The BAIKAL neutrino project: status, results and perspectives

V.Aynutdinova, V.Balkanovb, I.Belolaptikovc, N.Budnevd, L.Bezrukovb, A.Chenskyd, D.Chernove,I.Danilchenkob, Zh.-A.Dzhilkibaevb, G.Domogatskyb, A.N.Dyachokd, S.Fialkovskyf, O.Gaponenkob,O.Gressd, T.Gressd, K.Kazakovd, A.Klabukovb, A.Klimovg, S.Klimushinb, K.Konischevb,A.Koshechkinb, L.Kuzmicheve, V.Kulepovf , Vy.Kuznetzovb, B.Lubsandorzhievb, S.Mikheyevb,M.Mileninf , R.Mirgazovd, E.Osipovae, A.Pavlovd, G.Pan’kovd, L.Pan’kovd, A.Panfilovb, Yu.Parfenovd,E.Pliskovskyc, P.Pokhilb, V.Polecshukb, E.Popovae, V.Prosine, M.Rosanovh, V.Rubtzovd, Y.Semeneyd,B.Shaibonovb, Ch.Spieringi, O.Streicheri, B.Tarashanskyd, R.Vasilievc, E.Vyatchinb, R.Wischnewskii,I.Yashine, V.Zhukovb

aInstitute for Nuclear Research, Moscow, Russia

bInstitute for Nuclear Research, Moscow, Russia

cJoint Institute for Nuclear Research, Dubna, Russia

dIrkutsk State University, Irkutsk,Russia

eSkobeltsyn Institute of Nuclear Physics MSU, Moscow, Russia

fNizhni Novgorod State Technical University

gKurchatov Institute, Moscow, Russia

hSt.Peterburg State Marine University, St.Petersburg, Russia

iDESY–Zeuthen, Zeuthen, Germany

We review the present status of the Baikal Neutrino Project and present selected results obtained from datataken in 1998 - 2000 (780 live days). We describe the moderate upgrade of NT-200 planned for the next yearsand discuss a possible detector on the Gigaton scale.

1. DETECTOR AND SITE

The Baikal Neutrino Telescope is operated inLake Baikal, Siberia, at a depth of 1.1 km. Thepresent stage of the telescope, NT-200 [1], wasput into operation at April 6th, 1998 and consistsof 192 optical modules (OMs). An umbrella-likeframe carries 8 strings, each with 24 pairwise ar-ranged OMs. Three underwater electrical cablesand one optical cable connect the detector withthe shore station. The OMs are grouped in pairsalong the strings. They contain 37-cm diameterQUASAR - photo multipliers (PMs), which havebeen developed specially for our project [2]. Thetwo PMs of a pair are switched in coincidence in

order to suppress background from biolumines-cence and PM noise. A pair defines a channel. Atrigger is formed by the requirement of ≥ N hits(with hit referring to a channel) within 500 ns. Nis typically set to 3 or 4. For such events, ampli-tude and time of all fired channels are digitizedand sent to shore. A separate monopole triggersystem searches for clusters of sequential hits inindividual channels which are characteristic forthe passage of slowly moving, bright objects likeGUT monopoles.

Lake Baikal deep water is characterized by anabsorption length of Labs(480 nm)=20 ÷24 m, ascattering length of Ls =30 ÷70 m and a strongly

Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342

0920-5632/$ – see front matter © 2005 Published by Elsevier B.V.

www.elsevierphysics.com

doi:10.1016/j.nuclphysbps.2005.01.126

Zenith cosine

-1 -0.8 -0.6 -0.4 -0.2 -0

Nu

mb

er o

f ev

ents

0

10

20

30Experiment (84 ev)

MC without oscl. (80.5 ev)

∆m2= 2.5x10-3eV (71.6 ev)

Figure 1. Angular distribution of experimentalevents and MC data.

anisotropic scattering function f(θ) with a meancosine of the scattering angle cos(θ) = 0.85÷ 0.9.In contrast to underground detectors, open con-figurations in highly transparent media like wateror ice allow to observe a huge volume beyond theirgeometrical boundaries.

Here we present selected results obtained fromdata taken in 1998 - 2000 (780 live days). Datataken in 2001 are presently being analyzed. Wealso describe NT-200+, an upgrade of NT-200 bythree sparsely instrumented distant outer stringswhich increase the fiducial volume for high energycascades to order of 10 Mtons. Two of three outerstrings where deployed, and electronics, data ac-quisition and calibration systems for NT-200+have been tested in March 2004.

2. Atmospheric Muon NeutrinosThe clearest signature of neutrino induced

events is a muon crossing the detector from below.Track reconstruction algorithms as well as back-ground rejection have been described elsewhere[3]. The energy threshold of NT-200 for this par-ticular analysis (15-20 GeV) is much smaller thanof Amanda (∼50 GeV) but still too high for aclear appearance of oscillation effects, given thelow statistics and the systematic uncertainties.

24 h 0 h

-90 o

90 o

Figure 2. Skyplot (equatorial coordinates) of neu-trino events.

Atmospheric neutrinos serve as an important cal-ibration tool and demonstrate the understandingof the detector performance. The data set of years1998+1999 yields 84 upward going muons. TheMC simulation of upward muon tracks due to at-mospheric neutrinos gives 80.5 events. The angu-lar distribution for both experiment and simula-tion as well as the skyplot of upward muons areshown in Fig. 1,2.3. Search for Neutrinos from WIMP An-

nihilationThe search for WIMPs with the Baikal neu-

trino telescope is based on a possible signal ofnearly vertically upward going muons, exceedingthe flux of atmospheric neutrinos. The methodof event selection relies on the application of aseries of cuts which are tailored to the responseof the telescope to nearly vertically upward mov-ing muons [4]. The applied cuts select muonswith -1< cos(θ) <-0.75 and result in a detec-tion area of about 1800 m2 for vertically up-ward going muons. The energy threshold forthis analysis is Ethr ∼ 10 GeV i.e. significantlylower then for the analysis described in section 2(Ethr ∼ 15− 20 GeV). Therefore the effect of os-cillations is stronger visible. We expect a muonevent suppression of (25-30)% due to neutrino os-cillations assuming δm2 =2.5·10−3 eV2 with fullmixing, θm ≈ π/4.

From 502 days of effective data taking between

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342336

Figure 3. Angular distributions of selected neu-trino candidates as well as expected distributionsin a case with and without oscillations (solid anddashed curves respectively).

April 1998 and February 2000, 24 events with -1< cos(θ) <-0.75 have been selected as clear neu-trino events. The angular distribution of theseevents as well as the MC - predicted distribu-tions are shown in Fig. 3. For the MC simula-tions we used the Bartol96 atmospheric neutrinoflux [5] without (dashed curve) and with (solidcurve) oscillations. Within 1σ statistical uncer-tainties the experimental angular distribution isconsistent with the prediction including neutrinooscillations.

Regarding the 24 detected events as being in-duced by atmospheric neutrinos, one can derivean upper limit on the additional flux of muonsfrom the center of the Earth due to annihila-tion of neutralinos - the favored candidate forcold dark matter. The 90% C.L. muon flux lim-its for six cones around the opposite zenith aswell as muon flux limits for different neutralinomasses obtained with NT-200 (Ethr >10 GeV) in1998/99 are shown in Fig. 4 and Fig. 5, and com-pared to limits obtained by Baksan [6], MACRO[7], Super-Kamiokande [8] and AMANDA [9].

Figure 4. Limits on the excess muon flux from thecenter of the Earth versus half-cone of the searchangle.

Figure 5. Limits on the excess muon flux fromthe center of the Earth as a function of WIMPmass.

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342 337

10-17

10-16

10-15

0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1β=v/c

uppe

r lim

it (c

m-2

s-1

sr-1

)

Soudan

KGF

Ohya

MACRO

Baikal (98-00)

AMANDA

Figure 6. Upper limits on the flux of fastmonopoles obtained in different experiments.

4. Search for Relativistic MagneticMonopoles

Events due to relativistic monopoles (β > 0.75)are distinguished by their high light output, al-lowing identification of events beyond the geo-metrical boundaries of the detector. The searchstrategy has been described in [10]. An improvedanalysis including data from 1996 to 2000 yields alimit about a factor of four below the limit pub-lished earlier. This limit is compared to thosefrom other experiments ([11–15]).

5. A Search for Extraterrestrial High En-ergy Neutrinos

The BAIKAL survey for high energy neutrinossearches for bright cascades produced at the neu-trino interaction vertex in a large volume aroundthe neutrino telescope. Lack of significant lightscattering allows to monitor a volume exceedingthe geometrical volume by an order of magnitude.This results in sensitivities of NT-200 compara-ble to those of the much larger AMANDA detec-tor. The background to this search are brightbremsstrahlung flashes along downward muonspassing far outside the array.

Figure 7. Energy dependence of effective volume.

For the analysis of data recorded in 1998 -2000 (780 live days) we used 18384 events withhit channel multiplicity Nhit >15 and tmin =min(ti − tj) > −10 ns. The parameter tmin isthe smallest of all arrival time differences of hitchannels on each hit string. Positive and neg-ative values of tmin relate to upward and down-ward propagation of a light signal in the detector,respectively.

The experimental event distributions in the(tmin, Nhit)-parameter space are consistent withthe background expectation. No statistically sig-nificant excess over the background expectationfrom atmospheric muons has been observed.

Looking for events outside the area popu-lated by background events in the (tmin, Nhit)-parameter space we can derive upper limits onthe fluxes of high energy neutrinos which are pre-dicted by different models of neutrino sources.The detection volume Veff for neutrino producedevents was calculated as a function of neutrinoenergy and zenith angle θ. Veff rises from 2·105

m3 for 10 TeV up to 6·106 m3 for 104 TeV andsignificantly exceeds the geometrical volume Vg ≈105 m3 of NT-200 (Fig. 7).

Given an E−2 behaviour of the neutrino spec-trum and a flavor ratio νe : νµ : ντ = 1 : 1 : 1, the

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342338

90% C.L. upper limit obtained with the Baikalneutrino telescope NT-200 (780 days) is:

Φ(νe+νµ+ντ )E2 < 1.0×10−6cm−2s−1sr−1GeV.(1)

The model independent limit on ν̃e at the W -resonance energy is:

Φν̃e ≤ 4.2 × 10−20cm−2s−1sr−1GeV−1. (2)

Fig. 8 shows our upper limits on (νe + νµ +ντ ) diffuse fluxes from AGNs shaped accordingto the model of Stecker and Salamon (SS) [16], ofSemikoz and Sigl (SeSi) [17] and on E−2 spectrumaccording to Nellen et al. (NMB) [18] as well asthe model independent limit on the resonant ν̄e

flux (diamond).Also shown are the limits obtained by

AMANDA and MACRO experiments [19,20],theoretical bounds obtained by Berezinsky (B)[21], by Waxman and Bahcall (WB) [22], byMannheim et al.(MPR) [23], predictions for neu-trino fluxes from topological defects (TD) [17] andfrom GRB (WBGRB) [24].

10-10

10-9

10-8

10-7

10-6

10-5

10-4

3 4 5 6 7 8 9 10 11

νe:νµ:ντ=1:1:1

SeSi

TD

NMB

WBMPR

B

SS

WB(GRB)

lg(E/GeV)

E2 Φ

(E),

GeV

cm

-2 s

-1 s

r-1

MPR

atm.ν

νprνpr

BAIKAL(E-2,SS,SeSi) (res. νe)

♦AMANDA

MACRO

Figure 8. Experimental upper limits on the neu-trino fluxes as well as flux predictions in differentmodels of neutrino sources (see text).

6. A Search for Prompt Muons and Neu-trinos

Atmospheric neutrinos and muons are the mostimportant source of background for high-energyneutrino telescopes.

Figure 9. Experimental upper limits on theprompt muons (BAIKAL µ) and neutrinos(BAIKAL ν) obtained by the Baikal experimentas well as the experimental bounds on the promptmuons from LVD [27] and AKENO [28]. Alsoshown is a compilation of prompt muon fluxesat sea level [26]. Muon prompt fluxes are fromtwo charm production models in [29] (ZHVaand ZHVe); the empirical model in [30] (RVS);the quark-gluon string model and recombinationquark-parton model in [31] (QGSM, RQPM); per-turbative QCD in [32] (PRS); in [33,34] (GGV),and [35] (TIG).

At GeV energies they dominate by “conven-tional” muon and neutrino fluxes from decaysof relatively long-lived π and K mesons. Withincreasing energy, the probability increases that

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342 339

such particles interact in the atmosphere beforedecaying. This implies that even a small fractionof short-lived particles can give the dominant con-tribution to high energy atmospheric muon andneutrino fluxes at PeV energies. These “prompt”muons and neutrinos arise through semileptonicdecays of hadrons containing heavy quarks, mostnotably charm, and have significantly flatter spec-tra compared to conventional fluxes.

Our a search strategy for prompt muons andneutrinos is the same as for high energy neu-trinos which was described in the previous sec-tion. With the given shape of the energy spec-tra of prompt muons and neutrinos at sea leveland an expected number of events 2.44 at 90%C.L., upper limits on prompt muons and neu-trinos have been obtained. Fig. 9 shows ourupper limit for prompt muons with E−2.6 be-haviour of the energy spectrum at sea level aswell as limit for prompt νe with the spectrumsuggested by Volkova et al. [25]. Also a compi-lation of model dependent predictions on promptmuon fluxes [26] is shown in Fig. 9.

Figure 10. Sketch of NT-200+.

7. NT-200+ and Beyond

Recently derived upper limits on νe fluxes byBAIKAL and AMANDA are about E2Φ(ν) ≈(3 ÷ 5)10−7 cm−2 s−1 sr−1 GeV and cover theregion of optimistic theoretical predictions. How-ever, a flux sensitivity at the level of E2Φ(ν) <10−7 cm−2 s−1 sr−1 GeV which would test a va-riety of other models, requires detection volumesof order of 10 Mtons.

We envisage an upgrade of NT-200 to thisscale by three sparsely instrumented distant outerstrings. The basic principle will be the search forcascades produced in a large volume below NT-200. This configuration, christened NT-200+,will not only result in an increased detection vol-ume for cascades, but also allow for a precise re-construction of cascade vertex and energy withinthe volume spanned by the outer strings.

Figure 11. Detection volume of NT-200+ for νe

and νµ events which survive all cuts.

A schematic view of NT-200+ is shown in Fig.10. A water volume of 4.4 · 106 m3 is surroundedby the outer strings and NT-200.

The detection volumes for isotropic νe and νµ

fluxes are shown in Fig. 11. Most of the expectedevents would be produced by neutrinos from the

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342340

7

Figure 12. Reconstructed vs. simulated coordi-nates of cascades in NT-200+ (rectangles) andNT-200 (crosses).

energy range Eν > 102 TeV. In Fig. 12, recon-structed vs. simulated coordinates of cascades inNT-200+ (rectangles) and NT-200 (crosses) areshown. The reconstruction accuracy significantlyimproves in the case of NT-200+.

Assuming γ = 2 and a flavor ratio νe : νµ :ντ =1:1:1, a 90% C.L. limit on the νe flux of

Φ(νe+ν̃e)E2 < 9 · 10−8cm−2s−1sr−1GeV (3)

could be established from three years recordeddata .

MC simulations have shown that the detectionvolume of NT-200+ for PeV cascades would varyonly moderately, if NT-200 as the central part ofNT-200+ is replaced by a single string of OMs.For neutrino energies higher than 100 TeV, such aconfiguration could be used as a basic sub-array ofa Gigaton Volume Detector (GVD). Rough esti-mations show that 0.7 ÷ 0.9 Gton detection vol-ume for neutrino induced high energy cascadesmay be achieved with about 1300 OMs arrangedat 91 strings. A top view of GVD as well as sketch

Figure 13. Top view of GVD as well as sketch ofone of its sub-arrays.

of one basic sub-array are shown in Fig. 8 (rightpanel). The physical capabilities of GVD at veryhigh energies cover the typical spectrum of cu-bic kilometer arrays. We are presently workingon simulations to optimize the response for TeVmuons, maintaining at the same time the cubickilometer scale for cascades with energy above100 TeV.

8. Conclusions and Outlook

The deep underwater neutrino telescope NT-200 in Lake Baikal is taking data since April 1998.Using the first 502 live days, 84 neutrino inducedupward going muons have been selected. Limitson the diffuse high energy fluxes as well as onthe ν̄e flux at the W-resonance energy have beenderived. Also limits on an excess of the muon flux

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342 341

due to WIMP annihilation in the center of theEarth and on the flux of fast magnetic monopoleshave been obtained.

In 2005 we plan to put in operation the 10 Mtondetector NT-200+ with a sensitivity of approxi-mately 10−7cm−2s−1sr−1GeV for a diffuse neu-trino flux within the energy range E >102 TeV.NT-200+ will search for neutrinos from AGNs,GRBs and other extraterrestrial sources, neutri-nos from cosmic ray interactions in the Galaxyas well as high energy atmospheric muons withEµ > 10 TeV. In parallel to this short term goal,we started research & development activities to-wards a Gigaton Volume Detector in Lake Baikal.

This work was supported by the Russian Min-istry of Education and Science, the German Min-istry of Education and Research and the RussianFund of Basic Research ( grants 04-02-31006,02-02-17031 and 02-07-90293, Grant of Presidentof Russia NSh-1828.2003.2 and by the RussianFederal Program “Integration” (project no. 248).

REFERENCES

1. I. Belolaptikov et al.: Astr.Ph. 7, 263 (1997);2. R.I.Bagduev et al.,: Nucl. Instr. Meth. A420,

138 (1999).3. I. Belolaptikov et al.: Astr.Ph. 12, 75 (1999);4. V. Balkanov et al.: Nucl.Ph.(Proc.Sup.) B91,

438 (2001)5. V. Agrawal, T. Gaisser, P. Lipari & T.

Stanev: Phys. Rev. D53 1314 (1996)6. M.M.Boliev et al.: Nucl. Phys. (Proc.

Suppl.) 48 83 (1996); O. Suvorova. //hep-ph/9911415.

7. M.Ambrosio et al.: Phys. Rev, D60, 082002(1999),

8. A. Habig et al.: Proc.of the XVII ICRC, 1558(2001)

9. J. Ahrens et al.: Ph. Rev. D 66, 032006(2002)

10. V. Balkanov et al.: Nucl.Ph.(Proc.Sup.) B91,438 (2003)

11. M. Ambrosio et al.: Eur.Ph.J. C25, 511(2002)

12. J. Thorn et al.: Phys. Rev.D46 4846 (1992)13. H. Adarkar et al.: Proc. 21st ICRC. Adelaide.

95 (1990)14. S. Orito et al.: Phys. Rev. Lett. 66 1951

(1992)15. E. Andres et al.: Proc. of the 8th Int. Work-

shop on neutrino telescopes, 63 (1999) (astro-ph/9906255)

16. F. Stecker and M. Salamon, Space Sci. Rev.,1996, V. 75, P. 341-355

17. D.Semikoz and G.Sigl.: e-preprint hep-ph/0309328, 2003

18. L.Nellen, K.Mannheim and P.Biermann.:Phys. Rev., 1993, V. D47, P. 5270-5274

19. M. Kowalski et al.: Proc.of the XVIII ICRC,1301-1304 (2003)

20. M.Ambrosio et al.: Nucl. Phys (Proc. Sup.)B110 519 (2002);

21. V. Berezinsky et al.,: Astrophysics of CosmicRays, North Holland (1990)

22. E. Waxman and J. Bahcall: Phys. Rev. D59023002 (1999)

23. K.Mannheim et al.: astro-ph/9812398 (1998).24. E.Waxman and J.Bahcall.: Phys. Rev. Lett.,

1997, V. 78, P. 2292-2295.25. L.V. Volkova and G.T. Zatsepin.: Phys. Lett.,

1999, V. B462, P. 211-216.26. G.Gelmini, P.Gondolo, and G.Varieschi.:

Phys. Rev., 2003, V. D67, 017301.27. M. Aglietta et al.: Phys. Rev., 1999, V. D60,

112001.28. M. Nagano et al.: J. Phys.,1986, V. G12, P.

69.29. E.Zas, F.Halzen, and R.Vazquez.: Astropart.

Phys., 1993, V. 1, P. 297.30. O.Ryazhskaya, L.Volkova, and O.Saavedra.:

Nucl. Phys. B (proc. Suppl.), 2002, V. 110,P. 531.

31. E.Bugaev, V.Naumov, S.Sinegovsky, andE.Zaslavskaya.: Nuovo Cimento, 1989, V.C12, P.41.

32. L.Pasquali, M.Reno, and I.Sarcevic.: Phys.Rev., 1999, V. D59, 034020.

33. G.Gelmini, P.Gondolo, and G.Varieschi.:Phys. Rev., 2000, V. D61, 030605.

34. G.Gelmini, P.Gondolo, and G.Varieschi.:Phys. Rev., 2000, V. D61, 050611.

35. M.Thunman, G.Ingelman, and P.Gondolo.:Astropart. Phys., 1996, V. 5, P. 309.

V. Aynutdinov et al. / Nuclear Physics B (Proc. Suppl.) 143 (2005) 335–342342