival angles create tension with the standard model(SM) neutrino cross section. Each of the two events have a posteriori background estimates of < ∼ − events.If these are generated by τ -lepton decay, then either the charged-current ν τ cross section is suppressed at EeVenergies, or the events arise at moments when the peak flux of a transient neutrino source was much larger thanthe typical expected cosmogenic background neutrinos. The ANITA instrument is primarily designed for the detec-tion of the ultra-high energy (UHE) cosmogenic neutrino fluxvia the Askaryan effect in ice [1–3], but is able to trigger on awide variety of different impulsive radio signals. During thefirst ANITA flight, an unanticipated radio signal was discov-ered: 16 events due to ultra-high energy cosmic ray (UHECR)air showers were found during a blind search of the datafor isolated non-anthropogenic events [4]. ANITA observesUHECR via radio impulses that occur when geomagnetically-induced charged-particle acceleration occurs in the propaga-tion of an extensive air shower in the atmosphere. Conven-tional down-going ultra-high energy cosmic-ray (UHECR) airshowers produce downward-propagating radio impulses thatare observed in reflection off the surface of the ice, leadingto phase inversion of the signal. UHECR events detectedby ANITA also include a subset of horizontally-propagatingstratospheric air showers seen just above the horizon, whichpoint directly at the payload, and show no phase inversion ofthe signal [5]. These observations have established a baselinefor identification of events of UHECR origin in ANITA data.In the ANITA-I flight one such UHECR-like event wasobserved with characteristics similar to the direct, horizon-tal cosmic rays, but from a direction well below the hori- zon, without the phase inversion due to a reflection [5]. Thebackground for this event was estimated to be ≤ − events,suggesting the possibility that such events could arise from ahigh-energy ν τ charged-current interaction in the ice, leadingto a τ -lepton which exits the ice surface and decays, producingan air shower that propagates upward in the atmosphere. How-ever, a possible anthropogenic origin for the ANITA-I eventcould not be ruled out at sufficient confidence to be conclu-sive.The third flight of the ANITA instrument took place fromDec. 18, 2014 through Jan. 8, 2015, with 22 days at float atan altitude of ∼
34 to 38 km. Unexpected strong continuous-wave (CW) interference from geosynchronous satellites lim-ited the effective full-payload exposure to about 7 days ofequivalent time. Despite this loss of sensitivity, a set of 20radio-detected UHECR events were identified in a template-based analysis [6]. Because the polarity of the events was theprimary characteristic that would distinguish phase-invertedevents from the direct events, including possible upward-going showers, we blinded the event polarity throughout theanalysis to avoid bias. The geomagnetic field in Antarctic ispredominantly vertical, and thus the Lorentz-force accelera-tion of the e + e − pairs in the shower leads to lateral charge- a r X i v : . [ a s t r o - ph . H E ] M a r separation that produces an almost completely horizontally-polarized (Hpol) signal, with nearly unique temporal andspectral properties compared to anthropogenic backgroundevents observed. Despite their small size, the residual hori-zontal components of the geomagnetic field still provide for adetailed confirmation of the geomagnetic correlation of UHE-CRs. If we write the geomagnetic field in a local cartesianbasis, then B = ( B x , B y , B z ) , with B x , B y (cid:28) B z as noted above.ANITA’s observation geometry also favors air showers withprimary particle momenta with zenith angles of 60 ◦ or more,and thus their longitudinal velocity will follow v x , v y (cid:29) v z ingeneral.From Feynman’s rule [7], the radiation field per particlewill be aligned with the observer’s apparent angular accelera-tion of the charge, which is given by the magnetic portion ofthe Lorentz force, F = q v × B . Neglecting terms that are sec-ond order in the acceleration, and recognizing that the mag-netic deflection is nearly perpendicular to the direction of radi-ation, the observed radiation field vector can be approximatedas E ∝ ( v y B z ˆ x − v x B z ˆ y ) + ( v x B y − v y B x ) ˆ z . (1)The first term in parentheses on the right hand side givesthe Hpol component of the field, and because it involvesthe strongest components of both v and B , it is the muchstronger of the two radiation fields. The second term givesthe vertically-polarized (Vpol) field component, and is signif-icantly weaker because it depends on the much weaker trans-verse magnetic field vector components. In addition, there isa small contribution from Askaryan emission, but because ofthe strong Antarctic geomagnetic field, this is limited to about4% of the total and is neglected here. Because ANITA is de-signed to do accurate pulse-phase polarimetry with both Hpoland Vpol receiving antennas, the transverse B -field compo-nent is readily detectable. Since the geomagnetic field is well-modeled in Antarctica, it provides a strong confirmation of ge-omagnetic association for a given UHECR impulse, whereassignals of anthropogenic origin are uncorrelated to the geo-magnetic field. Fig. 1 shows the geomagnetic-correlated re-sults for the UHECR events selected in ANITA-III, The ex-pected polarization is corrected for the Fresnel coefficient ofreflection where appropriate. Measurement errors were deter-mined by measurements of comparable calibration pulses, andinclude systematics.The unblinded polarity of the ANITA-III CR events showedthat the two above-horizon events among the sample had theexpected non-inverted pulse phase, consistent with their ori-gin as stratospheric, atmosphere-skimming air showers. How-ever, as noted above, one of the remaining events also hada clearly non-inverted polarity, inconsistent with a reflection,but in all other ways consistent with UHECR origin. Fig. 2shows the overlain normalized Hpol waveforms from eachof the 20 candidate events, with the 17 inverted-polarity re-flected events now un-inverted for direct comparison of thewaveform shape. The events have the instrumental response -20 -10 0 10 20 30 expected plane of polarization, deg. -20-100102030 m ea s u r ed p l ane o f po l a r i z a t i on , deg . reflected CRsgeomagneticanomalous CRdirect CRs FIG. 1:
Geomagnetic correlation of 20 UHECR events detected inANITA-III, with event planes-of-polarization determined via Stokesparameters for each event. The two above-horizon non-invertedCRs are shown in red, and the anomalous non-inverted,below-horizon CR-like event 15717147 is shown in magenta. deconvolved, and are normalized in amplitude to their maxi-mum magnitude. They are remarkably similar in shape oncethe inversion is removed.FIG. 2:
Horizontally-polarized waveforms of 20 UHECR eventsdetected in ANITA-III, with the polarity and amplitude allnormalized to the peak.
For the final 20-event UHECR selection, candidates wereverified to be spatially and temporally isolated from any otherevents like them, and showed a high degree of correlation witha waveform template determined by well-established modelsfor UHECR radio emission. We have identified no knownphysics backgrounds for these events. Potential backgroundcomes from anthropogenic radio signals that might mimic theUHECR characteristics, or unknown processes which mightlead to non-inverted polarity on reflection from the ice; fur-ther investigation of polarity is given in ref. [8]. Two inde-pendent background estimates for anthropogenic origin weremade. The first, using the likelihood that the event was a sta-tistical outlier of sub-threshold events within its nearby locale,gave a background estimate of B = . × − events for the20-UHECR sample [6]. The second method uses a probabil-ity for a single isolated UHECR-like background event, de-rived from the frequency of UHECR-like events that appearedin known anthropogenic clusters of events and charted basesor camps. Because the rate of actual UHECR events is suchthat some inevitably do get included (and therefore lost to theanalysis) as part of these clusters, this latter estimate providesonly an upper limit to the background, B ≤ .
015 events, alsofor the entire 20 UHECR sample. Thus by all indications theresulting selection of events represents a very pure sample ofradio-detected UHECRs.Fig. 3 shows the incident field strength waveforms for allthree of the events with non-inverted polarity, along with oneof the “normal” UHECR events, chosen because its arrival an-gle at the payload was similar to that of the anomalous event15717147. Detailed simulations of the UHECR radio emis-sion process find that the power spectral density (PSD) of theradio signal is dependent on the observer’s viewing angle rel-ative to the axis of the air shower, and the PSD can thus beused, along with other parameters of the shower signal, to es-timate the primary energy of the event [10]. To provide moreconfidence in our estimate, we cross-checked event 15717147against 12 of the 16 ANITA-I cosmic ray events for whichthe parameters could be directly compared and scaled. Theresults are quite consistent, yielding an estimated shower en-ergy of E = . . − . × eV for this event, assuming thatshower was initiated close to the event’s projected position onthe ice sheet. For a shower initiated at a height of 4 km abovethe ice, the energy is reduced by about 30% to E = .
40 EeV.The errors here are statistical, based on the root-mean-squareof the cross-check sample. -10 0 10 20 30 40 50 time, ns -1-0.500.51 f i e l d s t r eng t h , m V / m ° HpolVpol -10 0 10 20 30 40 50 time, ns -2-101 f i e l d s t r eng t h , m V / m ° HpolVpol -10 0 10 20 30 40 50 time, ns -2-101 f i e l d s t r eng t h , m V / m ° HpolVpol -10 0 10 20 30 40 50 time, ns -0.4-0.200.20.40.6 f i e l d s t r eng t h , m V / m ° HpolVpol atmosphere-skimming air showeratmosphere-skimming air shower normal, reflected air showeranomalousupward air shower
ANITA-III UHECR Air ShowersA BC D
FIG. 3:
The three non-inverted polarity events are shown in panelsA,B,C. Panel A shows the anomalous event, with the same polarityas the above-horizon events B and C. Panel D shows the waveformfor an inverted UHECR that had an upcoming angle close to that ofthe anomalous CR 15717147. The inversion of the normal reflectedCR event is clearly evident.
In addition to the targeted search for UHECR events, weperformed two completely independent optimized multivari-ate blind analyses of all events, favoring impulsive, highly-linearly-polarized events, without consideration of correlationto any UHECR waveform template [17]. In both of these anal-yses, complete isolation from any anthropogenic source orfrom any other events was a stringent requirement, and event15717147 passed in both cases. These two analyses confirmthat event 15717147 is unique, impulsive, and isolated, evenwhen not selected by its UHECR-related properties. The aposteriori background estimates for both 15717147 and forthe similar anomalous event seen in ANITA-I [5] are at the > ∼ σ level. There is thus significant evidence for a physicalprocess that leads to direct upward-moving cosmic-ray-likeair showers above the ice surface. Horizon m ap s i gna l − t o − no i s e r a t i o Frequency (MHz) l o g ( A S D ) ( p W M H z m ) . Upgoing Air ShowerANITAALFA
FIG. 4:
Top: Interferometric map of the arrival direction of theanomalous CR event 15717147. Bottom: ANITA combinedamplitude spectral density (ASD) for the event, from 50-800 MHz,including data from the ANITA Low Frequency Antenna (ALFA). Asimulated upward-propagating extensive air showerspectral-density curve is overlain.
For detected radio impulses, the large fields-of-view for thequad-ridged horns used in ANITA allow up to 15 antennas,drawn from up to 5 azimuthal sectors of the payload, to beused for coherent beam forming. Pulse-phase interferometrybetween these antennas then yields a map of the arrival direc-tion of the radio impulse to typical precisions of 0 . ◦ , . ◦ in elevation and azimuth, respectively [9]. Fig. 4(top) showsthe resulting false-color map for event 15717147 in coordi-nates local to the payload, scaled by the signal-to-noise ratioof the map. Elevation is with respect to the payload horizon-tal, and the azimuthal angle φ is with respect to the payloadheading at the event arrival time. Mapping is done for 360 ◦ in φ to verify that the beamforming solution is unique.ANITA-III flew a separate low-frequency horizontally-polarized quad-slot antenna, the ANITA low-frequency an-tenna (ALFA), covering the frequency band from 30 to80 MHz. ALFA’s goal was to provide radio-spectral overlap ofANITA UHECR measurements with ground-based data whichgenerally favors bands below 100 MHz. Roughly 3/4 of theUHECR event sample reported here were also detected in theALFA, and of those detections, the ALFA data for 15717147was among the events with the highest signal-to-noise ratio,in this case ≥ σ above the thermal noise. Fig. 4(bottom)shows the combined ASD for this event, including the ALFAdata. The overlain curve gives the simulated spectral densityexpected from a τ -lepton initiated air shower, with character-istics consistent with this event [15]. While similar spectraldensity would be expected for a normal CR air shower seen inreflection, these data which fit this non-inverted event furtherstrengthen its identification as an anomalous air shower.An alternative explanation of the similar ANITA-I eventas due to transition radiation of an Earth-skimming eventhas also been proposed [11]. In this model, the plane-of-polarization correlation to geomagnetic angles would be coin-cidental. Since the event observed in ANITA-III is also well-correlated to the local geomagnetic angle, and both events areconsistent within 3-5 degrees of measurement error, coinci-dental alignment for both appears probable only at the fewpercent level. The waveform of these events showed a high de-gree of correlation to radio-detected UHECRs in each flight,which supported their identification as UHECRs. Ref. [11]did not provide any detailed modeling of time-domain wave-forms for transition radiation that confirm its similarity tothose made by the UHECR emission process. This step ap-pears necessary before this hypothesis can be further evalu-ated.TABLE I: ANITA-I,-III anomalous upward air showers. event, flight 3985267, ANITA-I 15717147, ANITA-IIIdate, time 2006-12-28,00:33:20UTC 2014-12-20,08:33:22.5UTCLat., Lon. ( ) -82.6559, 17.2842 -81.39856, 129.01626Altitude 2.56 km 2.75 kmIce depth 3.53 km 3.22 kmEl., Az. − . ± . ◦ , . ± . ◦ − . ± . ◦ , . ± . ◦ RA, Dec ( ) E ( ) shower . ± . . + . − . EeV Latitude, Longitude of the estimated ground position of the event. Sky coordinates projected from event arrival angles at ANITA. For upward shower initiation at or near ice surface.
Table I gives measured and estimated parameters for both ofthe anomalous CR events, with sky coordinates derived fromthe arrival direction of the radio impulses.In our report of the ANITA-I anomalous CR event, we con-sidered the hypothesis that such events could arise throughdecay of emerging τ -leptons generated by ν τ interactions be-neath the ice surface. However, the interpretation of theseevents as τ -lepton decay-driven air showers, arising from adiffuse flux of cosmic ν τ , faces the difficult challenge that the chord lengths through the Earth are such that the StandardModel (SM) neutrino cross section [18], even including the ef-fect of ν τ regeneration [12], will attenuate the flux by a factorof 10 − [15, 16]. Event 15717147 emerged from the ice witha zenith angle of ∼ . ◦ , implying a chord distance throughthe Earth of ∼ × km water equivalent, atotal of 18 SM interaction lengths at 1 EeV. Even with com-bined effects of ν τ regeneration, and significant suppressionof the SM neutrino cross section above ∼ eV, an alterna-tive model, such as a strong transient flux from a source withcompact angular extent, is required to avoid exceeding currentbounds on diffuse, isotropic neutrino fluxes.Suppression of the cross section may occur even within theSM for the extremely low values of the Bjorken- x parameterthat obtain at ultra-high energies. For example, ref. [19] showsexamples where higher-than-expected gluon saturation at x < − causes the UHE deep-inelastic neutrino cross sectionto saturate at 10 eV, remaining essentially constant abovethat energy. This yields a factor of 3-4 suppression comparedto the SM at 10 eV, approaching an order of magnitude at10 eV. More recent studies show similar types of suppres-sion are possible, giving factors of 2-3 at 10 − eV [20, 21].Such SM-motivated scenarios would certainly decrease theexponential attenuation for the Earth-crossing neutrinos rel-evant to our case, but unless the suppression is an order ofmagnitude or more, a large transient point-source flux is likelystill required. Thus we consider also a search for potentialcandidate transients that may be associated with this event.Under the hypothesis that event 15717147 is a τ -lepton-initiated air shower, the angular error relative to the parentneutrino direction is ∼ . ◦ , arising from both the width ofthe emission cone [10], and the instrinsic statistical errors inour estimate of the arrival direction of the RF signal. To in-vestigate this hypothesis further, we point back along the ap-parent arrival direction, giving sky coordinates shown in Ta-ble I. With these parameters, we search existing catalogs forassociations with two transient source types for which sourceconfusion is not excessive: gamma-ray burst (GRB) sources,and supernovae. GRBs have been considered as possible UHEneutrino sources for many years, although there are no detec-tions to date. Supernovae (SNe) have also been proposed asUHE sources in a variety of scenarios, both in core-collapseSNe, and more recently even in type Ia SNe, which are be-lieved to originate in the ignition of a white dwarf (WD) pro-genitor. In the latter case, tidal ignition of a WD by interactionwith an intermediate-mass black hole has been proposed as apotential source of UHECRs [23–25].For the 1.5 ◦ radius error circle derived from the angularemission pattern for UHECR events, no concurrent GRBsare observed. A SN candidate is found to be associated:SN2014dz, a nearby type Ia SN at z = . . ◦ ,well within our expected angular uncertainty on the sky. Thisrelatively bright SN was discovered ∼ a posteriori probability of a chance association withany confirmed SN, at any redshift, within the estimated likelytime period of detectability for this SN, is P (cid:39) . × − , or2 . σ .If SN2014dz is the source of the putative neutrino can-didate, the implied peak isotropic neutrino luminosity mustlikely far exceed the estimated bolometric luminosity of L B = . × ergs s − . The lower limit comes already fromassuming a much lower cross section than the SM. Alterna-tively, a beaming hypothesis would significantly relax theseconstraints.Both the IceCube [13] and Auger observatories are sensi-tive to τ -leptons, IceCube through events transiting the detec-tor, or via τ − decay within the detector, and Auger via Earth-skimming τ − decay-initiated air showers within a few degreesof the horizon [14]. In this case, the declination for IceCubeimplies an additional ∼ ∼ geometric area of IceCube is still comparable to ANITA’s ef-fective point-source geometric area of ∼ at this arrivalangle. Auger has potentially a much larger effective point-source area, but only limited exposure around the time of ourevent. However if the transient flux was as large as it appears,coincident detections in archival data may be possible.A search of the projected position given by the similaranomalous event from ANITA-I in 2006 yielded no SNe orany other significant association, but the sky position for thisevent is within ∼ ◦ from the galactic plane, and thus extinc-tion leads to low SNe detection efficiency for this region ofthe sky.We thank NASA for their generous support of ANITA, andthe Columbia Scientific Balloon Facility for their excellentfield support, and the National Science Foundation for theirAntarctic operations support. This work was also supportedin part by the US Dept. of Energy, High Energy Physics Divi-sion. [1] G. A. Askaryan, Excess Negative Charge of an Electron-PhotonShower And Its Coherent Radio Emission, JETP , 441(1962); also JETP , 658 (1965).[2] D. Saltzberg, P. Gorham, D. Walz, et al. , “Observation of theAskaryan Effect: Coherent Microwave Cherenkov Emissionfrom Charge Asymmetry in High Energy Particle Cascades,”Phys. Rev. Lett., , 2802 (2001). [3] The ANITA Collaboration, P. W. Gorham, S. Barwick, J.Beatty, et al., “Observations of the Askaryan Effect in Ice,”Phys. Rev. Letters, (2007) in final revision; http://arxiv.org/abs/hep-ex/0611008 .[4] S. Hoover et al , [ANITA collaboration], Observation of Ultra-high-Energy CRs with the ANITA Balloon-Borne Radio Inter-ferometer , Phys. Rev. Lett. , (2010).[5] P. W. Gorham et al , [ANITA collaboration], Phys. Rev. Lett.117, 071101, (2016).[6] B. Rotter, 2017, unpublished PhD. dissertation, University ofHawaii at Manoa.[7] R. P. Feynman,
The Feynman Lectures , section 34-1.[8] See Supplemental Material at [URL will be inserted by pub-lisher] for details on this analysis.[9] A. Romero-Wolf, S. Hoover, A.G. Vieregg, et al , An interfero-metric analysis method for radio impulses from ultra-high en-ergy particle showers , Astropart. Phys. , 72, (2015).[10] H. Schoorlemmer, et al. [ANITA Collaboration], Astropart.Phys. 77, 32, (2016); also arXiv:1506.05396.[11] P. Motloch, J. Alvarez-Muiz, P. Privitera, and E. Zas Phys. Rev.D 95 043004 (2017).[12] F. Halzen & D. Saltzberg Phys. Rev. Lett. 81, 4305, (1998); E.V. Bugaev et al. Astropart. Phys. 21, 491, (2004); O. BlanchBigas, et al.
Phys. Rev. D 78:063002, (2008).[13] M.G. Aartsen et al. [IceCube collaboration], Phys. Rev. D 93,022001 (2016)[14] A. Aab, et al. [Pierre Auger Collaboration] Phys. Rev. D 91,092008 (2015).[15] J. Alvarez-Mu˜niz et al. , Phys. Rev. D 97, 023021 (2018); https://arxiv.org/abs/1707.00334 .[16] A. Romero-Wolf [ANITA collaboration], 35th Internat. CosmicRay Conf., Vol. 301, (2017)[17] P. Allison et al. [ANITA Collaboration], https://arxiv.org/abs/1803.02719 [18] A. Connolly, R. S. Thorne, D. Waters, Phys. Rev. D 83:113009,2011.[19] E. Henley and J. Jalilian-Marian, Phys. Rev. D 73, 094004(2006).[20] N. Armesto, C. Merino, G. Parente, E. Zas, Phys. Rev. D 77,013001, (2008);[21] A.Y. Illarionov, B. A. Kniehl, & A. V. Kotikov, Phys. Rev. Lett.106, 231802 (2011).[22] https://wis-tns.weizmann.ac.il/object/2014dzhttps://wis-tns.weizmann.ac.il/object/2014dz