Physics Capabilities of the IceCube DeepCore Detector
aa r X i v : . [ a s t r o - ph . I M ] J u l PROCEEDINGS OF THE 31 st ICRC, Ł ´OD ´Z 2009 1
Physics Capabilities of the IceCube DeepCore Detector
Christopher Wiebusch ∗ for the IceCube Collaboration †∗ III.Physikalisches Institut, RWTH Aachen, University, Germany † See the special section of these proceedings
Abstract . IceCube-DeepCore is a compactCherenkov detector located in the clear ice of thebottom center of the IceCube Neutrino Telescope.Its purpose is to enhance the sensitivity of IceCubefor low neutrino energies ( < TeV) and to lowerthe detection threshold of IceCube by about anorder of magnitude to below 10 GeV. The detector isformed by 6 additional strings of 360 high quantumefficiency phototubes together with the 7 centralIceCube strings. The improved sensitivity willprovide an enhanced sensitivity to probe a rangeof parameters of dark matter models not coveredby direct experiments. It opens a new window foratmospheric neutrino oscillation measurements of ν µ disappearance or ν τ appearance in an energyregion not well tested by previous experiments, andenlarges the field of view of IceCube to a full skyobservation when searching for potential neutrinosources. The first string was succesfully installed inJanuary 2009, commissioning of the full detector isplanned early 2010. Keywords : Neutrino-astronomy, IceCube-DeepCore
I. I
NTRODUCTION
Main aim of the IceCube neutrino observatory [1] isthe detection of high energy extraterrestrial neutrinosfrom cosmic sources, e.g. from active galactic nuclei.The detection of high energy neutrinos would help toresolve the question of the sources and the accelerationmechanisms of high energy cosmic rays.IceCube is located at the geographic South Pole. Themain instrument of IceCube will consist of cablestrings, each with highly sensitive photo-detectorswhich are installed in the clear ice at depths between m and m below the surface. Charged leptonswith an energy above GeV inside or close to thedetector produce enough Cherenkov light to be de-tected and reconstructed using the timing information ofthe photoelectrons recorded with large area phototubes(PMT). While the primary goal is of highest scientificinterest, the instrument can address a multitude of scien-tific questions, ranging from fundamental physics suchas physics on energy-scales beyond the reach of currentparticle accelerators to multidisciplinary aspects e.g. theoptical properties of the deep Antarctic ice which reflectclimate changes on Earth.IceCube is complemented by other major detectorcomponents. The surface air-shower detector
IceTop isused to study high energy cosmic rays and to calibrate (cid:1) (cid:1) (cid:1)(cid:2)(cid:3)(cid:4)(cid:5)(cid:6)(cid:7)(cid:8) (cid:1)(cid:2)(cid:3)(cid:9)(cid:3)(cid:8)(cid:10)(cid:11)(cid:12)(cid:13)(cid:14)(cid:15)(cid:1)(cid:16)(cid:17)(cid:2)(cid:3)(cid:18)(cid:1)(cid:19)(cid:20)(cid:2)(cid:3)(cid:9)(cid:21)(cid:13)(cid:14)(cid:3)(cid:22)(cid:23)(cid:24)(cid:25) (cid:17)(cid:2)(cid:3)(cid:4)(cid:5)(cid:6)(cid:7)(cid:8) (cid:16)(cid:3)(cid:9)(cid:3)(cid:8)(cid:10)(cid:11)(cid:12)(cid:13)(cid:14)(cid:15)(cid:26)(cid:1)(cid:2)(cid:16)(cid:27)(cid:26)(cid:28)(cid:17)(cid:2)(cid:3)(cid:9)(cid:21)(cid:13)(cid:14)(cid:3)(cid:15)(cid:22)(cid:23)(cid:23)(cid:14)(cid:25) (cid:24)(cid:29)(cid:8)(cid:30)(cid:3)(cid:31)(cid:11) (cid:23)(cid:22)!(cid:6)!"(cid:4)!(cid:4)(cid:23)(cid:23)(cid:10) (cid:4) (cid:29) (cid:8) (cid:30)(cid:3) (cid:12) $(cid:14) (cid:12) (cid:23)(cid:14) (cid:30) (cid:22) (cid:11) (cid:30) (cid:13) $(cid:14) (cid:21) (cid:29)(cid:14)(cid:24)(cid:23) (cid:22) (cid:23) (cid:8) (cid:30) (cid:13) (cid:9) (cid:11) (cid:30) (cid:23) (cid:8) (cid:3) (cid:24)(cid:23)(cid:23)(cid:10) (cid:3) (cid:13) (cid:12) (cid:23) (cid:3) (cid:12) (cid:31) (cid:11) (cid:22) (cid:13) (cid:30) (cid:25) (cid:12) (cid:31) (cid:23)(cid:11) (cid:22) (cid:23) (cid:8) (cid:30)(cid:3) (cid:13) (cid:12) (cid:23) (cid:24)(cid:23)(cid:10)(cid:31)$ (cid:23)(cid:24)(cid:29)(cid:14)(cid:24)(cid:23)(cid:10)(cid:31)$ (cid:23)(cid:24) %(cid:12)(cid:23)
Fig. 1: The geometry of the DeepCore Detector. Thetop part shows the surface projection of horizontal stringpositions and indicates the positions of AMANDA andDeepCore. The bottom part indicates the depth of sensorpositions. At the left the depth-profile of the opticaltransparency of the ice is shown.
IceCube . R&D studies are underway to supplement Ice-Cube with radio (
AURA ) and acoustic sensors (
SPATS )in order to extend the energy range beyond EeV en-ergies. DeepCore will consist of six additional denselyinstrumented strings deployed in the bottom center ofIceCube and is the focus of this paper.The first DeepCore string has been taking data sinceits successful installation in January 2009. The Deep-Core detector will be completed in 2010 and will replacethe existing
AMANDA-II detector, which was decomis-sioned in May 2009. DeepCore will lower the detectionthreshold of IceCube by an order of magnitude to below GeV and, due to its improved design, provide newcapabilities compared to AMANDA. In this paper wedescribe the design of DeepCore and the enhancedphysics capabilities that may be addressed.II. D
EEP C ORE D ESIGN AND G EOMETRY
The geometry of DeepCore is sketched in figure 1.
CHRISTOPHER WIEBUSCH et al.
CAPABILITIES OF ICECUBE DEEPCORE fi ciency (%) D O M s p y ( )2 High QEStandard IceCube
Fig. 2: Results of the quantum efficiency calibration at λ = 405 nm of the DeepCore phototubes compared tostandard IceCube phototubes.DeepCore is comprised of additional strings, eachof which are instrumented with phototubes, in con-junction with the central IceCube strings. The detectoris divided into two components. Ten sensors of eachnew string are at shallow depths between m and m, above a major dust-layer of poorer opticaltransparency and will be used as a veto-detector for thedeeper component. The deep component is formed by sensors on each string and is installed in the clearice at depths between m and m. It will form,together with the neighboring IceCube sensors the mainphysics volume.The deep ice is on average twice as clear as theaverage ice above m [2]. The effective scatteringlength reaches m and the absorption length m.Compared to AMANDA a substantially larger numberof unscattered photons will be recorded allowing foran improved pattern recognition and reconstruction ofneutrino events in particular at lower energies.Another important aspect is a denser spacing of photo-sensors compared to IceCube: The horizontal inter-stringspacing is m (IceCube: m). The vertical spacingof sensors along a string is only m (IceCube: m).The next major improvement with respect to IceCubeand AMANDA is the usage of new phototubes (HAMA-MATSU R7081-MOD) of higher quantum efficiency.This hemispherical 10” PMT is identical to the standardIceCube PMT [3], but employs a modified cathodematerial of higher quantum efficiency (typically %at λ = 390 nm). Calibrations of the phototubes forDeepCore confirm a sensitivity improvement of %- % with respect to the standard IceCube PMT (figure2). Also regular IceCube strings will be equipped withthese phototubes within the DeepCore volume.The net effect of the denser instrumentation is a factor ∼ gain in sensitivity for photon detection and superioroptical clarity of the ice. This is an imporant prerequisitefor a substantially lower detection threshold.III. D EEP C ORE P ERFORMANCE
The electronic hardware of the optical sensors isidentical to the standard IceCube module [3] and thissignificantly reduces the efforts for maintenance and operations compared to AMANDA. The DeepCore de-tector is integrated into a homogeneous data aquisitionmodel of IceCube which will be only supplemented byan additional trigger. Initial comissioning data of thefirst installed DeepCore string verifies that the hardwareworks reliably and as expected.The IceCube detector is triggered if typically a mul-tiplicity of sensors within ∼ µ s observe a signalcoincident with a hit in a neighboring or next to neigh-boring sensor. For each trigger, the signals of the fulldetector are transferred to the surface.For the sensors within the considered volume thedata taking is supplemented with a reduced multiplicityrequirement of typically − . As shown in figure 7, sucha trigger is sufficient to trigger atmospheric neutrinoevents down to a threshold of GeV, sufficiently belowthe anticipated physics threshold.The chosen location of DeepCore allows to utilize theouter IceCube detector as an active veto shield againstthe background of down-going atmospheric muons.These are detected at a ∼ higher rate than neutrinoinduced muons. The veto provides external informa-tion to suppress this background and standard up-goingneutrino searches will strongly benefit from a largersignal efficiency and a lower detection threshold as thedemands on the maturity of recorded signals decrease.Even more intriguing is the opportunity to identifydown-going ν induced µ , which may, unlike cosmic rayinduced atmosperic µ , start inside the DeepCore detec-tor. Simulations [4] show that three rings of surroundingIceCube strings and the instrumentation in the upperpart of IceCube are sufficient to achieve a rejection ofatmospheric muons by a factor > maintaining alarge fraction of the triggered neutrino signals. A furtherinteresting aspect is the proposal [6] to veto also atmo-spheric ν by the detection of a correlated atmospheric µ .This could provide the opportunity to reject a substantialpart of this usually irreducable background for extra-terrestrial neutrino searches.Triggered events which start inside the detector willbe selected online and transmitted north by satellite.Already simple algorithms allow to suppress the back-ground rate by a factor > and meet the bandwidthrequirements while keeping 90% of the signal [4]. Atypical strategy requires that the earliest hits are locatedinside DeepCore and allows for later hits in the veto-region only if the time is causally consistent with thehypothesis of a starting track. A filter which selectsstarting tracks in IceCube has been active since 2008and allows performance verification of such filters withexperimental data and to benchmark the subsequentphysics analysis.The filtered events are analyzed offline with moresophisticated reconstruction algorithms. Here, the focusis to improve the purity of the sample and to reconstructdirection, energy and the position of the interactionvertex. A particularily efficient likelihood algorithm( finiteReco [4]) capable of selecting starting muons ROCEEDINGS OF THE 31 st ICRC, Ł ´OD ´Z 2009 3 [GeV] ν E [ m ] R ec o L Entries 642Mean x 57.25Mean y 142.1RMS x 46.98RMS y 101.9Entries 642Mean x 57.25Mean y 142.1RMS x 46.98RMS y 101.9 preliminary
Fig. 3: The reconstructed length of µ contained tracksin DeepCore, based on the reconstructed start- and stop-vertex with the finiteReco algorithm. The data are ν induced µ -tracks from the upper hemisphere, whichare reconstructed to start within DeepCore. (Primary Neutrino Energy - GeV) log1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 ) E ff ec t i ve N e u t r i no A r ea ( m -6 -5 -4 -3 -2 -1 Preliminary
Fig. 4: Effective neutrino detection area of IceCube(trigger level) versus the energy for up-going neutrinos.The squares are IceCube only. The circles represent thearea if DeepCore is included.evaluates the hit probabilities of photomultipliers withand without a signal in dependence of the distance tothe track. It estimates the most probable position of thestart-vertex and provides the probablity that a track mayhave reached this point undetected by the veto.The reconstruction algorithms are still under develop-ment but initial results are promising. As an example,figure 3 shows the reconstructed length of µ tracks asfunction of the ν energy. Already the currently achievedresolution of ∼ m results in a visible correlation withthe neutrino energy in particular for energies ≤ GeV.Note, that the resolution is substantially better for verti-cal tracks.The effective detection area of IceCube for neutrinosfor triggered events is shown in figure 4. Despite Deep-Core being much smaller than IceCube, a substantialgain of up to an order of magnitude is achieved bythe additional events detected in DeepCore. Higher levelevent selections for specific physics analysis benefitstrongly from the higher information content of eventsand the gain of DeepCore further improves. Fig. 5: Interesting celestrial objects with known emissionof TeV gamma rays.IV. P
HYSICS P OTENTIAL
A. Galactic point sources of neutrinos
The analysis of IceCube data greatly benefits fromthe location at the geographic South Pole because thecelestial sphere fully rotates during one sideral day.Azimuthal detector effects are largly washed out becauseeach portion of the sky is observed with the same expo-sure and same inclination. However, the aperture of theconventional up-going muon analysis is restricted to onlythe Northern hemisphere and leaves out a large fractionof the galactic plane and a number of interesting objectssuch as the galactic center (see figure 5). Extending thefield of view of IceCube at low energies ( ≤ TeV) to afull sky observation will greatly enlarge the number ofinteresting galactic sources in reach of IceCube .The energy spectrum of gamma rays from supernovaremnants show indications of a potential cut-off at a fewTeV [7]. Under the assumption of a hadronic productionmechanism for these gamma rays the correspondingneutrino fluxes would show a similar cut-off at typ-ically half of that cut-off value. The high sensitivityof DeepCore for neutrinos of TeV energies and belowwill complement the sensitivity of IceCube which isoptimized for energies of typically TeV and above.
B. Indirect detection of dark matter
The observation of an excess of high energy neutrinosfrom the direction of the Sun can be interpreted bymeans of annihilations of WIMP-dark matter in itscenter. The energy of such neutrinos is a fraction of themass of the WIMP particles (expected on the TeV-scale)and it depends on the decay chains of the annihilationproducts. The large effective area of DeepCore and thepossibility of a highly efficient signal selection greatly Note, that at high energies > PeV the background of atmosphericmuons rapidly decreases and also here neutrinos from the Southernhemisphere can be detected by IceCube [8]. However, galactic sourcesare usually not expected to produce significant fluxes of neutrinos atenergies around the cosmic ray knee and above.
CHRISTOPHER WIEBUSCH et al.
CAPABILITIES OF ICECUBE DEEPCORE
Neutralino mass (GeV) ) N e u t r a li no - p r o t on S D c r o ss - sec t i on ( c m -41 -40 -39 -38 -37 -36 -35 -34 -33 -32 CDMS(2008)+XENON10(2007) limSI σ < SI σ CDMS(2008)+XENON10(2007) limSI σ < 0.001x SI σ < 0.20 h χ Ω -41 -40 -39 -38 -37 -36 -35 -34 -33 -32
10 10 -41 -40 -39 -38 -37 -36 -35 -34 -33 -32 Fig. 6: The expected upper limit of IceCube DeepCore at90% confidence level on the spin-dependent neutralino-proton cross section for the hard (W + W − ) annihilationchannel as a function of the neutralino mass for IceCubeincluding Deep Core (solid line). Also shown are limitsfrom previous direct and indirect searches. The shadedareas represent MSSM models which are not disfavouredby direct searches, even if their sensitivity would beimproved by a factor 1000.improves the sensitivity of IceCube. In particular it ispossible to probe regions of the parameter space withsoft decay chains and WIMP masses below ∼ GeV and which are not disfavored by direct searchexperiments.An example of the sensitivity for the hard annihilationchannel of supersymmetric neutralino dark matter isshown in figure 6.
C. Atmospheric neutrinos
DeepCore will trigger on the order of atmo-spheric neutrinos/year in the energy range from GeV to
GeV. Atmospheric neutrinos are largly unexploredin this energy range. Smaller experiments like Super-Kamiokande cannot efficiently measure the spectrumfor energies above GeV and measurements done byAMANDA only start at TeV. In the range between − GeV decays of charged kaons become dominantover decays of charged pions [10] for the production ofatmospheric neutrinos and the systematic error of fluxcalculations increases. A measurement of this transitioncould help to reduce systematic errors of the flux ofatmospheric neutrinos at TeV energies.The first maximum of disappearence of atmospheric ν µ due to oscillations appears at an energy of about GeV for vertically up-going atmospheric neutrinos[5]. The energy threshold of about GeV would allowto measure atmospheric neutrino oscillations by meansof a direct observation of the oscillation pattern inthis energy range. In addition, DeepCore would aim toobserve the appearance of ν τ by the detection of smallcascade-like events in the DeepCore volume at a ratewhich is anti-correlated with the disappearance of ν µ . Primary Neutrino Energy - GeV10 20 30 40 50 60 70 80 90 10005001000150020002500300035004000
Preliminary
Fig. 7: Number of triggered vertical atmospheric neu-trinos per year (per 3GeV) versus the neutrino energy.Events from . π sr are accepted. Shown are the num-bers without (squares) and with (circles) the inclusionof oscillations ( ∆ m atm = 0 . eV , sin(2 θ ) = 1 ).Similar to ν τ , the signature of ν e events are cascade-events with a large local light deposition without thesignature of a track. The dominant background to theseevents are charged current ν µ interactions with a smallmomentum transfer to the µ . Analyses like these willhave to be performed considering all three flavors andtheir mixing. Note, that only for a further reduction ofthe energy threshold smaller than GeV matter effectsin the Earth’s core would become visible [5].
D. Other physics aspects
Two remaining items are only briefly mentioned here.Slowly moving magnetic monopoles, when catalizingproton decays, produce subsequent energy depositionsof ∼ GeV along their path with time-scales of µ s toms. Initial studies are under-way to develop a dedicatedtrigger for this signature using delayed coincidences.DeepCore extends the possibility to search for neu-trino emission in coincidence with gamma ray bursts(GRBs) to lower energies. According to [11] GRBs mayemit a burst of neutrinos. However, predicted energiesare only a few GeV and the event numbers are small( ∼ yr − km − ). Additional studies are required toevaluate the sensitivity for such signals.V. S UMMARY AND O UTLOOK
This paper summarizes the enhancement of thephysics profile of IceCube by the DeepCore detector.The geometry of DeepCore has been optimized andconstruction has started. Detailed MC studies and ex-perimental analyses are currently under way to optimizeand finalize the analysis procedures. First data from thefull detector will be available in spring 2010, the vetowill be fully completed latest 2011.A
CKNOWLEDGEMENT
This work is supported by the German Ministry forEducation and Research (BMBF). For a full acknowl-edgement see [1].
ROCEEDINGS OF THE 31 st ICRC, Ł ´OD ´Z 2009 5 R EFERENCES[1] J. Ahrens et al. (IceCube Collaboration), Astropart. Phys. (2004) 507-532, 2004[2] M. Ackermann et al. (IceCube Collaboration), J. of Geophys.Res. (2006) D13203, July 2006[3] R. G. Stokstad et al. (IceCube Collaboration), Nucl. Phys. B(Proc. Suppl.)118