SuperPINGU for measurement of the leptonic CPphase with atmospheric neutrinos
Abstract
We explore a possibility to measure the CPviolating phase using multimegaton scale ice or water Cherenkov detectors with low, GeV, energy threshold assuming that the neutrino mass hierarchy is identified. We elaborate the relevant theoretical and phenomenological aspects of this possibility. The distributions of the (track) and (cascade) events in the neutrino energy and zenith angle plane have been computed for different values of . We study properties and distinguishability of the distributions before and after smearing over the neutrino energy and zenith angle. The CPviolation effects are not washed out by smearing, and furthermore, the sensitivity to increases with decrease of the energy threshold. The events contribute to the CPsensitivity as much as the events. While sensitivity of PINGU to is low, we find that future possible upgrade, SuperPINGU, with few megaton effective volume at () GeV and e.g. after 4 years of exposure will be able to disentangle values of from with “distinguishability” ( significance in ’s) correspondingly. Here the intervals of are due to various uncertainties of detection of the low energy events, especially the flavor identification, systematics, etc.. SuperPINGU can be used simultaneously for the proton decay searches.
pacs:
14.60.Pq, 14.60.StI Introduction
Discovery of the leptonic CP violation and measurement of the Dirac CP phase are among the main objectives in neutrino physics and, in general, in particle physics. They may have fundamental implications for theory and important consequences for phenomenology of atmospheric and accelerator neutrinos, high energy cosmic neutrinos, etc. reviews .
The present experimental results have very low sensitivity to giving only weak indications of the preferable interval of its values. Thus, the T2K and reactor data favor the interval with central value t2k . Analysis of the SuperKamiokande atmospheric neutrino data gives preferable range himmel . The global fit of all oscillation data, e.g. from fogli , agrees with these results: at level and no restriction appears at level. The values around are disfavored. Similar results with the best fit value (NH) have been obtained in valle and with in nufit .
A possibility to measure is generally associated with accelerator long baseline (LBL) neutrino experiments. There is certain potential to improve our knowledge of with further operation of T2K and NOvA ghosh . Proposals of more remote experiments, which will measure with reasonable accuracy, include LBNE lbne , JPARC  HyperKamiokande hyperk , ESS ess and LBNO lbno . Further developments can be related to the low energy neutrino and muon factories, beta beams, etc., see reviews .
Another possibility to determine is to use the atmospheric neutrino fluxes and large underground/underwater detectors. Sensitivity of future atmospheric neutrino studies by HyperKamiokande (HK) has been estimated in hyperk : During 10 years of running with fiducial volume Mton the HK will be able to discriminate the values of phases at about CL. ICAL at INO alone will have very low sensitivity, but combined with data from T2K and NOvA, it will reduce degeneracy of parameters, and thus, increase the global sensitivity ghosh2 .
Various theoretical and phenomenological aspects of the CPviolation in atmospheric neutrinos have been explored in a number of publications before Peres:2003wd ; Kimura1 ; GonzalezGarcia:2004wg ; our2 ; our3 ; Mena:2008rh ; ARS ; latimer ; Agarwalla:2012uj ; ohlsson ; ge1 ; ge2 . In particular, pattern of the neutrino oscillograms (lines of equal probabilities in the plane) with CP violation has been studied in details in our3 . It was realized that structure of the oscillograms is determined to a large extent by the grid of the magic lines of three different types Barger:2001yr ; Huber:2003ak ; Smirnov:2006sm ; our3 (solar, atmospheric and interference phase lines). Although at the probability level the effects of the CPviolation can be of order 1, there are a number of factors which substantially reduce the effects at the level of observable events ARS .
Capacities of new generation of the atmospheric neutrino detectors (PINGU, ORCA) have been explored recently pingu ; ARS ; pingu2 ; orca ; Winter:2013ema . It was found ARS ; pingu2 that these detectors with GeV have good sensitivity to the neutrino mass hierarchy and the parameters of the 23 sector (the 23 mixing and mass splitting). However, the CPviolation effects turn out to be subleading. This helps in establishing the hierarchy without serious degeneracy with in contrast to the accelerator experiments, but the information on the CPphase will be rather poor.
The goal of this paper is twofold: (i) detailed study of the CPviolation effects in atmospheric neutrinos, and (ii) tentative estimation of sensitivity to the CPphase of future large detectors, assuming that the neutrino mass hierarchy is identified. We will show that in spite of averaging of oscillation pattern over the neutrino energy and direction, the CP violation effects are not washed out, and furthermore, increase with lowering the energy threshold . This opens up a possibility to measure using multimegaton scale ice or water Cherenkov detectors with GeV. We study dependence of the energy and zenith angle distributions of events produced by and on the CP phase. We estimate distinguishability of different values of . According to the present proposal pingu2 PINGU will have low sensitivity to and only further upgrades, which we will call SuperPINGU, can measure with potentially competitive accuracy. We discuss requirements for such detectors. We identify problems and challenges of these CP measurements, and propose ways to resolve or mitigate the problems. We formulate conditions, in particular on accuracies of knowledge of external parameters and level of flavor misidentification, to achieve the goal.
The paper is organized as follows. In Sec. II we summarize relevant information on the oscillation probabilities and their dependence on CPphase. We present analytical formulas for the probabilities in quasiconstant density approximation. The grid of the magic lines will be described and we will show how the grid determines structure of oscillograms. In Sec. III we consider a possible upgrade of PINGU, called SuperPINGU, which will be able to measure and outline a procedure of computation of numbers of events. In Sec. IV we compute the distributions as well as relative differences of distributions of the events in the plane (the relative CPdifferences) for different values of . We study dependence of these distributions on before and after smearing over the neutrino energy and direction. In Sec. V we perform similar studies of the cascade (mainly ) events. Sec. VI contains estimations of the total sensitivity of SuperPINGU to and discussion of our results. We conclude in Sec. VII.
Ii Oscillation probabilities, CPdomains
ii.1 Oscillation amplitudes and probabilities
We will study the CPviolation phase defined in the standard parametrization of the PMNS mixing matrix, , where is the matrix of rotation in the plane and . We consider evolution of the neutrino states in the propagation basis, determined by the relation . In this basis the CP dependence is dropped out from the evolution and appears via projection of the propagation states back onto the flavor states at the production and detection. Due to this, dependence of the probabilities and numbers of events on is simple and explicit. Therefore the results will be presented in terms of amplitudes in this basis (see our3 for details), where the matrix of amplitudes is defined as
Here we have taken into account the equalities and valid for symmetric density profile and in absence of the fundamental CP and T violation in the propagation basis. In the low energy domain, GeV, i.e. below the 13 resonance, one can further decrease the number of amplitudes involved down to 3 (see Peres:2003wd and comment ^{1}^{1}1The basis used in the paper Peres:2003wd differs from the basis considered here by the additional 13 rotation on the 13 mixing in matter. This basis is useful for description of oscillations at low energies (in the subGeV range) since it allows to make certain approximations which simplify description. Namely, neglecting changes of 13 mixing in matter with distance one can reduce 3neutrino evolution problem to 2neutrino evolution problem. Correspondingly all the probabilities can be expressed in terms of just three real functions , and . The main dependence on 13 mixing as well as on is explicit here. The formulas in Peres:2003wd are approximate, and in general they are not valid at high energies (in multiGeV range). Since the highest sensitivity to CP is at low energies these formulas give accurate description of CPeffects.).
The oscillation probabilities can be written as
(1) 
where and are the independent and dependent parts of the probability , respectively. Notice that , since contains terms which are proportional to , generally even on , and these terms do not disappear when . Then the total probability is . The probabilities equal our3
(2)  
The amplitude is doubly suppressed by small quantities and our3 . Therefore terms that are quadratic in can be neglected in the first approximation of our analytical study. For the dependent parts we have then our3 ,
(3) 
where , and
(4) 
Here
The term is small if the 23 mixing is close to the maximal one, and as we said, in addition the amplitude is small. Let us emphasize that in the phase dependence, , factors out, whereas in it appears in combination with the oscillation phase .
In matter with symmetric density profile one has for the inverse channels
in particular, . For antineutrinos the probabilities have the same form as for neutrinos with substitution:
(5) 
where and are the mixing angles and phases in matter for antineutrinos, and is the matter potential. In particular,
(6) 
ii.2 Quasiconstant density approximation
One can further advance in analytical study using explicit expressions for the amplitudes in the constant (or quasiconstant) density approximation our3 (see also kimura2 and Blennow:2013vta ). According to this approximation, at high energies for a given trajectory in mantle one can use the mixing angles computed for the average value of the potential . For low energies, where adiabaticity condition is fulfilled, the mixing angle is determined by the surface density. The oscillation phases, however, should be computed by integration over the neutrino trajectory. For corecrossing trajectories one can use the three layer model with constant densities in each layer; corrections are computed in our3 .
In the case of constant density our3
(7)  
(8) 
The halfphases equal in the high energy range (substantially larger than the 12 resonance, GeV):
(9) 
Here with being the radius of the Earth,
and the upper (lower) sign corresponds to neutrinos (antineutrinos). For two other phases we obtain
(10)  
(11) 
where is given in (9). In practical cases the terms can be neglected. For low energies (close to the 12 resonance):
(12) 
Notice that in the energy range above the 12 resonance and the amplitude (8) is reduced to the two neutrino form, which corresponds to factorization our3 .
ii.3 Numerical results
We have computed the probabilities by performing numerical integration of the evolution equation for the complete system. We used the PREM density profile of the Earth prem and the values of the neutrino parameters eV, eV, , and , which are close to the current best fit values fogli . We assume the normal neutrino mass hierarchy in the most part of the paper.
In Fig. 1 we show the oscillation probabilities and as functions of the neutrino energy for different values of CPphase and zenith angles. In the low energy range where sensitivity to is high and consider . In Fig. 1 the resonantly enhanced probability due to the 12 mixing and mass splitting is modulated by fast oscillations driven by the 13 mass and mixing. The 12 resonance energy in the mantle is at GeV. For core crossing trajectories (upper panels) the parametric effects distort the dependence of probability on energy.
The key feature which opens up a possibility to measure is the presence of systematic shift of the oscillatory curves (probabilities) at low energies, GeV, with increase of the phase. The shift occurs in the same way in wide energy interval GeV, and essentially for all trajectories which cross the mantle only. This systematic shift can be understood using analytical expressions for the probabilities. Averaging (15) over fast oscillations driven by the 13 splitting we find
(18) 
The first term does not change the sign with , whereas the second one does. Notice that above the 12 resonance , and so
(19) 
The difference of probabilities for a given value and equals:
(20) 
The first term is positive for all and , and it is this term that produces a systematic shift of the probabilities.
For the values of phase shown in Fig. 1 we obtain from (18)
(21)  
These equations show that is the smallest one. The probability increases with and reaches maximum at . For the trajectory with , the oscillation phase equals . That leads to , and consequently, to equal total probabilities. For the phase equals which gives different values of probability: , and furthermore . These results are in agreement with plots shown in Fig. 1.
Although there is certain phase shift with change of , the sizes of energy intervals where the difference has positive and negative signs are strongly different. One sign dominates, and therefore there is no averaging over energy. Maximal relative upward shift of the probability curves compared to the curve is around GeV. For the corecrossing trajectories () due to the parametric effects the transition probability first increases with increase of , it reaches maximum at and then decreases.
The probability, (17), averaged over the 13 oscillations equals
(22) 
where the term is neglected. Notice immediately that the CPeffect in the channel has an opposite sign with respect to that in the channel (18). Therefore the presence of both and original fluxes weakens the total CPeffect, and consequently, the sensitivity to which is unavoidable. We will call this the flavor suppression.
According to (22) dependence of the probability on factors out and therefore turns out to be very simple. The maximal effect is for ,
and , so that the total probabilities are equal for and which in perfect agreement with result of Fig. 1.
The difference of probabilities for a given value of and zero phase equals
(23) 
Only CPeven contribution is present.
The probabilities in antineutrino channels are shown in Fig. 2. Their dependencies on and can be immediately understood from our analytical treatment. According to (5) the averaged probabilities equal
(24)  
(25) 
For energies far above the 12 resonance, the expressions are further simplified since (recall, for neutrinos ):
(26) 
Comparing this with (20) and (22) we find that for antineutrinos the probabilities have opposite sign with respect to the probabilities for neutrinos. Indeed, according to Fig. 2 for mantle trajectories the biggest amplitude is for and the smallest one is for which is opposite to the case. This means that summation of signals from neutrinos and antineutrinos reduces the effect of CPphase, and consequently, the sensitivity to this phase. This Csuppression can be reduced if and signals are separated at least partially (see Sec. IV C).
As follows from Fig. 2 for the mantle crossing trajectories, only the largest CP effect on is in the range GeV where maximal values equal and for and correspondingly. These numbers are about 2 times smaller than for neutrinos. The reason is that, in the case of NH for neutrinos both and are enhanced in matter whereas for antineutrinos both and are suppressed. The antineutrino probabilities decrease with increase of energy above 0.8 GeV. This, as well as smaller crosssections suppresses number of events and therefore reduces cancellation of the CPeffect.
Similar consideration can be performed for the channel for which . Notice that in vacuum , i.e. the probability is even function of . In the matter dominated region we have due to change of sign of the potential. The differences of the antineutrino probabilities for a given and equals at
(27)  
(28) 
They also have an opposite sign with respect to the differences for neutrinos (20) and (23), and equal up to change of mixing angles and phases in matter.
ii.4 Magic lines and CPdomains
In what follows we will study differences of probabilities as well as distributions of events in the plane for different values of . The patterns of distributions are determined to a large extent by the grid of the magic lines Barger:2001yr ; Huber:2003ak ; Smirnov:2006sm ; our3 . The lines fix the borders of the CPdomains – the regions in the plane of the same sign of the CPdifference.
Let us summarize relevant information about properties of the magic lines. Recall that the magic lines are defined as the lines in the plane along which the oscillation probabilities do not depend on phase in the so called factorization (quasi ) approximation our3 . Correspondingly, the CPdifferences vanish along these lines.

The solar magic lines are determined by the condition
(29) where in neutrino channels is given by the expression (12) for valid in framework below 13 resonance but extended to all the energies. For antineutrinos in the NH case everywhere. Along these lines below the 13 resonance. That would be the line of zero solar amplitude in the approximation. The minimum of probability at GeV for in Fig. 1 corresponds to the first magic line with . The minimum at GeV for (Fig. 1) is on the second magic line with .
Notice that the energy of minimal level splitting (maximal oscillation length) is given by GeV which is much bigger than GeV due to large 12 mixing. So, below 0.7 GeV the splitting increases and correspondingly the oscillation length decreases. Therefore the same phase can be obtained for smaller , and consequently, the solar magic lines bend toward smaller . At energies much above the 12 resonance these lines do not depend on energy and are situated at
(30) for , and correspondingly.

The atmospheric magic lines are determined by the equality
(31) Along these lines . It would vanish exactly in the approximation, when that is far above the 12 resonance. Zeros of the probability at GeV (see Fig. 1) which do not depend on are situated on the atmospheric magic lines. E.g., for these points are at GeV and GeV. For , zeros are at GeV. For the solar and atmospheric magic lines coincide with those for in the limit .

The interference phase lines are important for distinguishing different values of the CPphase: a given value and a different value . Along these lines . According to (3) for the condition reads
where and the latter is the vacuum oscillation phase. This condition corresponds to intersection of probability curves for different values of phases and in Fig. 1. For the condition can be written as or
(32) For the inverse channel, , the sign of should be changed. According to (4) dependencies of the probability on and factor out in the approximation , and the corresponding interference phase line is determined by the condition , or
The condition can be written as
(33) where is the Earth radius and in general should lead to the vanishing CPdifference of probabilities.
The exact value of interference phase does not coincide with . In the constant density approximation equals the phase of the expression in brackets of (8):
(34) Notice that would be equal , if . The latter is satisfied for high energies , where . However, if we can not neglect the second term in the denominator of (34). Notice that in the limit we obtain from (16)
(35) where one can see immediately all three “magic” conditions.
Notice that magic lines could be introduced immediately in the framework as the lines along which . In this case they would, indeed, determine the borders of domains with different sign of the CPdifference of the probabilities. We use the original definitions of magic lines to match with previous discussions. Still as we said before, the solar, atmospheric and interference lines nearly coincide with the exact lines of zero CPdifferences in certain energy intervals or in asymptotics. The corresponding phases are related as
where is the phase in vacuum. So, the true lines of zero difference of probabilities interpolate between the magic lines (see our3 for details).
Iii PINGU, SuperPINGU and CP
The key conclusion of the previous section is that integration over the neutrino energy and direction does not suppress the CP effect significantly. Furthermore, for all trajectories which cross the mantle of the Earth only, the CP violation effect is similar: it has the same sign and the same change with . Effect is different for the core crossing trajectories, . So, there could be partial cancellation due to smearing over the zenith angle. Another important feature is that the relative CP effect at the probability level increases with decrease of energy. In this connection we will explore a possibility to measure using multimegaton scale neutrino detectors with low energy threshold. As it was already realized in ARS , sensitivity of PINGU to is low. So, we will consider future possible upgrades of PINGU. We will also quantify capacity of PINGU to obtain information about . For definiteness we will speak about PINGU for which more information is available. Similar upgrades can be considered for ORCA detector orca .
iii.1 PINGU and SuperPINGU
We calculate event rates for the proposed PINGU detector and for possible future PINGU upgrade which we call SuperPINGU. The PINGU detector pingu2 will have 40 strings additional to the DeepCore strings with 60 digital optical modules (DOM’s) at 5 m spacing in each string. A compact array like PINGU could detect neutrinos with energies as low as GeV. Strict criteria allow over efficiency of event reconstruction for all 3 flavors pingu2 . We parametrize the PINGU effective mass as
(36) 
and
(37) 
respectively for and . Here is the effective volume and is the density of the ice. These parametrizations well represent simulated volumes pingu2 from GeV up to 25 GeV. We will use an accuracy of the energy and angle reconstruction for PINGU from pingu2 .
Along with the PINGU proposal the idea has been discussed to construct “ultimate” multimegatonscale detector MICA with a threshold about 10 MeV allowing to detect the solar and supernova neutrinos mica . Clearly reducing the threshold by more that 2 orders of magnitude is very challenging. In this connection we would like to consider a kind of intermediate step  the detector with an effective energy threshold about (0.1  0.2) GeV, i.e. one order of magnitude below the threshold in the present PINGU proposal. For this, a denser array of DOM’s is required which will lead to increase of the effective volume of a detector at low energies. For definiteness we will take the effective volume which corresponds to the PINGU detector simulations with a total of 126 strings and 60 DOM’s per string each cowen . The effective mass can be parameterized as
(38) 
for both and events. We call this version SuperPINGU. According to (36) and (38) the effective mass, , in the range (1  2) GeV equals Mton for PINGU and Mton for SuperPINGU, i.e. 4 times larger. For the bin below 1 GeV the corresponding numbers are 0.3 and 2.2 Mton (7 times larger). This can be compared with MICA, which may have 220 strings and 140 DOMs per string. We will extrapolate to lower energies some PINGU characteristics from the proposal pingu2 .
Going to further upgrade has double effect:

increase of the effective volume, especially in the low energy bins, and

improvements of reconstruction of the neutrino energy and direction as well as the flavor identification of events for all energies.
SuperPINGU will have three times denser DOM array than PINGU. Therefore it will collect about 3 times more photons from the same event (with the same neutrino energy). Recall that, in PINGU the average distance between DOM’s is smaller than the photon scattering length (50 m).
We describe uncertainties of reconstruction of the neutrino energy and direction by smearing functions
where and ( and ) are the true (reconstructed) energy and zenith angle of the neutrinos. For PINGU we use and from pingu2 determined down to energies GeV. The distributions are normalized in such a way that
Notice that PINGU distributions have longer tails than the Gaussian functions.
Characteristics of the SuperPINGU reconstruction are expected to be better. We estimate parameters of and for SuperPINGU using the DeepCore resolutions and the simulated PINGU resolutions dcres , gross in the following way. For a given event the number of photons collected is proportional to the density of DOMs, that is for fixed total volume of the detector. Therefore the relative statistical error in determination of characteristics is proportional to , so we can assume that
(39) 
Estimations of resolutions of the DeepCore and PINGU confirm (39). Indeed, DeepCore has about DOM’s, while PINGU (40 strings with 60 DOM’s per string) will have DOM’s (also with higher quantum efficiency), that is, . Since the density of DOM’s in PINGU is about 4.5 times larger, amount of light detected from the same event will be about 4.5 times larger. According to dcres , gross and pingu2 for the events the ratio of resolutions (median errors)
(40) 
The ratio equals 0.66 at GeV, however estimation of DC parameters become not very reliable at low energies. For the events the improvement is even better: The ratio of median errors (40) is () in the interval GeV and it becomes 0.6 at 5 GeV.
For neutrino energy reconstruction (median energy resolution) of the events we have
(41) 
in the interval GeV. It decreases down to 0.52 at GeV. Similar improvement is expected for the events.
The SuperPINGU will have 3 times larger number (and therefore density) of DOM’s, than PINGU. Therefore according to (39) the resolutions will be further improved by factor . So for SuperPINGU we use the resolution functions from Fig. 7 and 8 of pingu2 , scaling their widths as
(42) 
We extrapolate these functions down to GeV and for simplicity neglect possible dependences of the factors in Eq. (42) on energy. (Notice that according to pingu2 the median value of angle is very similar for cascades and tracks.)
This estimation of improvement can be considered as conservative. Indeed, the DeepCore characteristics have been obtained after stringent kinematical cuts which allows one to select a sample of high quality events. That reduces efficiency of reconstruction (fraction of reconstructed events) down to , whereas PINGU characteristics have been obtained with efficiency. With stronger cuts in PINGU the reconstruction characteristics could be even better. Also, developments of electronics may lead to further improvements. Clearly, configuration of SuperPINGU should be optimized taking into account also the cost of construction. For large density of strings the issue of the ice stability may become important. One can reduce number of strings by increasing number of DOMs per string (decreasing vertical spacing). Since typical size of an event is about 100 m, for distances between strings (17 m) the total number of DOMs in the unit volume matters and geometry plays only secondary role. Another option is to consider underwater detector, i.e., an upgrade of ORCA.
iii.2 Distributions of events in the neutrino energy and zenith angle plane
To evaluate sensitivity of SuperPINGU to we will compute the distributions of events of different types and explore their dependence on . The numbers of events , produced by neutrinos () with energies and zenith angles in small bins and marked by subscript equal
(43) 
Here is the exposure time, is the Avogadro’s number. The density of events of type , , (the number of events per unit time per target nucleon) is given by
(44) 
where and are the fluxes of neutrinos and antineutrinos at the detector which produce events of the type , and and are the corresponding crosssections. In turn, the fluxes at the detector equal
and are the original muon and electron neutrino fluxes at the production.
With decrease of energy, resonance processes (pion production) and quasielastic processes will contribute, and the latter dominates below 1 GeV. In our estimations we use the total neutrinonucleon crosssections down to (0.2  0.3) GeV as they are parametrized in pingu2 , We assume that different contributing processes would produce visible effect at the detector with the same efficiency. For antineutrinos there is no data below 1 GeV and we use extrapolation given in pingu2 . Clearly in future these computations should be refined.
We use the atmospheric neutrino fluxes, and (and corresponding fluxes of antineutrinos) from Refs. Honda:1995hz ; Athar:2012it . At low energies the geomagnetic effects become important which break azimuthal symmetry.
After smearing in the () plane, we obtained the unbinned distribution of events as
(45) 
, and then binned them according to