Документ взят из кэша поисковой машины. Адрес оригинального документа : http://nuclphys.sinp.msu.ru/news/kamland/KamLAND_PRL21802.pdf Дата изменения: Wed Sep 17 10:09:39 2003 Дата индексирования: Tue Oct 2 00:37:18 2012 Кодировка: Поисковые слова: m 5

VOLU M E 90, N U M BER 2

PH YSICA L R E VI E W L E T T E RS

week end i ng 17 JA N UARY 20 03

Fi rst Results f rom Ka m L A N D: Ev idence for React or A nt i neut r i no Disappea ra nce
K. Eg uch i ,1 S. Enomot o ,1 K. Fu r uno,1 J. Gold ma n ,1 H. Ha nada ,1 H. I ke da ,1 K. I ke da ,1 K. I noue ,1 K. Ish i ha ra ,1 W. It oh ,1 T. Iwa mot o ,1 T. Kawag uch i ,1 T. Kawa sh i ma ,1 H. K i nosh it a ,1 Y. K ish i mot o ,1 M. Koga ,1 Y. Kosek i ,1 T. Mae da ,1 T. Mit sui,1 M. Motok i ,1 K. Na kaji ma ,1 M. Na kaji ma ,1 T. Na kaji ma ,1 H. Ogawa ,1 K. Owada ,1 T. Sa kabe ,1 I. Sh i m i z u,1 J. Sh i ra i ,1 F. Sueka ne ,1 A. Suz u k i ,1 K. Tada ,1 O. Taji ma ,1 T. Ta kaya ma ,1 K. Ta mae ,1 H. Wat a nabe ,1 J. Busen it z ,2 Z. Djurcic, 2 K. McK i n ny,2 D.- M. Mei ,2 A. P iepke ,2 E. Ya k ushev,2 B. E. Berger,3 Y. D. C ha n ,3 M. P. Decowsk i ,3 D. A. Dw yer,3 S. J. Free d ma n ,3 Y. Fu,3 B. K. Fuji kawa ,3 K. M. Heeger,3 K. T. L esko,3 K.- B. Lu k ,3 H. Muraya ma ,3 D. R. Nyg ren ,3 C. E. Okada ,3 A. W. P. Poon ,3 H. M. St ei ner,3 L. A. Wi n slow,3 G. A. Hor t on -Sm it h , 4 R. D. McKeown , 4 J. R it t er, 4 B. Tipt on ,4 P. Vogel ,4 C. E. L a ne ,5 T. Mi let ic, 5 P. W. Gorha m ,6 G. G ui l l ia n ,6 J. G. L ea r ne d ,6 J. Ma r icic,6 S. Mat suno,6 S. Pa k va sa ,6 S. Da zeley,7 S. Hat a keya ma ,7 M. Mu ra ka m i ,7 R. C. Svoboda ,7 B. D. Diet erle ,8 M. Di Mauro ,8 J. Det wi ler,9 G. G rat t a ,9 K. Ish i i ,9 N. Tol ich ,9 Y. Uch ida ,9 M. Bat ygov,10 W. Bugg,10 H. Coh n ,10 Y. E f remen ko,10 Y. Ka mysh kov,10 A. Koz lov,10 Y. Na ka mu ra ,10 L. De Brae ckeleer,11 C. R. Gould ,11 H. J. Ka r wowsk i ,11 D. M. Ma rkoff ,11 J. A. Messi more ,11 K. Na ka mu ra ,11 R. M. Roh m ,11 W. Tor now,11 A. R. Young,11 a nd Y.- F. Wa ng12 ( Ka m L A N D Col laborat ion)
1 Resea rch Cent er for Neut r i no Science , Tohok u Un iversit y, Senda i 980-8578, Japa n Depa r t ment of Physics a nd A st ronomy, Un iversit y of A laba ma , Tusca lo osa , A laba ma 35487 Physics Depa r t ment , Un iversit y of Ca l i for n ia at Berkeley a nd L aw rence Berkeley Nat iona l Laborat or y, Berkeley, Ca l i for n ia 9 4720 4 W. K. Kel logg Rad iat ion L aborat or y, Ca l i for n ia I n stit ut e of Te ch nolog y, Pasadena , Ca l i for n ia 91125 5 Physics Depa r t ment , Drexel Un iversit y, Ph i ladelph ia , Pen n sylva n ia 1910 4 6 Depa r t ment of Physics a nd Ast ronomy, Un iversit y of Hawa i i at Ma noa , Honolulu, Hawa i i 96822 7 Depa r t ment of Physics a nd A st ronomy, Louisia na St at e Un iversit y, Bat on Rouge , Louisia na 70803 8 Physics Depa r t ment , Un iversit y of New Mexico , A lbuquerque , New Mexico 87131 9 Physics Depa r t ment , St a n ford Un iversit y, St a n ford , Ca l i for n ia 94305 10 Depa r t ment of Physics a nd A st ronomy, Un iversit y of Ten nessee , K noxv i l le , Ten nessee 37996 11 Tr ia ngle Un iversit ies Nuclea r Laborator y, Durha m , Nor t h Ca rol i na 27708 a nd Physics Depa r t ment s at Duke Un iversit y, Nor t h Ca rol i na St at e Un iversit y, a nd t he Un iversit y of Nor t h Ca rol i na at Chap el H i l l 12 I n stit ut e of Hig h Energ y Physics, Beiji ng 10 0 039, People's Republ ic of C h i na ( Re ceived 6 De cember 20 02; publ ishe d 17 Ja nua r y 20 03) 2

3

Ka m LA N D ha s mea sure d t he flux of e 's f rom d ist a nt nuclea r react ors. We find fewer e event s t ha n expe ct e d f rom st a nda rd a ssumpt ion s about e propagat ion at t he 9 9.95% C. L. I n a 162 ton yr exp osure t he rat io of t he obser ved i nverse -decay event s t o t he exp e ct ed nu mber w it hout e d isapp ea ra nce is 0:611 0:085stat 0:041syst for e energies > 3:4 MeV. I n t he cont ext of t wo-flavor neut r i no osci l lat ion s wit h CP T i nva r ia nce , a l l solut ion s t o t he sola r neut r i no problem except for t he ``la rge m i xi ng a ngle'' reg ion a re exclude d.
DOI: 10.1103/ PhysRevL et t.90. 021802 PACS nu mbers: 14.60. Pq , 26.65.+t , 28. 50. Hw, 91.65. Dt

T he pr i ma r y goa l of t he Ka m ioka Liquid Sci nt i l lat or A nt i - Neut r i no Det e ct or ( Ka m L A N D) [1] is a sea rch for t he osci l lat ion of e 's em it t e d f rom d is t a nt p ower react ors. T he long ba sel i ne , t y pica l ly 180 k m , enables Ka m L A N D t o add ress t he osci l lat ion solut ion [2 ,3] of t he ``sola r neut r i no problem'' [4] wit h e's under laborat or y cond it ion s. T he i nverse -decay react ion , e p ! e n is used t o det e ct e 's i n l iquid sci nti l lat or ( LS) [5]. Det e ct i ng bot h t he e a nd t he de laye d 2. 2 MeV -ray f rom neut ron capt u re on a prot on is a p ower f ul t o ol for re duci ng backg round. T h is L et t er present s first result s f rom a n a na lysis of 162 ton yr (145.1 d) of t he react or e dat a. 021802 -1 0 031-90 0 7= 03 =9 0 (2)= 021802(6)$20. 0 0

Ka m L A N D o ccupies t he sit e of t he ea rl ier Ka m io ka nde [6], under 2700 m.w. e. of rock result i ng i n 0.34 Hz of cosm ic- ray muon s i n t he det e ct or. Shown i n Fig. 1, t he neut r i no det e ct or/t a rget is 1 k t on of ult rapu re LS cont a i ne d i n a 13- m - d ia met er spher ica l ba l lo on made of 135-m-thick t ra n spa rent nylon / EVOH (et hylene v i nyl a lcohol cop oly mer) comp osit e fil m. A net work of Kevla r ropes suppor t s a nd con st ra i n s t he ba l loon. T he LS is 80% dode ca ne , 20% pseudocu mene (1,2 , 4 -t r i met hylben zene) , and 1:52 g=liter of PPO (2 ,5- d iphenyloxa zole) a s a fluor. A bu ffer of dode ca ne a nd isopa ra ffin oi ls bet ween t he ba l loon a nd a n 18- m - d ia met er spher ica l sta i n less-st eel cont a i n ment vessel sh ields t he LS f rom ext er na l rad iat ion. 2003 T he A mer ica n Physica l Societ y 021802-1


VOLU M E 90, N U M BER 2

PH YSICA L R E VI E W L E T T E RS
0.04 0.02 E/E 0 0.02 0.04 0 2000 1 2 3 4 5 6

week end i ng 17 JA N UARY 20 03

(a)

7

8

9

Energy (MeV) (b)
balloon radius fiducial 0.2 0.4 0.6 0.8 1 1.2 1.4

Events/Bin

1600 1200 800 400 0 0

F IG. 1.

Schemat ic d iag ra m of t he Ka m L A N D det e ct or.

Dur i ng fill i ng wat er ext ract ion a nd n it rogen st r ippi ng [7] pu r ifie d t he LS a nd bu ffer oi l ( BO). T he bu ffer oi l den sit y is 0. 0 4% lower t ha n t he LS. A 1879 phot omult ipl ier t ube a r ray, mount e d on t he cont a i n ment vessel , complet es t he i n ner det e ct or ( I D). T here a re 1325 newly develop e d fa st 17- i nch - d ia met er phot omult ipl ier t ubes ( PM Ts) a nd 554 older Ka m ioka nde 20- i nch PM Ts [8]. T he t ot a l phot o cat hode coverage is 34% but on ly t he 17- i nch PMTs wit h 22% coverage a re used for t he present a na lysis. A 3- m m t h ick acr ylic ba r r ier at 16.6 - m d ia met er re duces radon f rom PM T gla ss i n t he LS. T he cont a i n ment vessel is sur rounde d by a 3. 2 k t on wat er-Cheren kov det e ct or wit h 225 20- i nch PMTs. T h is out er det e ct or (OD) absorbs rays a nd neut ron s f rom sur roundi ng rock a nd act s a s a t ag for cosm ic- ray muon s. T he pr i ma r y I D t r igger t h reshold is 200 PM T h it s, cor resp ondi ng t o about 0.7 MeV. T he t h reshold go es t o 120 h it s for 1 m sec a f t er a pr i ma r y t r igger. T he OD t r igger t h reshold cor resp onds t o > 99% t agg i ng efficiency. Energ y resp on se i n t he 0. 5 t o 7. 5 MeV ra nge is ca l i brat e d w it h 68 Ge, 65 Zn, 60 Co, a nd A m - Be -ray sources deploye d at va r ious posit ion s a long t he ver t ica l a x is. Det e ct e d energ y is obt a i ne d f rom t he number of obser ve d phot oele ct ron s ( p. e.) a f t er cor re ct ion s for ga i n va r iat ion , sol id a ngle , den sit y of PMTs, shadow i ng by suspen sion rop es, a nd t ra n spa rencies of t he LS a nd BO. Fig u re 2 (a) shows t he f ract iona l dev iat ion of t he re con st r uct e d energ ies f rom t he k nown source energies. T he 68 Ge and 60 Co sou rces em it t wo coi ncident rays a nd a re plot t e d at a n average energ y pn Fig. 2 (a). T he obser ved energ y resolui t ion is 7:5%= EMeV. T he energ y sca le is aug ment e d f rom st udies of t he rad iat ion f rom 40 K and 208 Tl a nd Bi - Po cont a m i na nt s, a s wel l a s 12 B- a nd 12 N-spallation pro duct s, a nd rays f rom neut ron capt u re on prot ons a nd 12 C. T he re con st r uct e d energ y va r ies by less t ha n 0. 5% wit h i n a 10 - m d ia met er volume except for 1.6% va r iat ion s nea r t he ch i m ney. T he energ y sca le is st able t o 0.6% dur i ng t he r un. Cor re ct ion s for quench i ng a nd C heren kov l ig ht pro duct ion a re i nclude d , a nd cont r ibut e t o t he systemat ic er ror i n Fig. 2 (a). T he est i mat e d systemat ic er ror i n t he 021802 - 2

(R/6.5m)3

F IG. 2. (a) T he f ract iona l d i fference of t he re con st r uct e d average energies a nd k nown sou rce energ ies. T he da she d l i nes show t he adopt e d systemat ic er ror. ( b) T he R3 ver t ex dist r ibution of 2. 2 MeV neut ron capt u re 's. T he level of un i for m it y over t he fiducia l volu me is used i n t he est i mat e of the fiducia l volume uncer t a i nt y.

energ y is 1.9% at our 2.6 MeV a na lysis t h reshold g iv i ng a 2.1% uncer t a i nt y i n t he rat e above t h reshold. Event lo cat ion s a re re con st r uct e d f rom t he t i m i ng of PMT h it s. A f t er energ y-dep endent rad ia l adjust ment s, t he k nown source p osition s a re re con st r uct e d t o 5 cm; t he t y pica l posit ion resolut ion is 25 cm. Ver t ex re con st r uct ion p er for ma nce t h roug hout t he LS volume is ver ifie d by repro duci ng t he un i for m d ist r ibut ion of 2.2 MeV capt u re 's f rom spa l lat ion neut rons, a s shown i n Fig. 2 ( b). T he dat a present e d i n t h is L et t er were col le ct e d f rom Ma rch 4 t h roug h Oct ober 6 , 2002. We obt a i ne d 370 106 event s i n 145.1 d of l ive t i me at a n average t r igger rat e of ' 30 Hz. Event s w it h less t ha n 10 0 0 0 p. e. (30 MeV) a nd no prompt OD t ag a re ``reactor- e ca ndidat es''; more energet ic event s a re ``muon ca ndidat es.'' T he sele ct ion cut s for e event s a re t he fol lowi ng: (i) fiducia l volu me (R< 5 m), (i i) t i me cor rela t ion (0:5 sec < T < 660 sec), (i i i) ver t ex cor re lat ion (R< 1:6 m), (iv) delaye d energ y (1:8 MeV

VOLU M E 90, N U M BER 2

PH YSICA L R E VI E W L E T T E RS

week end i ng 17 JA N UARY 20 03

event s wit h Eprompt < 2:49 MeV. Model Ia i n Ref. [9] pre d ict s about 9 geo event s i n our dat a set. However, t he abunda nces a nd d ist r ibut ions of U a nd T h a re not wel l k nown. We employ (v i) a prompt energ y cut , Eprompt > 2:6 MeV, t o avoid a mbig uit y i n t he present a na lysis. The fiducia l volu me is est i mat e d usi ng t he expe ct e d un i for m d ist r ibut ion of spa l lat ion -pro duct neut ron capt u re event s show n i n Fig. 2 ( b). T he rat io of event s i n the fiducia l volu me t o t he t ota l volu me ag rees wit h t he geomet r ic fiducia l f ract ion t o wit h i n 4.1%. T h is met hod is a lso used for h ig her energ y event s f rom 12 N, 12 B 's fol lowi ng muon spa l lat ion ; t he ag reement is wit h i n 3. 5%. Account i ng for t he 2.1% uncer t a i nt y i n t he t ot a l LS ma ss, we est i mat e a 4.6% uncer t a i nt y i n t he fiducia l volume. T he LS den sit y is 0:780 g=cm3 at 11:5 C; t he expe ct e d hyd rogen -t o - ca r bon rat io of 1.97 was ver ified by element a l a na lysis t o 2%. T he spe cific g rav it y is mea sure d t o 0. 01% pre cision a nd we a ssign a n add it iona l 0.1% er ror f rom t he uncer t a i nt y i n t he t emp erat u re. T he 408 t on fiducia l ma ss t hus cont a i n s 3:46 1031 f ree t a rget prot ons. T he t r igger efficiency wa s det er m i ne d t o be 9 9.98% w it h L E D l ig ht sources. T he combi ne d efficiency of t he ele ct ron ics, dat a acquisit ion , a nd event re con st r uct ion wa s st ud ie d usi ng t i me d ist r ibut ion s of uncor relat e d event s f rom ca l ibrat ion sources. We find t hat t h is com bi ne d efficiency is bet t er t ha n 99:98%. T he ver t ex fitt er y ields > 99:9% efficiency wit h i n 2 m of k nown source p osit ion s. Wit h ca l ibrat e d 60 Co and 65 Zn sources, t he overa l l efficiency wa s che cke d t o t he 3% source -st reng t h uncer t a i nt ies. T he det e ct ion efficiency for delaye d event s f rom t he A m - Be source (4. 4 MeV prompt a nd 2. 2 MeV delaye d neut ron capt u re wit h i n 1.6 m) wa s ver ified to 1% cer ta int y. St udies of Bi - Po se quent ia l de cays i ndicat e t hat t he effe ct ive e qui l ibr iu m concent rat ion s of 238 U and 232 Th in t he LS a re 3:5 0:5 10Ъ18 g=g and 5:2 0:8 10Ъ17 g=g, resp e ct ively. T he obser ve d back g round energ y spe ct r u m i ndicat es t hat 40 K cont a m i nat ion is less t ha n 2:7 10Ъ16 g=g. T he ext remely low level of U a nd T h cont a m i nat ion i n t he LS prov ides a n opt i m ist ic prospe ct for f ut u re sola r neut r i no exper i ment s w it h Ka m L A N D. T he flat accident a l backg round, obser ved i n a delaye d t i me w i ndow of 0. 020 ­ 20 se c, is 0:0086 0:0005 event s for t he present dat a set. T he most ser ious source of ext er na l rays f rom 208 Tl (E 3 MeV) is st rongly suppressed by t he fiducia l volume cut (i). At h ig her energ ies, t he backg round is dom i nat e d by spa l lat ion pro duct s f rom energet ic muon s. We obser ve 3 000 neutron events=day=kton. We a lso expe ct 1 300 events=day=kton [10] for va r ious unst able pro duct s. Single neut ron s a re ea si ly suppressed wit h a 2 - mse c veto fol low i ng a muon. Ca re is re qui re d t o avoid neut rons wh ich m i m ic t he e delaye d coi ncidence sig na l. Most 021802 -3

ext er na l fa st neut ron s a re pro duce d by muon s wh ich pa ss t h roug h bot h t he OD a nd t he sur roundi ng rock. T h is backg round is st udie d by det e ct i ng delaye d coi nci dence event s t agge d wit h a muon det e ct e d by on ly t he OD. As expe ct e d , event s concent rat e nea r t he ba l lo on e dge. T he backg round i n t he fiducia l volu me is est i mat e d by ext rap olat i ng t he d ist r ibut ion of ver t ex posit ion s a nd account i ng for t he 92% OD re con st r uct ion efficiency. T he number of backg round event s due t o neut rons f rom t he sur roundi ng rock is est i mat e d f rom t he OD-tagge d dat a sca le d by t he relat ive neut ron product ion a nd t he sh ield i ng fact or of t he releva nt mat er ia ls. T he est i mat e d t ota l fa st neut ron backg round is less t ha n 0. 5 event s i n t he ent i re dat a set. Most rad ioact ive spa l lat ion pro duct s si mply bet a de cay, a nd a re effe ct ively suppressed by re qui r i ng a de laye d neut ron. Delaye d neut ron em it t ers such a s 8 He (T1=2 119 msec)a nd 9 Li (178 msec) a re el i m i nat e d by t wo t i me/geomet r y cut s: (a) a 2 -se c vet o i n t he ent i re fiducia l volume fol low i ng a ``shower i ng muon'' (more tha n 106 p:e:, 3 GeV, ext ra energ y dep osit ion) , ( b) for ot her muon s, delaye d event s wit h i n 2 se c a nd 3 m of t he muon t rack a re reje ct e d. T he cut efficiency is est i mat e d f rom t he obser ve d cor relat ion of spa l lat ion neut rons wit h muon t rack s. T he rema i n i ng 8 He and 9 Li backg round is est i mat e d t o be 0:94 0:85. T he dead t i me f rom spa l la t ion cut s is 11. 4%. T h is is che cke d by con st r uct i ng t he t i me d ist r ibution of t he event s fol lowi ng a det e ct e d muon t o sepa rat e t he shor t - l ive d spa l lat ion -pro duce d act iv it ies f rom e ca ndidat es. T he uncor relat e d e d ist r ibut ion ha s a t i me con st a nt of 1=R ' 3 se c, where R is t he muon rat e. Spa l lat ion product s a l l have a shor t er t i me con sta nt ( 0.2 sec). T he t wo sele ct e d met hods ag ree t o 3% accuracy. As show n i n Table I , t he t ot a l number of expe ct e d backg round event s is 1 1, where t he fa st neut ron con t r ibut ion is i nclude d i n t he er ror est i mat e. I n st a nt a neous t her ma l p ower, bu r nup, a nd f uel excha nge re cords for a l l com mercia l Japa nese p ower react ors a re prov ide d by t he p ower compa n ies. T he t her ma l p ower generat ion is che cke d wit h t he i ndep endent re cords of ele ct r ic p ower generat ion. T he fission rat e for each fissile isotope is ca lculat e d a s a f unct ion of t i me a nd t he systemat ic uncer t a i nt y i n t he e flux is 1%. Average d over l ive t i me , t he relat ive fission y ields f rom f uel comp onent s are 235 U: 238 U: 239 Pu: 241 Pu 0:568 : 0:078 : 0:297 : 0:057. T he e spect r um per fission w it h a 2. 5% er ror

TA BL E I. Backg round Accident a l 9 Li=8 He Fa st neut ron Tot a l B.G. event s

Backg round sum ma r y. Number of events 0:0086 0:0005 0:94 0:85 <0:5 11

021802-3


VOLU M E 90, N U M BER 2
TA BL E I I.

PH YSICA L R E VI E W L E T T E RS

week end i ng 17 JA N UARY 20 03

Est i mat e d systemat ic uncer t a i nt ies (%). 2.1 4.1 2.1 2.1 0. 07 React or power F uel comp osition Ti me lag spe ct ra [11] C ross sect ion [14] 2. 1. 0. 2. 0. 0 0 28 5 2

Tot a l LS ma ss Fiducia l ma ss rat io Energ y t h reshold Efficiency of cut s Live t i me Tot a l systemat ic er ror

6. 4%

T he obser ve d space -t i me cor relat ion of t he prompt a nd delaye d event s ag rees w it h expe ct at ion s, a nd t he mea sure d capt u re t i me of 188 23 sec is con sist ent wit h pre dict ion s for LS. A f t er apply i ng a l l t he prompt a nd delaye d energ y cut s, 54 event s rema i n. Account i ng for 1 backg round event t he probabi l it y of a fluct uat ion f rom 86.8 expe ct e d is <0:05% by Poisson stat ist ics. T he rat io of obser ve d react or e event s t o expe ct e d i n t he absence of neut r i no d isapp ea ra nce is Nobs Ъ NBG 0:611 0:085stat 0:041syst: Nexpected Fig u re 4 shows t he rat io of mea sure d t o expe ct e d flux for Ka m L A N D a s wel l a s prev ious react or exper i ment s a s a f unct ion of t he average d ist a nce f rom t he source. T he expe ct e d prompt p osit ron spe ct r u m w it h no osci l lat ion s a nd t he best fit wit h re duce d 2 0:31 for 8 deg rees of f ree dom for t wo-flavor neut r i no osci l la t ion s above t he 2.6 MeV t h reshold a re show n i n Fig. 5. A clea r deficit of event s is ev ident. At t he 93% C. L. t he dat a a re con sist ent wit h a d ist or t e d spe ct r u m shap e expe ct e d f rom neut r i no osci l lat ions, but a sca le d noosci l lat ion shap e is a lso con sist ent at 53% C. L. a s det erm i ne d by Monte Ca rlo. T he neut r i no osci l lat ion pa ra met er reg ion for t woneut r i no m i x i ng is shown i n Fig. 6. T he da rk shade d a rea is t he MSW- LM A [19] reg ion at 95% C. L. der ive d f rom [16]. T he shade d reg ion out side t he solid l i ne is exclude d at 95% C. L. f rom t he rat e a na lysis wit h 2 3:84 and
1.4 1.2 1.0 Nobs/Nexp 0.8 0.6 0.4 0.2 0.0 10
1

a re t a ken f rom [11]. T h is neut r i no spe ct r u m ha s been t est e d t o a few p ercent wit h shor t -ba sel i ne react or e exper i ment s [5,12]. T he fin it e l i fet i mes of fission pro d uct s i nt ro duce a 0.28% uncer t a i nt y t o t he e flux. T he cont r ibut ion f rom Korea n react ors is est i mat e d t o be 2:46 0:25% ba sed on rep or t e d ele ct r ic p ower gen erat ion. T he rest of t he World's react ors cont r ibut e 0:70 0:35% f rom a n est i mat e usi ng react or spe cifica t ion s f rom t he I nt er nat iona l Nuclea r Sa fet y Cent er [13]. I n t he absence of e d isapp ea ra nce t he expe ct e d number of e event s is 86:8 5:6; t he systemat ic er ror cont r ibu t ion s a re l ist e d i n Table I I. T he a nti neut r i nos at Ka m L A N D a re prov ide d by ma ny nuclea r react ors but t he flux is act ua l ly dom i nat e d by a few p ower f ul react ors at a n average d ist a nce of 180 km. More t ha n 79% of t he flux is f rom 26 react ors bet ween 138­214 km away. One close react or at 88 k m cont r ibut es 6.7%; ot her react ors a re more t ha n 295 k m away. T he relat ively na r row ba nd of d ist a nces a l lows Ka m L A N D t o be sensit ive t o spe ct ra l d ist or t ion s for cer t a i n osci l lat ion pa ra met ers. Fig u re 3 shows t he energ y d ist r ibut ion of delaye d co i ncidence event s wit h no energ y cut s. A wel l -sepa rat e d clust er of 2.2 MeV capt u re 's is ev ident. One obser ve d event w it h delaye d energ y a round 5 MeV a nd prompt energ y of about 3.1 MeV (not shown i n Fig. 3) is con sist ent w it h t he expe ct e d neut ron rad iat ive capt u re rat e on 12 C.

ILL Savannah River Bugey Rovno Goesgen Krasnoyarsk Palo Verde Chooz

KamLAND 10 10 10 Distance to Reactor (m)
2 3 4

10

5

F IG. 3 (color). Dist r ibut ion of e ca nd idat es a f t er fiducia l volume , t i me , ver t ex cor relat ion , a nd spa l lat ion cut s a re ap pl ie d. For e event s t he prompt energ y is at t r ibute d t o p osit ron s a nd t he delaye d energ y t o neut ron capt u re. Event s wit h i n t he hor i zont a l l i nes bracket i ng t he delaye d energ y of 2. 2 MeV a re con sist ent wit h t her ma l neut ron capt ure on prot on s.

F IG. 4 (color). T he rat io of mea sure d t o exp e ct e d e flux f rom react or exp er i ment s [15]. T he sol id ci rcle is t he Ka m LA N D result plot t e d at a flux -weig ht e d average d ist a nce of 180 km. T he shade d region i nd icat es t he ra nge of flux pre dict ion s cor respondi ng t o t he 95% C. L. LM A reg ion f rom a globa l a na lysis of t he sola r neut r i no dat a [16]. T he dot t e d cu r ve , sin2 2 0:833 and m2 5:5 10Ъ5 eV2 [16], is rep resent at ive of a best -fit LM A pre d ict ion a nd t he da she d cu r ve is exp e ct e d for no osci l lat ion s.

021802 - 4

021802- 4


VOLU M E 90, N U M BER 2
20 15 10 Events/0.425 MeV

PH YSICA L R E VI E W L E T T E RS
reactor neutrinos geo neutrinos background

week end i ng 17 JA N UARY 20 03

10

-3

m 2 (eV )

2

5 0 25 20 15 10 5 0 0 2 4 Prompt Energy (MeV) 6 8
2.6 MeV analysis threshold

10

-4

KamLAND data no oscillation best-fit oscillation
sin 2 = 1.0 2 -5 2 m = 6.9 x 10 eV
2

10

-5

10

-6

Rate excluded Rate+Shape allowed LMA Palo Verde excluded Chooz excluded

0

0.2

0.4

sin 22

0.6

0.8

1

F IG. 5 (color). Upper pa nel: E xp e ct e d react or e energ y spe ct r u m a long wit h geo (mo del Ia of [9]) a nd backg round. L ower pa nel: Energ y spe ct r u m of t he obser ved prompt event s (sol id ci rcles wit h er ror ba rs), a long wit h t he exp e ct e d no osci l lat ion spe ct r u m (upp er h ist og ra m , wit h geo a nd back g round shown) a nd best fit ( lower h ist og ra m) i nclud i ng neut r i no osci l lat ion s. T he shade d ba nd i nd icat es t he systemat ic er ror i n t he best -fit spe ct r u m. T he ver t ica l da she d l i ne cor re sponds t o t he a na lysis t h reshold at 2.6 MeV.

F IG. 6 (color). E xclude d region s of neut r i no osci l lat ion pa ra met ers for t he rat e a na lysis a nd a l lowed region s for t he combi ne d rat e a nd shape a na lysis f rom Ka m LA N D at 95% C. L. At t he t op a re t he 95% C. L. exclude d region f rom CHOOZ [17] a nd Pa lo Verde [18] exper i ment s, respe ct ively. T he 95% C. L. a l lowe d reg ion of t he la rge m i x i ng a ngle ( LM A) solut ion of sola r neut r i no exper i ment s [16] is a lso shown. T he sol id ci rcle shows t he best fit t o t he Ka m L A N D dat a i n t he physica l reg ion : sin2 2 1:0 and m2 6:9 10Ъ5 eV2 . A l l region s lo ok ident ica l under $ =2 Ъ except for t he LM A reg ion f rom sola r neut r i no exper i ment s.

2

0:611 Ъ Rsin2 2; m2 2 : 0:0852 0:0412

Here , Rsin2 2; m2 is t he expe ct e d rat io wit h t he osci l lat ion pa ra met ers. The fina l event sa mple is eva luat e d using a ma x i mu m l i kel i hoo d met hod t o obt a i n t he opt i mu m set of osci l la t ion pa ra met ers w it h t he fol lowi ng 2 defin it ion : 2 2 sin2 2; m2 ;NBG12 ; rate
14


14

Ъ 2 logLshape sin2 2; m2 ;NBG12 ; 2 NBG12 2 BG distortion
14



:

Lshape is t he l i kel i hoo d f unct ion for t he spe ct r u m i nclud i ng exper i ment a l d ist or t ion s. NBG12 a re t he est i mat e d 9 Li and 8 He backg rounds a nd 14 a re pa ra met ers t o account for t he spe ct ra l effe ct s of energ y sca le uncerta int y, fin it e resolut ion , e spe ct r u m uncer t a i nt y, a nd fiducia l volu me systemat ic er ror, resp e ct ively. Para met ers a re va r ie d t o m i n i m i ze t he 2 at each pa i r of m2 ; sin2 wit h a bound f rom 2 NBG12 and BG 2 distortion 14 . T he best fit t o t he dat a i n t he physica l region yields sin2 2 1:0 and m2 6:9 10Ъ5 eV2 wh i le t he globa l m i n i mu m o ccu rs slig ht ly out side of t he physica l reg ion at sin2 2 1:01 wit h t he sa me m2 . T hese numbers ca n be compa re d t o t he best fit LM A 021802 - 5

va lues of sin2 2 0:83 and m2 5:5 10Ъ5 eV2 f rom [16]. T he 95% C. L. a l lowe d reg ion s f rom t he spe ct r u m shape a na lysis for 2 5:99 a nd t wo pa ra met ers a re shown i n Fig. 6. T he a l lowe d reg ions d isplaye d for Ka m L A N D cor resp ond t o 0 < < con sist ent wit h 4 t he sola r LM A solut ion , wh i le t he a l lowe d reg ion s i n 4 << 2 a re t he sa me [20] but do not i nclude t he sola r solution. T he result s f rom a spe ct ra l shap e a na lysis wit h a 0.9 MeV t h reshold a re con sist ent w it h t he above result. I n t h is low-energ y a na lysis, t he mea sure d geo fluxes a re f ree pa ra met ers. T he numbers of geo event s for t he best fit a re four for 238 U and five for 232 Th, wh ich cor resp onds t o 40 TW rad iogen ic heat generat ion accord i ng t o mo del Ia i n [9]. However, for t he sa me model , geo pro duct ion p owers f rom 0 t o 110 T W a re sti l l a l lowe d at 95% C. L. wit h t he sa me osci l lat ion pa ra met ers. I f t h ree neut r i no generat ion s a re con sidere d , t he e sur v iva l probabi l it y dep ends on t wo m i x i ng a ngles 12 and 13 . I n t he reg ion close t o t he best fit Ka m L AND solution t he sur v iva l probabi l it y is, t o a ver y goo d ap proxi mat ion , g iven by 2 4 2 2 m12 L P e ! e cos 13 1 Ъ sin 212 sin ; 4E wit h m
2 12

m2 f rom t he t wo-flavor a na lysis above. 021802- 5


VOLU M E 90, N U M BER 2

PH YSICA L R E VI E W L E T T E RS

week end i ng 17 JA N UARY 20 03

T he CHOOZ exper i ment [17] est abl ishe d a n upper l i m it of sin2 213 < 0:15, or cos4 13 0:92. O ur best fit result cor resp onds approx i mat ely t o 0:86 < sin2 212 < 1:0. I n sum ma r y, Ka m L A N D demon st rat e d react or e d isapp ea ra nce at long ba sel i nes a nd h igh confidence level (99:95%) for t he first t i me. One expe ct s a negl ig ible re duct ion of e flux f rom t he SM A , LOW a nd VAC , sola r neut r i no solut ion s, a nd t he LM A reg ion is t he on ly rema i n i ng osci l lat ion solution con sist ent wit h t he Ka m L A N D result a ssum i ng C P T i nva r ia nce. T he a l lowe d LM A reg ion is con st ra i ne d by Ka m L A N D. Fut u re Ka m L A N D mea surement s wit h i mprove d st at ist ica l pre cision a nd re duce d systemat ic er rors wi l l prov ide pre ci sion det er m i nat ion s of neut r i no osci l lat ion pa ra met ers. T he Ka m L A N D exper i ment is suppor t e d by t he Cent er of E xcel lence prog ra m of t he Japa nese Mi n ist r y of E ducat ion , Cult u re , Spor t s, Science a nd Te ch nolog y, a nd f und i ng f rom t he Un it e d St at es Depa r t ment of Energ y. T he react or dat a a re prov ide d cou r t esy of t he fol lowi ng ele ct r ic a ssociat ion s i n Japa n ; Hok ka ido , Tohok u, Tok yo, Hok u r i k u, C hubu, Ka n sa i , C hugok u, Sh i kok u a nd Ky ushu Ele ct r ic Power Compa n ies, Japa n At om ic Power Co. , a nd Japa n Nuclea r C ycle Develop ment I n st it ut e. Ka m ioka Mi n i ng a nd Smelt i ng Compa ny prov ide d ser v ices for act iv it ies i n t he m i ne.

[1] Ka m L A N D Col laborat ion , A. Suz u k i et a l. , i n P rocee di ngs of t he 7t h Eu rom icro Workshop on Pa ral lel a nd Dist r ibut e d P rocessi ng, F uncha l , Por t uga l , 19 9 9 [ Nucl. Phys. ( P roc. Suppl.) B77, 171 (19 9 9)]; ht t p: //w ww.awa.t ohok u.ac.jp/ Ka m L A N D/.

[2] Z. Ma k i et a l. , P rog. T heor. Phys. 28, 870 (1962). [3] B. Ponte cor vo , Sov. Phys. J ET P 6, 429 (1957) [Z h. E k sp. Teor. Fi z. 33, 549 (1957)]. [4] J. N. Ba hca l l a nd R. Davis, Science 191, 26 4 (1976); J. N. Ba hca l l , Neut r i no A st rophysics (Ca mbr idge Un iversit y P ress, Ca mbr idge , Un it e d K i ngdom , 1989); J. N. Ba hca l l , Ast rophys. J. 46 7, 475 ( 19 96). [5] F. Rei nes et a l. , Phys. Rev. 117, 159 (1960); for a review, see C. Bemp orad et a l. , Rev. Mod. Phys. 74 , 297 (20 02). [6] K. S. Hi rat a et a l. , Phys. Rev. D 38, 4 48 (1988). [7] J. B. Ben z iger et a l. , Nucl. I nst r um. Met ho ds Phys. Res. , Se ct. A 417, 278 (1998). [8] H. Kume et a l.. , Nucl. I n st r u m. Met ho ds Phys. Res. 205, 4 43 (1986). [9] R. S. Rag hava n et a l. , Phys. Rev. Let t. 80, 635 (1998). [10] T. Hagner et a l.. , A st ropa r t. Phys. 14 , 33 (200 0). [11] 235 U: K. Sch re ckenbach et a l. , Phys. L et t. B 160, 325 (1985); 239 241 Pu: A. A. Ha h n et a l., Phys. L et t. B 218, 365 (1989); 238 U: P. Vogel et a l. , Phys. Rev. C 24 , 1543 (1981). [12] B. Ach ka r et a l., Phys. L et t. B 374, 2 43 (19 96); G. Z acek et a l., Phys. Rev. D 34 , 2621 (1986). [13] ht t p: //w ww. i n sc.a n l.gov/ [14] P. Vogel a nd J. F. Beacom , Phys. Rev. D 60, 0530 03 (19 9 9); rad iat ive cor re ct ion f rom A. Kur ylov et a l. , hep -ph /0211306. [15] Pa r t icle Dat a G roup, Phys. Rev. D 66 , 010001- 406 (20 02). [16] G. L. Fogl i et a l.. , Phys. Rev. D 66 , 053010 (20 02). [17] M. Apol lon io et a l. , Phys. L et t. B 466, 415 (19 9 9). [18] F. Bo eh m et a l. , Phys. Rev. D 64, 112001 (20 01). [19] L. Wol fen st ei n , Phys. Rev. D 17, 2369 (1979); S. P. Mi k heyev a nd A. Yu. Sm i r nov, Sov. J. Nucl. Phys. 42 , 913 (1986). [20] A. de Gouvea et a l. , Phys. L et t. B 49 0, 125 (20 0 0).

021802 -6

021802-6