Antarctic Surface Reflectivity Measurements from the ANITA-3 and HiCal-1 Experiments P. W. Gorham,1 P. Allison,2,3 O. Banerjee,2 J. J. Beatty,2,3 K. Belov,4 D. Z. Besson,5 W. R. Binns,6 V. Bugaev,6 P. Cao,7 C. Chen,8 P. Chen,8 J. M. Clem,7 A. Connolly,2,3 B. Dailey,2 P. Dasgupta,9 C. Deaconu,10 L. Cremonesi,11 P. F. Dowkontt,12 B. D. Fox,1 J. Gordon,2 B. Hill,1 R. Hupe,2 M. H. Israel,6 P. Jain,9 J. Kowalski,1 J. Lam,12 J. G. Learned,1 K. M. Liewer,4 T.C. Liu,8 S. Matsuno,1 C. Miki,1 M. Mottram,11 K. Mulrey,7 J. Nam,8 R. J. Nichol,11 A. Novikov,5,13 E. Oberla,10 S. Prohira,5 B. F. Rauch,6 A. Romero-Wolf,4 B. Rotter,1 K. Ratzlaff,5 J. Russell,1 D. Saltzberg,12 D. Seckel,7 H. Schoorlemmer,1 S. Stafford,2 J. Stockham,5 M. Stockham,5 B. Strutt,11 K. Tatem,1 G. S. Varner,1 A. G. Vieregg,10 S. A. Wissel,14 F. Wu,12 and R. Young5 1Dept. of Physics and Astronomy, Univ. of Hawaii, Manoa, HI 96822. 2Dept. of Physics, Ohio State Univ., Columbus, OH 43210. 3Center for Cosmology and Particle Astrophysics, Ohio State Univ., Columbus, OH 43210. 4Jet Propulsion Laboratory, Pasadena, CA 91109. 5Dept. of Physics and Astronomy, Univ. of Kansas, Lawrence, KS 66045. 6Dept. of Physics, Washington Univ. in St. Louis, MO 63130. 7Dept. of Physics, Univ. of Delaware, Newark, DE 19716. 8Dept. of Physics, Grad. Inst. of Astrophys.,& Leung Center for Cosmology and Particle Astrophysic s, National Taiwan University, Taipei, Taiwan. 9Dept. of Physics, Indian Institute of Technology, Kanpur, Uttar Pradesh 208016, India 10Dept. of Physics, Enrico Fermi Institute, Kavli Institute for Cosmological Physics, Univ. of C hicago , Chicago IL 60637. 11Dept. of Physics and Astronomy, University College London, London, United Kingdom. 12Dept. of Physics and Astronomy, Univ. of California, Los Angeles, Los Angeles, CA 90095. 13National Research Nuclear University, Moscow Engineering Physics Institute, 31 Kashirskoye Highway, Rossia 115409 14Dept. of Physics, California Polytechnic State Univ., San Luis Obispo, CA 93407. (Dated: March 22, 2017) The primary science goal of the NASA-sponsored ANITA project is measurement of ultra-high energy neutrinos and cosmic rays, observed via radio-frequency signals resulting from a neutrino- or cosmic ray- interaction with terrestrial matter (atmospheric or ice molecules, e.g.). Accurate inference of the energies of these cosmic rays requires understanding the transmission/reflection of radio wave signals across the ice-air boundary. Satellite-based measurements of Antarctic surface reflectivity, using a co-located transmitter and receiver, have been performed more-or-less continu- ously for the last few decades. Our comparison of four different reflectivity surveys, at frequencies ranging from 2–45 GHz and at near-normal incidence, yield generally consistent maps of high vs. lowreflectivity,asafunctionoflocation,acrossAntarctica. UsingtheSunasanRFsource,andthe ANITA-3 balloon borne radio-frequency antenna array as the RF receiver, we have also measured the surface reflectivity over the interval 200-1000 MHz, at elevation angles of 12-30 degrees. Con- sistent with our previous measurement using ANITA-2, we find good agreement, within systematic errors (dominated by antenna beam width uncertainties) and across Antarctica, with the expected reflectivity as prescribed by the Fresnel equations. To probe low incidence angles, inaccessible to theAntarcticSolartechniqueandnotprobedbyprevioussatellitesurveys,anovelexperimentalap- proach(“HiCal-1”)wasdevised. Unlikepreviousmeasurements,HiCal-ANITAconstituteabi-static transmitter-receiverpairseparatedbyhundredsofkilometers. DatatakenwithHiCal,between200– 600 MHz shows a significant departure from the Fresnel equations, constant with frequency over that band, with the deficit increasing with obliquity of incidence, which we attribute to the com- bined effects of possible surface roughness, surface grain effects, radar clutter and/or shadowing of the reflection zone due to Earth curvature effects. We discuss the science implications of the HiCal results, as well as improvements implemented for HiCal-2, launched in December, 2016. I. INTRODUCTION Within the last 30 years, the sub-field of ultra-high energy cosmic ray (UHECR; E > 1018 eV) astronomy has emergedasavibrantexperimentalandtheoreticalsub-fieldwithinthelargerfieldofparticleastrophysics,comprising studies of both charged and neutral particles at macroscopic kinetic energies. The physics interest in UHECR lies in understanding i) the nature of the cosmic accelerators capable of producing such enormously energetic particles at energies millions to billions of times higher than we are capable of producing in our terrestrial accelerators, ii) the detailsoftheinteractionofUHECRwiththecosmicraybackground,evidentintheobservedenergyspectrumofcosmic rays as an upper ‘cut-off’[1–3], or maximum observed energy, at approximately 1020 eV, and iii) correlations in the arrival directions of UHECR with exotic objects such as neutron stars, gamma-ray bursts (GRB), and active galactic 2 nuclei (AGN). Experimentally, charged-particle UHECR astronomy is currently dominated by two experiments – the SouthernHemispherePierreAugerObservatory(PAO)[4]basedinMalargue,ArgentinaandtheNorthernHemisphere TelescopeArray(TA)[5]basedinUtah,USA.Theconstructionoftheseobservatoriesisverysimilar,basedonalarge number of ground detectors sampling the charged component of Extensive Air Showers (EAS) and with stations deployed over hundreds of square kilometers at ∼1 km spacing, coupled with a much smaller number of atmospheric nitrogen fluorescence detectors at sparser spacing having a much more restricted duty cycle, but individually capable of providing a more comprehensive image of atmospheric shower development. Within the last decade, the PAO has been complemented by an array of radio-wave antennas, capable of measuring the signal generated primarily by the separation of the charged particles comprising the down-moving air shower in the geomagnetic field[6, 7]. This technique has also demonstrated sensitivity to shower development and, therefore, the composition of the primary UHECR[8]. II. RADIOFREQUENCY UHECR DETECTION WITH THE ANITA EXPERIMENT Although originally purposed for detection of neutrinos, the ANITA-1 mission (2006) unexpectedly observed 14 extremelyhigh-energyRFsignalswithanon-neutrino-likeradiowavesignalpolarization(horizontal[HPol]vs. vertical [VPol],asexpectedforneutrinos)whichtracedbacktotheAntarcticsurfacebeneaththeballoon[9]. Afterconsiderable work,itwasdemonstratedthattheseeventsweretheresultofcollisionsofdown-comingprotonsatultra-highenergies (corresponding to energies 10,000,000 times greater than the energy of particles accelerated in the Large Hadron Collider in Geneva, Switzerland) with atmospheric molecules. The RF signals produced in the collision, as the combined result of the so-called “Askaryan effect” resulting from the net charge excess acquired by the shower as it descends through the atmosphere, plus the “geomagnetic” signal resulting from separation of different charged species due to the Earth’s magnetic field, were subsequently reflected off the Antarctic surface and back up to the ANITA balloon. Perhaps most striking was the discovery of two additional, steeply inclined events in the ANITA-1 sample entering the atmosphere nearly parallel to the Earth; these RF signals were observed directly rather than via surface reflection and had a signal polarity exactly opposite those observed via surface reflection, as expected. Within the last year, expanded analysis of the ANITA-1 event sample has uncovered one upcoming event with signal polarity inconsistent with a reflection[10]. Such an event does not readily find a conventional explanation, and has been suggested as a possible ν candidate. τ RecentanalysisoftheenergyscaleofthoseUHECRevents,basedondetailedmodelingofthesignalgeneration[11], as well as fraction of primary cosmic ray-induced extensive air showers reflected off the surface and back up the ANITA experiment indicate that the efficiency-per-livetime for UHECR registration approaches that of the PAO and TA observatories. Accurate flux measurements require accurate inference of UHECR energies from their detected radio-wave signals. For ANITA’s detection of protons, this means understanding surface effects on the reflected radio signalgeneratedinanair-protoncollision. Similarly,measurementoftheenergiesofneutrinoscollidingin-icerequires understanding of the surface-transmitted signal observed in-air by ANITA. III. EXPERIMENTAL MEASUREMENT OF ANTARCTIC SURFACE ROUGHNESS AND REFLECTIVITY Several (interdependent) physical phenomena are likely responsible for the limited reflectivity of the Antarctic surface. Wind-blownsurfaceinhomogeneitiesarelikelytheprimarydeterminant;windmayalsodeterminethetypical locale-specific surface grain size. Sub-surface scattering effects will also play some role, depending on frequency; to the extent that annual thin ‘crusts’ of icy surface layers form as the temperature warms and then cools, volume and also layer scattering will both contribute. Disentangling and quantifying all these features is an ongoing geophysical exercise. A. Satellite-based High Frequency Measurements Atfrequenciesbeyond1GHz,AntarcticsurfacereflectivitydatahavebeentakenbytheEnvisat[12]andAquarius[13] satellites. Figure 1 shows compilations of these data across the continent. Visually, the four bands show considerable similarity; interpretation of the L-band satellite data is somewhat complicated by the fact that those data have been taken in a variety of polarizations. We note that the smallest wavelength probed in these satellite surveys is <1 cm, suggesting that the surface is uniformly incoherent at wavelengths up to 30 cm. 3 (a)Ka-band(26.5–40GHz)Antarcticsurfacereflectivity,drawn (b)Ku-band(12–18GHz)Antarcticsurfacereflectivity,drawn fromEnvisatsatellitedata. fromEnvisatsatellitedata. (c)S-band(2–4GHz)Antarcticsurfacereflectivity,drawnfrom (d)CorrelationmatrixbetweenS-band,Ku-band,Ka-band Envisatsatellitedata. reflecivities,aswellascorrelationwithwindvelocities. FIG. 1: Compilation of satellite reflectivity data, over the frequency range 2–40 GHz, and correlation matrix. Wehaveconsideredtheinternalconsistencyofthehigher-frequencysatellitesurveyswitheachother, aswellasthe correlationbetweenreflectivityandwindspeed. Figure1d)summarizesthecorrelationbetweenthethreehigherbands and also relative to wind velocity. We observe positive correlations in reflectivity across the continent for those three higher bands, indicating consistency in albedo measurements, as a function of position. We additionally, as expected, observe an anti-correlation between wind velocity and reflectivity, consistent with the expectation that higher wind velocities results in higher roughness and reduced albedo. We conclude that, above 2 GHz, the satellite-based surveys are generally consistent and also consistent with wind-driven surface effects reducing overall surface reflectivity. However, all these measurements probe only surface scattering at normal incidence. B. Solar Measurements with ANITA-2 and ANITA-3 We can probe surface roughness with radio wave receivers by measuring the ratio of the intensity of the surface- reflected radio-frequency Solar image to the Solar image observed directly by the balloon-borne ANITA experiment. That measurement can then be compared to the ratio expected for reflection off of a smooth surface (“specular” reflection). By taking the ratio of the surface-reflected Solar RF power to the direct Solar RF power measured with 4 ANITA, as a function of incident elevation angle relative to the surface, θ , we can thus estimate the surface power i reflection coefficients R(θ ). For θ > 15◦, our previous analysis found general consistency with the values of R(θ ) i i i expectedfromtheFresnelequations[14],whichprescribetheamountofsignalpowerreflectedattheinterfacebetween two smooth dielectrics given their indices of refraction, for both vertical- vs. horizontal-polarizations (“VPol” and “HPol”, respectively). At more glancing incident angles (θ < 15◦), the ANITA-2 data suggested slightly reduced i signal strength compared to the expectation from the Fresnel equations, perhaps indicating that surface roughness effects are becoming increasingly apparent at oblique incidence angles. Figure 2 shows one sample ANITA-3 HPol interferogram used to compile the reflection coefficients, as a function of incidence angle. In general, the ANITA-3 data follow the trend obtained from the ANITA-2 data. FIG. 2: Sample ANITA-3 sun-centered interferogram showing solar radio frequency image (at (φ,θ ∼0,0)) and reflection (at (φ,θ)∼(0,−32)). C. The HiCal Experiment TheHiCal(High-altitudeCalibration)balloon-bornetransmitterwasproposedtoemulatetheradiosignalsproduced byUHECRandtoderivetheeffectsofsurfacereflectivityandroughness, usingANITAasthereceiver. Totheextent that the generated HiCal signal matches the waveform expected from UHECR, HiCal triggers registered by ANITA also can be used as a signal ‘template’ in offline cosmic-ray search analysis. In January, 2015, this technique was successfully prototyped with the HiCal-1 flight in Antarctica, tracking ANITA-3. In this scheme, a ‘trailer’ balloon (“HiCal-1”), comprising an in-air RF transmitter emitting high amplitude signals measured by ANITA both directly (“D”),aswellasintheirsurfacereflection(“R”),islaunchedinproximitytotheANITAflightpath. Theratioofthe measured ANITA amplitude from a surface-reflected HiCal signal relative to a directly-received signal, over a wide range of incidence angles numerically defines the reflectivity; the short ∼10µs separation time of these two signals gives a unique, and easily recognizable signature in the ANITA data sample. Knowing the GPS coordinates of both ANITA as well as HiCal at any given time, we can calculate the expected time difference between the reflected and direct RF signals. At these large separations, inclusion of Earth curvature effects, as well as the ∼2-3 km elevation of the Antarctic plateau at the putative intervening RF reflection point, are critical. Note that the HiCal-ANITA transmitter-receiver pair represent a bi-static radar configuration, for which the de- pendence of the received signal on the radar beam direction relative to the sastrugi alignment can be opposite that expectedformonostaticradar. Inthecaseofsastrugialignedtransversetotheradarbeam,monostaticdevicessuchas satellites directly measure the enhanced radar backscatter, while in the bi-static configuration, signal can be affected by such things as, e.g., local shadowing. 5 D. HiCal hardware 1. Payload Schematic and Transmitter TheHiCal-1transmitterisbasedonasmallceramic piezo-electric. Such devices translate the mechanical energy of impact of a solid ‘actuator’ with a piezo ce- Iridium unit Physical Layout MIP PCB ramic into a ∼10 nanosecond-duration burst of elec- Relay and regulator 9” MIP enclosure To trical energy, and are capable of generating kiloVolt- HICAL PCB balloon scale radio-frequency signals. The full HiCal pay- MIP battery (9V) loadconsistsofthreesub-elements,schematicallyillus- HICAL motor battery Power cable 6” trated in Figure 3. The “Micro-Instrumentation Pack- Timing cable age” (MIP) is a NASA standard for sub-orbital mis- Motor/cam/piezo sions, and contains the hardware for communications Dipole with the payload and control operations during flight Rigging lines Pressure transducer (telemetry), as well as GPS payload time and location 24” information,whichrunsasynchronouslyrelativetothe HiCal triggers. Below the MIP, the “actuator” com- 20” 3” prises a motor turning at a rep rate of approximately 0.33 Hz which drives a camshaft, designed to depress the spring-loaded piezo electric at the same 0.33 Hz 20” HICAL PV enclosure frequency. Signals from the piezo are directed into the dipole antenna (built according to the RICE experi- FIG. 3: HiCal-1 payload schematic, showing electronics box, ment’s antenna specifications[15]), for which the feed comprisingmicro-instrumentationpackage(MIP)compartment, point has been coated with anti-coronal paint to sup- including GPS and MIP battery, and HiCal compartment, plus press possible arcing of the kiloVolt signals emitted by RICE dipole transmitter below. MIP is supplied by Columbia the piezo. Ultimately, given the enhanced arcing with Scientific Balloon Facility (CSBF), funded by NASA. which one must contend at the 38-km HiCal float al- titude (5 mB pressure), a dedicated pressure vessel, constructed from lightweight ABS (Acrylonitrile-Butadiene- Styrene), was built to enclose the dipole and piezo in a sealed, 1000 mB environment. A second GPS board time stamps the RF signals being emitted by the dipole. 2. Azimuthal Orientation Readout Sincetheemittedsignalstronglydependsontheorientationofthedipoletransmitterrelativetothedirectionofthe ANITA balloon and since the payload can freely rotate during flight, an additional custom printed-circuit board also provides azimuthal orientation information (“HiCaz”). This board consists of 8 photodiodes spaced evenly around the circumference of the PC board, allowing determination of the azimuthal orientation relative to the instantaneous positionofthesuninthesky,byinterpolationofthemeasuredphotodiodevoltages. Afteraslopecorrection,weobtain an azimuthal orientation resolution of approximately 1–2 degrees, which is smaller than the intrinsic uncertainty in the transmitter antenna directivity. 3. Signal Generation Details; Spark Gap Dependence E. HiCal-1 flight details Four science projects were approved for launch for the 2014-15 NASA Long-Duration Balloon campaign in Antarc- tica. These were the ANITA-3 project, the Super Pressure Balloon COSI[16] project, the SPIDER[17] project, and HiCal. The first HiCal launch attempt (“HiCal-1a”), on December 19, 2014 was unsuccessful. Insufficient lift, which resultedinaballoonwhichfailedtoascend,wascompoundedbyerraticpiezooutputpulses. Asecondlaunch(Fig. 4), using back-up HiCal hardware occurred on January 6, 2015 (“HiCal-1b”) during ANITA-3’s return after one circuit around the continent. Almost immediately after turning on the HiCal transmitter during ascent, ANITA began reg- istering D triggers, the ∼700 km separation distance notwithstanding. We note that no clear R signals were observed during that time, indicating very small reflection coefficients at the near-glancing angles typical of ascent. After HiCal reached its 38-km float altitude, during those times when the transmitter was activated by the HiCal motor, 6 FIG. 4: Zoom of HiCal payload directly following launch. FIG. 5: ANITA-3 flight track (blue for first orbit and red for second) overlaid with HiCal-1b (green). signals continued to be recorded for the subsequent 48 hours (after which time the ANITA flight was terminated), at ANITA-HiCal separations between 650 and 800 km. The flight tracks of ANITA-3 (blue for first orbit and red for second orbit) is overlaid with that of HiCal-1b (shown in green) in Figure 5 (reproduced from the CSBF/NASA website at http://www.csbf.nasa.gov/antarctica/ice.htm). In the initially telemetered ANITA data sample, three ‘doublet’ events were quickly identified for which the time separation between the first pulse (the direct HiCal trigger D) and the second (the surface-reflected signal R) was of order 7.2 microseconds. Figure 6 shows that these events agree excellently with calculations of the expected direct- reflected signal time delays based on the known HiCal-ANITA separation (640 km) at the time these events were recorded. In total, HiCal-1b produced ∼600 triggers observed by ANITA, many at ANITA/HiCal separation distances of ∼750 km, or 200 km further than the ANITA ground pulsers can be seen, due to Earth curvature effects. F. HiCal science results 1. Time Characeristics of R and D signals The known separation distance between HiCal and ANITA, combined with elevation information of the Antarctic plateau, can be used to infer an expected time delay between registration of the reflected vs. the direct signals, as shown in Fig. 7. At typical separation distances of order 600-800 km, we expect time differences between R and D triggers to be of order 7 microseconds. The camshaft actuator rotates with a period of approximately 3 seconds; Fig. 8 illustrates the time between successive HiCal direct triggers registered offline by ANITA during ascent (left) and after reaching float (center). We notetheclearpresenceofa3secondperiodicity,althoughthefactthatthetimeintervalbetweensuccessivetriggersis often6, 9, 12... secondsobviouslyindicatesthatseveraltriggersare’missing’from theANITA-3datastream. Thisis dueeithertoloweramplitudetransmittersignalsor, giventhefactthatthepayloadisrotating, anunfavorableHiCal transmitter azimuthal orientation relative to the ANITA payload. HiCal-2 includes improved azimuthal information to resolve this uncertainty. Figure 8 (right) shows the time difference between successive ANITA-3 triggers, taken while HiCal-1b was pulsing, and showing the expected time difference between R and D events. Figure9showstheangularreconstructionofthedirectandreflectedsignals. Asexpected,thedirectsignalsemanate from a point slightly below horizontal, but above a tangent to the Earth, whereas R signals appear to emanate from a point on the ice. 7 FIG.6: Measuredtimedifferencebetweensuccessivesignalsreg- FIG. 7: Calculated incidence angle of Reflection signals rela- istered in telemetered sub-sample of ANITA-3 data (points) vs. tive to horizon (x-axis) and time difference between Direct and expectation, given known GPS locations of transmitter and re- Reflected pulses (color scale, in microseconds) as a function of ceiver (curves). Bottom curve, from HiCal MC simulations, HiCal and ANITA separation distance (km; along y-axis). corresponds to known 3 km Antarctic plateau elevation and in- cludes Earth curvature effects. FIG.8: Left: DifferencebetweensuccessiveDtriggers,registeredforANITAruns401and402,duringHiCal-1bascent. Center: Same, registered for ANITA run 413, for which the ANITA-HiCal separation distance was of order 680 km, and for which the HiCal motor was on (and therefore HiCal-1 presumably pulsing). We note the evident ∼3 second period of the transmitter pulsing; we also observe a large fraction of times when HiCal was nominally pulsing but no triggers registered at ANITA-3. We attribute at least some of these to cases where the HiCal transmitter antenna was in a geometrically disfavorable orientation relative to ANITA-3. Right: Time difference between successive registered HiCal-direct signals during one pulsing period. 2. Reflectivity Measurements ThegoalofHiCalistoobtainanestimateofthesurfacereflectivityforsurfaceincidenceangleslessthan5degrees. One of the signature features of signals reflecting off higher index-of-refraction materials relative to direct signals is the expected signal inversion. Figure 10 shows that the inverted R signal gives, modulo an overall scale factor, a very good match to the directly-measured signal. In fact, in the entire sample of doublet events consisting of an observed reflectionaswellasanobserveddirectsignal,thecross-correlationforthereflectedsignalexceedsthecross-correlation of the direct event 100% of the time. The match of the signal shape also indicates that, to a good approximation, the frequency composition of the reflected signal matches that of the direct signal. Figure 11 shows the coherently summed,andaveragedwaveformsfordirectvs. reflectedsignals,againverifyingtheconsistencyofsignalshapeacross the two samples. We also note from Figure 11 that the direct VPol signal is reduced by approximately a factor of 10 in voltage (100 in power) relative to the HPol signal, indicating that the combined effects of cross-polarization broadcastfromthetransmitterdipole,pluscross-polarizationresponseoftheANITAhornantennas,plusanypossible vertically transmitted component resulting from the transmitter possibly being non-horizontal and having a non-zero polar angle offset, is reduced by approximately 20 dB relative to the expected HPol broadcast and received power. In these measurements, and more critically for the Solar measurements, one must correct for the different elevation 8 FIG. 9: Angular reconstruction of HiCal direct and reflected events, in elevation (θ) and azimuth (φ). In these coordinates, θ=0 corresponds to Horizontal relative to ANITA; θ ∼−6o corresponds to the Earth horizon. Azimuth has subtracted out the known azimuthal location of the HiCal-1 transmitter at a given time. FIG. 11: Coherently summed, averaged waveforms for Di- FIG. 10: Overlay of HiCal-1 Direct event 79998774 (red) rectHPol(upperleft), DirectVPol(upperright), Reflected overlaid with inverted event 79998775 (green) and non- HPol (lower left) and Reflected VPol (lower right). Hori- inverted event 79998775 (blue), illustrating expected signal zontal offset from zero, for all plots, is arbitrary. inversion for HiCal-1 reflected pulse. angles of the direct vs. reflected signals. Since the ANITA antennas are canted at a downwards angle of 10 degrees, reflected signals are closer to boresight than direct signals. Consequently, the measured signal power for R vs. D mustbemultipliedbyafactorexp(−(θ +θ )2/2σ2(f))/exp(−(θ +θ )2/2σ2(f)), withθ theelevationangle R Cant θ D Cant θ R to the reflection point, θ the elevation angle to the direct source, and σ the beam width (3 dB full-width-half- D θ max/2.36) at a given frequency, to obtain the actual reflected power ratio. Figure 12 presents the raw ratio of R:D using two different estimators: in the first, the R and D signal power estimated from the interferometric map are sideband subtracted, to remove any DC offsets. In the second, the peak of the interferometric maps, for the pixels correspondingtotheDandRsignalsaredirectlycompared. Thetwoestimatesprovidegenerallyconsistentmeasures of the reflectivity. Thevalueofthereflectioncoefficientcanthusbeinferredeitherdirectlyfromthescalefactorneededto‘boost’the R-waveform to match the D-waveform, in voltage, or, alternately, by a direct measurement of the received D vs. R power in an interferogram, which measures signal strength in units of voltage2. Compiling those two and combining with the Solar measurements, we present our measurements, along with comparison to calculation (detailed below) 9 Received R:D power from two estimates 1 3 o i t a R 0.9 r e 2.5 w 0.8 o P k 0.7 a 2 e P 0.6 0.5 1.5 0.4 1 0.3 0.2 0.5 0.1 0 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Background-subtracted ratio FIG. 12: Ratio of R:D signal strength using two different estimators. and also the raw Fresnel coefficients, as shown in Figure 13. Although the Solar measurements give good consistency with ‘raw’ Fresnel, the HiCal measurements show a clear deficiency of signal relative to the expectation from the Fresnel equations at oblique incidence angles <5◦. IV. ESTIMATES OF EXPECTED REFLECTIVITY Given the indices of refraction of two dielectric media, the Fresnel coefficients for reflection and transmission of power are derivable by imposing continuity at a smooth interface, taking into account the difference in wavespeeds in the two media. These equations numerically prescribe the fraction of power reflected by, and transmitted across that interface, assuming that all the scattering occurs at that interface. In practice, of course, there is some penetration across the dielectric boundary and into the dielectric, such that the actual reflected signal includes some sub-surface reflected signal. A. Reflection considerations In the most simple-minded approach, and neglecting roughness, we can approximate the Earth as a convex mirror, with focal length equal to half the Earth radius. At normal incidence, in the case where the object distance d is s quasi-infinite (the sun), the image distance is essentially the focal length, and the reflectance dictated by the Fresnel coefficients. Iftheobjectdistanceismuchsmallerthanthefocallength(thecasefor,e.g.,eitherHiCalsignalsorradio- frequency signals from UHECR), now the image distance is approximately equal to the object distance. If parallel rays traveling toward a convex mirror are not parallel to the main, there is still a ’virtual’ focal point f, although now f (cid:54)=R /2. In this case, the deviation from normal incidence is set by the scale of the ratio of one Fresnel zone, E 10 Reflection Coefficient Anita3 Sun >200 MHz n 0.7 o Anita3 Sun >400 MHz ti c a Fresnel HPol r F r 0.6 Plane Wave Calculation e w o arXiv:1506.05396 UHECR P d 0.5 HiCal data e t c e fl e R 0.4 0.3 0.2 0.1 5 10 15 20 25 30 Elevation Angle q with respect to surface FIG. 13: Summary of ANITA-3 Antarctic radio frequency surface reflectivity measurements. Curve is Fresnel power reflection coefficient,assumingsurfaceindex-of-refractionof1.35. FilledblackcirclesareANITA-3Solarobservationsoverfullband200- 1000 MHz with blue bands and yellow bands indicating estimated uncertainties due to antenna beam width uncertainties and interferometricfringingeffects,respectively;openbrownsquaresshowSolarreflectivityresult,afterhigh-passfilteringabove400 MHz. Inverted red-brown triangles show results obtained using HiCal-1b triggers observed by ANITA-3. Filled green triangles show currently applied corrections to UHECR energys; open blue circles show results of calculation described herein. projected at non-normal incidence, to the radius of curvature of the Earth. At λ∼1 m, and an observation distance (cid:112) d of order 100 km from the reflection point, one Fresnel zone is of order λd d /(d +d )∼300 m, so the curvature r s r s r correction is essentially negligible compared to the radius of curvature. A more sophisticated approach gives the reduction in net power at the receiver as the reduction in flux density due tocurvature,whichcanbeestimatedbygeometricconsiderations(thisargument,ofcourse,lacksaproperaccounting for phase variations which must be included). We can probe this possible effect experimentally by examining the widthoftheinterferometricpeakforHiCal-1datarecordedat3degreeelevationanglevs. 4degreeincidentelevation angle. However, within our limited statistics, we find that these two are consistent with each other. Alternately, we can estimate the effect of reflection from a sphere of radius R, with source at distance d and s elevation angle θ from the specular point and receiver at d and θ . The curvature of the reflecting surface results in s r r an ‘obscuration’ factor describing the fraction of the total possible reflecting Antarctic surface which remains visible from either HiCal (in transmission) or ANITA (in reception); we calculate this factor to be 61% at 5o incidence and 56% at 3.8o incidence. However, we stress that this is an extreme case, corresponding to the entire visible surface contributing to the HiCal signal observed by ANITA. We note that, although the signal is likely much greater than one Fresnel zone (in this case, of order 200 m), the limited time duration of the observed signals indicates a reflection area contributing to the detected signal of order 50 km or so in radius only.