1: %%

2: %% Beginning of file 'ms.tex'

3: %%

4: %% Modified 2007 May 3

5: %%

6: %% This is a manuscript marked up using the

7: %% AASTeX v5.x LaTeX 2e macros.

8: 

9: \documentclass[12pt,preprint]{aastex}

10: %% manuscript produces a one-column, double-spaced document:

11: %% \documentclass[manuscript]{aastex}

12: %% preprint2 produces a double-column, single-spaced document:

13: %% \documentclass[preprint2]{aastex}

14: 

15: %% Sometimes a paper's abstract is too long to fit on the

16: %% title page in preprint2 mode. When that is the case,

17: %% use the longabstract style option.

18: %% \documentclass[preprint2,longabstract]{aastex}

19: 

20: %% \usepackage{mlineno}

21: 

22: %% You can insert a short comment on the title page using the command below.

23: %% \slugcomment{Draft version 2007 May 8, 7.1}

24: 

25: 

26: \shorttitle{The Search for Muon Neutrinos from GRBs with AMANDA}

27: \shortauthors{The IceCube Collaboration}

28: 

29: 

30: \begin{document}

31: 

32: 

33: \title{The Search for Muon Neutrinos from Northern Hemisphere Gamma-Ray Bursts with AMANDA}

34: 

35: \author{

36: A.~Achterberg\altaffilmark{1},

37: M.~Ackermann\altaffilmark{2},

38: J.~Adams\altaffilmark{3},

39: J.~Ahrens\altaffilmark{4},

40: K.~Andeen\altaffilmark{5},

41: %% D.~W.~Atlee\altaffilmark{6},

42: J.~Auffenberg\altaffilmark{14},

43: J.~N.~Bahcall\altaffilmark{7,34},

44: X.~Bai\altaffilmark{8},

45: B.~Baret\altaffilmark{9},

46: %% M.~Bartelt\altaffilmark{10},

47: S.~W.~Barwick\altaffilmark{11},

48: R.~Bay\altaffilmark{12},

49: K.~Beattie\altaffilmark{13},

50: T.~Becka\altaffilmark{4},

51: J.~K.~Becker\altaffilmark{10},

52: K.-H.~Becker\altaffilmark{14},

53: P.~Berghaus\altaffilmark{15},

54: D.~Berley\altaffilmark{16},

55: E.~Bernardini\altaffilmark{2},

56: D.~Bertrand\altaffilmark{15},

57: D.~Z.~Besson\altaffilmark{17},

58: E.~Blaufuss\altaffilmark{16},

59: D.~J.~Boersma\altaffilmark{5},

60: C.~Bohm\altaffilmark{18},

61: J.~Bolmont\altaffilmark{2},

62: S.~B\"oser\altaffilmark{2},

63: O.~Botner\altaffilmark{19},

64: A.~Bouchta\altaffilmark{19},

65: J.~Braun\altaffilmark{5},

66: C.~Burgess\altaffilmark{18},

67: T.~Burgess\altaffilmark{18},

68: T.~Castermans\altaffilmark{20},

69: D.~Chirkin\altaffilmark{13},

70: B.~Christy\altaffilmark{16},

71: J.~Clem\altaffilmark{8},

72: D.~F.~Cowen\altaffilmark{6,21},

73: M.~V.~D'Agostino\altaffilmark{12},

74: A.~Davour\altaffilmark{19},

75: C.~T.~Day\altaffilmark{13},

76: C.~De~Clercq\altaffilmark{9},

77: L.~Demir\"ors\altaffilmark{8},

78: F.~Descamps\altaffilmark{22},

79: P.~Desiati\altaffilmark{5},

80: T.~DeYoung\altaffilmark{6},

81: J.~C.~Diaz-Velez\altaffilmark{5},

82: J.~Dreyer\altaffilmark{10},

83: J.~P.~Dumm\altaffilmark{5},

84: M.~R.~Duvoort\altaffilmark{1},

85: W.~R.~Edwards\altaffilmark{13},

86: R.~Ehrlich\altaffilmark{16},

87: J.~Eisch\altaffilmark{23},

88: R.~W.~Ellsworth\altaffilmark{16},

89: P.~A.~Evenson\altaffilmark{8},

90: O.~Fadiran\altaffilmark{24},

91: A.~R.~Fazely\altaffilmark{25},

92: %% T.~Feser\altaffilmark{4},

93: K.~Filimonov\altaffilmark{12},

94: M.~M.~Foerster\altaffilmark{6},

95: B.~D.~Fox\altaffilmark{6},

96: A.~Franckowiak\altaffilmark{14}

97: T.~K.~Gaisser\altaffilmark{8},

98: J.~Gallagher\altaffilmark{26},

99: R.~Ganugapati\altaffilmark{5},

100: H.~Geenen\altaffilmark{14},

101: L.~Gerhardt\altaffilmark{11},

102: A.~Goldschmidt\altaffilmark{13},

103: J.~A.~Goodman\altaffilmark{16},

104: R.~Gozzini\altaffilmark{4},

105: T.~Griesel\altaffilmark{4},

106: A.~Gro{\ss}\altaffilmark{27},

107: S.~Grullon\altaffilmark{5},

108: R.~M.~Gunasingha\altaffilmark{25},

109: M.~Gurtner\altaffilmark{14},

110: A.~Hallgren\altaffilmark{19},

111: F.~Halzen\altaffilmark{5},

112: K.~Han\altaffilmark{3},

113: K.~Hanson\altaffilmark{5},

114: D.~Hardtke\altaffilmark{12},

115: R.~Hardtke\altaffilmark{23},

116: %% T.~Harenberg\altaffilmark{14},

117: J.~E.~Hart\altaffilmark{6},

118: Y.~Hasegawa\altaffilmark{30},

119: T.~Hauschildt\altaffilmark{8},

120: D.~Hays\altaffilmark{13},

121: J.~Heise\altaffilmark{1},

122: K.~Helbing\altaffilmark{14},

123: M.~Hellwig\altaffilmark{4},

124: P.~Herquet\altaffilmark{20},

125: G.~C.~Hill\altaffilmark{5},

126: J.~Hodges\altaffilmark{5},

127: K.~D.~Hoffman\altaffilmark{16},

128: B.~Hommez\altaffilmark{22},

129: K.~Hoshina\altaffilmark{5},

130: D.~Hubert\altaffilmark{9},

131: B.~Hughey\altaffilmark{5},

132: P.~O.~Hulth\altaffilmark{18},

133: J.-P.~H\"ul{\ss}\altaffilmark{32},

134: K.~Hultqvist\altaffilmark{18},

135: S.~Hundertmark\altaffilmark{18},

136: M.~Inaba\altaffilmark{30},

137: A.~Ishihara\altaffilmark{5},

138: J.~Jacobsen\altaffilmark{13},

139: G.~S.~Japaridze\altaffilmark{24},

140: H.~Johansson\altaffilmark{18},

141: A.~Jones\altaffilmark{13},

142: J.~M.~Joseph\altaffilmark{13},

143: K.-H.~Kampert\altaffilmark{14},

144: A.~Kappes\altaffilmark{5},

145: T.~Karg\altaffilmark{14},

146: A.~Karle\altaffilmark{5},

147: H.~Kawai\altaffilmark{30},

148: J.~L.~Kelley\altaffilmark{5},

149: %% M.~Kestel\altaffilmark{6},

150: N.~Kitamura\altaffilmark{5},

151: S.~R.~Klein\altaffilmark{13},

152: S.~Klepser\altaffilmark{2},

153: G.~Kohnen\altaffilmark{20},

154: H.~Kolanoski\altaffilmark{28},

155: L.~K\"opke\altaffilmark{4},

156: M.~Kowalski\altaffilmark{28},

157: T.~Kowarik\altaffilmark{4},

158: M.~Krasberg\altaffilmark{5},

159: K.~Kuehn\altaffilmark{11},

160: M.~Labare\altaffilmark{9},

161: H.~Landsman\altaffilmark{5},

162: H.~Leich\altaffilmark{2},

163: D.~Leier\altaffilmark{10},

164: I.~Liubarsky\altaffilmark{29},

165: J.~Lundberg\altaffilmark{19},

166: J.~L\"unemann\altaffilmark{10},

167: J.~Madsen\altaffilmark{23},

168: K.~Mase\altaffilmark{30},

169: H.~S.~Matis\altaffilmark{13},

170: T.~McCauley\altaffilmark{13},

171: C.~P.~McParland\altaffilmark{13},

172: A.~Meli\altaffilmark{10},

173: T.~Messarius\altaffilmark{10},

174: P.~M\'esz\'aros\altaffilmark{6,21},

175: H.~Miyamoto\altaffilmark{30},

176: A.~Mokhtarani\altaffilmark{13},

177: T.~Montaruli\altaffilmark{5,35},

178: A.~Morey\altaffilmark{12},

179: R.~Morse\altaffilmark{5},

180: S.~M.~Movit\altaffilmark{21},

181: K.~M\"unich\altaffilmark{10},

182: R.~Nahnhauer\altaffilmark{2},

183: J.~W.~Nam\altaffilmark{11},

184: P.~Nie{\ss}en\altaffilmark{8},

185: D.~R.~Nygren\altaffilmark{13},

186: H.~\"Ogelman\altaffilmark{5},

187: %% Ph.~Olbrechts\altaffilmark{9},

188: A.~Olivas\altaffilmark{16},

189: S.~Patton\altaffilmark{13},

190: C.~Pe\~na-Garay\altaffilmark{7},

191: C.~P\'erez~de~los~Heros\altaffilmark{19},

192: A.~Piegsa\altaffilmark{4},

193: D.~Pieloth\altaffilmark{2},

194: A.~C.~Pohl\altaffilmark{19,36},

195: R.~Porrata\altaffilmark{12},

196: J.~Pretz\altaffilmark{16},

197: P.~B.~Price\altaffilmark{12},

198: G.~T.~Przybylski\altaffilmark{13},

199: K.~Rawlins\altaffilmark{31},

200: S.~Razzaque\altaffilmark{6,21},

201: %% F.~Refflinghaus\altaffilmark{10},

202: E.~Resconi\altaffilmark{27},

203: W.~Rhode\altaffilmark{10},

204: M.~Ribordy\altaffilmark{20},

205: A.~Rizzo\altaffilmark{9},

206: S.~Robbins\altaffilmark{14},

207: P.~Roth\altaffilmark{16},

208: C.~Rott\altaffilmark{6},

209: D.~Rutledge\altaffilmark{6},

210: D.~Ryckbosch\altaffilmark{22},

211: H.-G.~Sander\altaffilmark{4},

212: S.~Sarkar\altaffilmark{33},

213: S.~Schlenstedt\altaffilmark{2},

214: T.~Schmidt\altaffilmark{16},

215: D.~Schneider\altaffilmark{5},

216: D.~Seckel\altaffilmark{8},

217: B.~Semburg\altaffilmark{14},

218: S.~H.~Seo\altaffilmark{6},

219: S.~Seunarine\altaffilmark{3},

220: A.~Silvestri\altaffilmark{11},

221: A.~J.~Smith\altaffilmark{16},

222: M.~Solarz\altaffilmark{12},

223: C.~Song\altaffilmark{5},

224: J.~E.~Sopher\altaffilmark{13},

225: G.~M.~Spiczak\altaffilmark{23},

226: C.~Spiering\altaffilmark{2},

227: M.~Stamatikos\altaffilmark{5,38},

228: T.~Stanev\altaffilmark{8},

229: P.~Steffen\altaffilmark{2},

230: T.~Stezelberger\altaffilmark{13},

231: R.~G.~Stokstad\altaffilmark{13},

232: M.~C.~Stoufer\altaffilmark{13},

233: S.~Stoyanov\altaffilmark{8},

234: E.~A.~Strahler\altaffilmark{5},

235: T.~Straszheim\altaffilmark{16},

236: K.-H.~Sulanke\altaffilmark{2},

237: G.~W.~Sullivan\altaffilmark{16},

238: T.~J.~Sumner\altaffilmark{29},

239: I.~Taboada\altaffilmark{12},

240: O.~Tarasova\altaffilmark{2},

241: A.~Tepe\altaffilmark{14},

242: L.~Thollander\altaffilmark{18},

243: S.~Tilav\altaffilmark{8},

244: M.~Tluczykont\altaffilmark{2},

245: P.~A.~Toale\altaffilmark{6},

246: D.~Tur{\v{c}}an\altaffilmark{16},

247: N.~van~Eijndhoven\altaffilmark{1},

248: J.~Vandenbroucke\altaffilmark{12},

249: A.~Van~Overloop\altaffilmark{22},

250: V.~Viscomi\altaffilmark{6},

251: B.~Voigt\altaffilmark{2},

252: W.~Wagner\altaffilmark{10},

253: C.~Walck\altaffilmark{18},

254: H.~Waldmann\altaffilmark{2},

255: M.~Walter\altaffilmark{2},

256: Y.-R.~Wang\altaffilmark{5},

257: C.~Wendt\altaffilmark{5},

258: C.~H.~Wiebusch\altaffilmark{32},

259: G.~Wikstr\"om\altaffilmark{18},

260: D.~R.~Williams\altaffilmark{6},

261: R.~Wischnewski\altaffilmark{2},

262: H.~Wissing\altaffilmark{32},

263: K.~Woschnagg\altaffilmark{12},

264: X.~W.~Xu\altaffilmark{25},

265: G.~Yodh\altaffilmark{11},

266: S.~Yoshida\altaffilmark{30},

267: J.~D.~Zornoza\altaffilmark{5,37} (the IceCube Collaboration),

268: and 

269: M.~Boer\altaffilmark{39},

270: T.~Cline\altaffilmark{38},

271: G.~Crew\altaffilmark{40},

272: M.~Feroci\altaffilmark{41},

273: F.~Frontera\altaffilmark{42},

274: %% J.~Goldsten\altaffilmark{43},

275: %% S. Golenetskii\altaffilmark{46},

276: K.~Hurley\altaffilmark{43},

277: D.~Lamb\altaffilmark{44},

278: A.~Rau\altaffilmark{45},

279: F.~Rossi\altaffilmark{42},

280: G.~Ricker\altaffilmark{40},

281: A.~von~Kienlin\altaffilmark{46} (the InterPlanetary Network)

282: %% E. Mazets\altaffilmark{46},

283: }

284: 

285: \altaffiltext{1}{Dept.~of Physics and Astronomy, Utrecht University/SRON,

286:   NL-3584 CC Utrecht, The Netherlands}

287: \altaffiltext{2}{DESY, D-15735 Zeuthen, Germany}

288: \altaffiltext{3}{Dept.~of Physics and Astronomy, University of Canterbury,

289:   Private Bag 4800, Christchurch, New Zealand}

290: \altaffiltext{4}{Institute of Physics, University of Mainz, Staudinger

291: Weg 7,

292:   D-55099 Mainz, Germany}

293: \altaffiltext{5}{Dept.~of Physics, University of Wisconsin, Madison, WI

294: 53706,

295:   USA}

296: \altaffiltext{6}{Dept.~of Physics, Pennsylvania State University, University

297:   Park, PA 16802, USA}

298: \altaffiltext{7}{Institute for Advanced Study, Princeton, NJ 08540, USA}

299: \altaffiltext{8}{Bartol Research Institute, University of Delaware,

300: Newark, DE

301:   19716, USA}

302: \altaffiltext{9}{Vrije Universiteit Brussel, Dienst ELEM, B-1050 Brussels,

303:   Belgium}

304: \altaffiltext{10}{Dept.~of Physics, Universit\"at Dortmund, D-44221

305: Dortmund,

306:   Germany}

307: \altaffiltext{11}{Dept.~of Physics and Astronomy, University of California,

308:   Irvine, CA 92697, USA}

309: \altaffiltext{12}{Dept.~of Physics, University of California, Berkeley, CA

310:   94720, USA}

311: \altaffiltext{13}{Lawrence Berkeley National Laboratory, Berkeley, CA 94720,

312:   USA}

313: \altaffiltext{14}{Dept.~of Physics, University of Wuppertal, D-42119

314:   Wuppertal, Germany}

315: \altaffiltext{15}{Universit\'e Libre de Bruxelles, Science Faculty CP230,

316:   B-1050 Brussels, Belgium}

317: \altaffiltext{16}{Dept.~of Physics, University of Maryland, College Park, MD

318:   20742, USA}

319: \altaffiltext{17}{Dept.~of Physics and Astronomy, University of Kansas,

320:   Lawrence, KS 66045, USA}

321: \altaffiltext{18}{Dept.~of Physics, Stockholm University, SE-10691

322: Stockholm,

323:   Sweden}

324: \altaffiltext{19}{Division of High Energy Physics, Uppsala University, S-75121

325:   Uppsala, Sweden}

326: \altaffiltext{20}{University of Mons-Hainaut, 7000 Mons, Belgium}

327: \altaffiltext{21}{Dept.~of Astronomy and Astrophysics, Pennsylvania State

328:   University, University Park, PA 16802, USA}

329: \altaffiltext{22}{Dept.~of Subatomic and Radiation Physics, University of

330:   Gent, B-9000 Gent, Belgium}

331: \altaffiltext{23}{Dept.~of Physics, University of Wisconsin, River Falls, WI

332:   54022, USA}

333: \altaffiltext{24}{CTSPS, Clark-Atlanta University, Atlanta, GA 30314,

334:   USA}

335: \altaffiltext{25}{Dept.~of Physics, Southern University, Baton Rouge, LA

336:   70813, USA}

337: \altaffiltext{26}{Dept.~of Astronomy, University of Wisconsin, Madison, WI

338:   53706, USA}

339: \altaffiltext{27}{Max-Planck-Institut f\"ur Kernphysik, D-69177

340: Heidelberg, Germany}

341: \altaffiltext{28}{Institut f\"ur Physik, Humboldt Universit\"at zu

342: Berlin, D-12489 Berlin, Germany}

343: \altaffiltext{29}{Blackett Laboratory, Imperial College, London SW7 2BW, UK}

344: \altaffiltext{30}{Dept.~of Physics, Chiba University, Chiba 263-8522 Japan}

345: \altaffiltext{31}{Dept.~of Physics and Astronomy, University of Alaska

346:   Anchorage, 3211 Providence Dr., Anchorage, AK 99508, USA}

347: \altaffiltext{32}{III Physikalisches Institut, RWTH Aachen University,

348:   D-52074 Aachen, Germany}

349: \altaffiltext{33}{Dept.~of Physics, University of Oxford, 1 Keble Road,

350: Oxford

351:   OX1 3NP, UK}

352: \altaffiltext{34}{Deceased}

353: \altaffiltext{35}{on leave of absence from Universit\`a di Bari,

354: Dipartimento

355:      di Fisica, I-70126, Bari, Italy}

356: \altaffiltext{36}{affiliated with Dept.~of Chemistry and Biomedical

357: Sciences,

358:   Kalmar University, S-39182 Kalmar, Sweden}

359: \altaffiltext{37}{affiliated with IFIC (CSIC-Universitat de Val\`encia),

360:   A. C. 22085, 46071 Valencia, Spain}

361: \altaffiltext{38}{NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA}

362: \altaffiltext{39}{Observatoire de Haute Provence, F-04870 Saint Michel l'Observatoire, France}

363: \altaffiltext{40}{Massachusetts Institute of Technology, Kavli Institute for Astrophysics and Space Research, Cambridge, MA 02139, USA}

364: \altaffiltext{41}{Instituto di Astrofisica Spaziale e Fisica Cosmica, Instituto Nazionale di Astrofisica, c/o CNR Area di Ricerca di 

365: Roma - Tor Vergata, Via Fosso del Cavaliere 100, Roma, I-00133 Italy}

366: \altaffiltext{42}{Physics Department, University of Ferrara, Via Saragat 1, 44100 Ferrara, Italy}

367: %% \altaffiltext{43}{The Johns Hopkins University, Applied Physics Laboratory, Laurel, MD 20723 USA}

368: \altaffiltext{43}{University of California, Space Sciences Laboratory, 7 Gauss Way, Berkeley, CA 94720-7450 USA}

369: \altaffiltext{44}{Department of Astronomy and Astrophysics, University of Chicago, 5640 South Ellis Avenue Chicago, IL 60637 USA}

370: \altaffiltext{45}{Caltech Optical Observatories, California Institute of Technology, MS 105-24, 1200 E. California Blvd., Pasadena, CA 

371: 91125 USA}

372: \altaffiltext{46}{Max-Planck-Institut f\"ur Extraterrestrische Physik, Giessenbachstrasse, 85748, Garching, Germany}

373: 

374: \email{* Corresponding author K. Kuehn: kuehn@HEP.ps.uci.edu}

375: 

376: %% \linenumbers

377: 

378: \begin{abstract}

379: We present the results of the analysis of neutrino observations by the Antarctic Muon and 

380: Neutrino Detector Array (AMANDA) correlated with photon observations of more than 400 

381: gamma-ray bursts (GRBs) in the Northern Hemisphere from 1997 to 2003.  During this 

382: time period, AMANDA's effective collection area for muon neutrinos was larger than that of 

383: any other existing detector.  Based on our observations of zero neutrinos during and 

384: immediately prior to the GRBs in the dataset, we set the most stringent upper limit on muon 

385: neutrino emission correlated with gamma-ray bursts.  Assuming a Waxman-Bahcall spectrum and 

386: incorporating all systematic uncertainties, our flux upper limit has a normalization at 1 

387: PeV of

388: 

389: \noindent

390: $E^{2}$${\Phi}_{\nu}$ ${\leq}$ 6.0 $\times$ $10^{-9}$ $\mathrm{GeV cm^{-2} s^{-1} sr^{-1}}$,

391: 

392: \noindent

393: with 90\% of the events expected within the energy range of $\sim$10 TeV to $\sim$3 PeV.  

394: The impact of this limit on several theoretical models of GRBs is discussed, as well as the 

395: future potential for detection of GRBs by next generation neutrino telescopes.  Finally, we 

396: briefly describe several modifications to this analysis in order to apply it to other types 

397: of transient point sources.

398: \end{abstract}

399: 

400: \keywords{gamma-ray bursts, high energy astrophysics, neutrino astronomy, AMANDA}

401: 

402: \section{Introduction}

403: 

404: \subsection{Gamma-Ray Bursts}

405: 

406: Gamma-ray bursts (GRBs) are among the most energetic phenomena in the universe; based on 

407: their luminosity and the cosmological distances derived from redshift measurements of burst 

408: afterglows and/or host galaxies\citep{Beppo}, GRBs require the release of an enormous amount of 

409: energy (E $\approx$10$^{53}$ $\times$ $\Omega$/4$\pi$ erg, where $\Omega$ is the solid angle of 

410: the GRB jet) in as little as a fraction of a second\citep{Frail}.  Based on the observations of 

411: the Burst and Transient Source Experiment\citep[BATSE, see][]{BATSE} and other space-based 

412: detectors, they are expected to occur throughout the observable universe at a rate of 

413: ${\gtrsim}$700 per year, though current instrument do not have sufficient sky coverage or 

414: sensitivity to detect every burst.  Long duration (${\gtrsim}$2 sec) bursts are believed to 

415: originate from the collapse of a massive stellar progenitor into a black hole, whereas 

416: short duration (${\lesssim}$2 sec) bursts are believed to result from the merger of two 

417: compact objects into a black hole \citep{Eicher}\footnote{For a more recent treatment of 

418: the compact object merger scenario, see \citep{Pacz, Lewin}, and for an alternative 

419: description of the GRB progenitor scenario, see also \citep{Roming}.}.  Though these two 

420: types of bursts come from different progenitors, both are consistent with the canonical 

421: picture of gamma-ray bursts---the fireball scenario \citep{Fryer,Piran1}.  A fireball is 

422: generated during the formation of the black hole when the outflowing plasma is accelerated 

423: to ultrarelativistic speeds.  Subsequently, in an optically thin region (outside of the 

424: progenitor), the kinetic energy of the plasma is converted to radiation, either through 

425: interaction with an external medium or through self-interaction within the flow 

426: \citep{Piran2}.  If the circumstellar environment contains enough baryonic material, it 

427: will be entrained with the accelerated plasma.  Subsequent photo-pion production by baryon 

428: interaction with synchrotron or inverse Compton scattered photons will lead to several 

429: decay products, including neutrinos and antineutrinos in a ratio of 2:1.  The primary 

430: reaction is:

431: \begin{equation}

432: p + \gamma \rightarrow \Delta \rightarrow \pi^{+} + n

433: \end{equation}

434: followed by 

435: \begin{equation}

436: \pi^{+} \rightarrow \mu^{+} + \nu_{\mu} 

437: \end{equation}

438: after which the muon will further decay to

439: \begin{equation}

440: \mu^{+} \rightarrow e^{+} + \nu_{e} + \bar{\nu_{\mu}}.

441: \end{equation}

442: \noindent

443: Similarly, ``precursor'' neutrinos may be generated by p-p interactions either within the 

444: star or in the immediate circumburst environment (see Section 2).  Due to their minuscule 

445: interaction cross section \citep{Gandhi}, neutrinos will reach the AMANDA detector after 

446: traveling nearly unimpeded from the GRB environment.  AMANDA has been searching for 

447: high-energy neutrinos from various astrophysical fluxes (both discrete and diffuse) for 

448: nearly a decade; in this work we focus on the analysis of AMANDA data correlated with 

449: photon observations of more than 400 GRBs from 1997 to 2003.

450: 

451: \subsection{The AMANDA Detector}

452: 

453: The AMANDA detector \citep{AMANDA} is an array of Optical Modules (OMs) deployed at depths 

454: between 1.5 and 2 km beneath the surface of the ice at the South Pole.  An OM consists of a 

455: photomultiplier tube housed in a glass pressure sphere. During the years 1997-1999 the 

456: detector operated with 302 OMs on ten strings placed in a circular geometry with a diameter 

457: of about 100 m, and was known as AMANDA B-10.  From 2000 onward, nine additional strings 

458: were in operation, placed within a diameter of about 200 m, bringing the total number 

459: of optical modules to 677.  This phase of the neutrino observatory (dubbed AMANDA-II) 

460: operated through 2004, and continues as a high density component of IceCube, a km-scale 

461: detector currently being constructed \citep{IceCub}.

462: 

463: The optical modules in AMANDA are designed to detect the Cherenkov emission from 

464: neutrino-induced muons that travel through or near the instrumented volume of ice.  While 

465: other neutrinos may be detected with this search, the efficiency for ${\nu}_{e}$ or 

466: ${\nu}_{\tau}$ detection is significantly smaller.  Other multi-flavor GRB neutrino searches 

467: which don't require directional information have been performed \citep{Casc}; we focus 

468: here on the search for GRB muon neutrinos from the Northern Hemisphere ($\delta$ from 0$^{\circ}$ to 90$^{\circ}$).  Due to the limited 

469: volume of ice above the detector, few downgoing extraterrestrial neutrinos will interact 

470: above and be detected by AMANDA.  At the energies of interest to this analysis, the 

471: down-going events in the AMANDA dataset are primarily the atmospheric muon background which 

472: will completely overwhelm any potential downgoing signal.  Thus, our extraterrestrial 

473: signal is primarily confined to the horizontal or up-going direction.  As these muon 

474: neutrinos travel through the ice, they may interact with nearby nucleons to create 

475: energetic muons:

476: \begin{equation}

477: \nu_{\mu} + N \rightarrow \mu + X,

478: \end{equation}

479: 

480: \noindent

481: where N is a nucleon and X represents other reaction products.  Muons produced in this 

482: reaction can carry a significant fraction of the original neutrino energy

483: \citep{Gandhi2}.  Depending on its energy, the muon can travel up to tens 

484: of kilometers through the ice; for $\nu_{\mathrm{\mu}}$ in the energy range of greatest 

485: interest to AMANDA (${\sim}10^{5}$ GeV), the muon path length is ${\sim}$10 

486: km \citep{Lipari}.

487: 

488: Since AMANDA can detect such a muon anywhere along its substantial path length, the 

489: effective detector volume is significantly larger than the actual instrumented volume.  A 

490: muon that has sufficient energy will continuously emit Cherenkov radiation, and will also 

491: generate additional particles due to stochastic processes.  The ice at a depth of more than 

492: one kilometer is extremely clear, and thus the Cherenkov photons have large scattering 

493: ($L^{\mathrm{{eff}}}_{\mathrm{s}}$) and absorption ($L_{\mathrm{a}}$) lengths---at 

494: $\lambda$ = 400 nm, $L^{\mathrm{eff}}_{\mathrm{s}}$${\approx}$25 m and 

495: $L_{\mathrm{a}}$${\approx}$100 m \citep{Kurt}.  The Cherenkov light therefore has the 

496: potential to reach numerous OMs as the muon travels through the detector, and the relative 

497: timing of the hit OMs provides the basis for a set of maximum-likelihood reconstruction 

498: algorithms to determine the muon's direction of origin \citep{Reco}.  The algorithms 

499: applied to this analysis are based on variations from a randomly-seeded ``first guess'' 

500: track using the Pandel function to parametrize the sequence of OM hits.  The likelihood of 

501: the initial track is calculated, and then the procedure is iterated (up to 32 times) to 

502: determine the most likely muon track.  Iterations beyond the first incorporate increasingly 

503: complex features of the detector response to the Cherenkov photons, the details of which 

504: are beyond the scope of this work\footnote{An alternative track reconstruction known as a 

505: ``paraboloid fit'' is also relevant for our secondary data selection criteria, see 

506: Section 3.3 for further details.}.  Detector simulations, along with observations of 

507: downgoing cosmic ray muons, have shown that this procedure provides track reconstructions 

508: accurate to within a mean value of ${\sim}$2$^{\circ}$.  Atmospheric muons are almost 

509: entirely removed from the dataset by constraining our search to those bursts occurring in 

510: the Northern Hemisphere, allowing the detector to be shielded from a substantial background 

511: flux by the bulk of the earth.  Upgoing atmospheric neutrinos caused by cosmic ray 

512: interactions in the Northern hemisphere may also be detected by AMANDA, as their spectrum 

513: extends into the energy range of relevance to the GRB search.  However, they likewise are 

514: removed from the dataset by requiring strict spatial and temporal correlation with photon 

515: observations of GRBs.  With these selection criteria applied, we expect less than 0.01 

516: atmospheric neutrino events in our dataset.

517: 

518: In Section 2 we describe several models for GRB neutrino emission.  In Section 3 we discuss 

519: the method for determining periods of stable detector performance and for separating the 

520: expected GRB neutrino signal from all misreconstructed background events, as well 

521: as the systematic uncertainties associated with this analysis procedure.  In Section 4 we 

522: compare the results of the AMANDA observations with the models, as well as provide a 

523: spectrum-independent method for determining the fluence upper limit from GRBs.  We conclude 

524: with the future potential of AMANDA/IceCube, for both the standard GRB search in the Swift 

525: era \citep{Mark} and for searches optimized for other transient point sources, such as 

526: jet-driven supernovae.

527: 

528: \section{Models of Neutrino Emission}

529: 

530: According to the canonical description provided above, gamma-ray bursts result from the 

531: dissipation of the energy of relativistic outflows from a central engine.

532: Based on the assumption that GRBs are the source of ultra-high energy cosmic rays (UHECRs), 

533: Waxman and Bahcall predicted an annual muon neutrino flux associated with GRBs of 

534: $E^{2}$$\Phi_{\mathrm{\nu}}$ $\sim$ 9 $\times$ $10^{-9}$ $\mathrm{GeV cm^{-2} 

535: s^{-1} sr^{-1}}$ from 100 TeV to 10 PeV \citep{Waxman} 

536: \footnote{For the original formulation of this neutrino flux prediction, see 

537: \citep{WaxBah}.  Note that this GRB neutrino flux is distinct from the Waxman-Bahcall upper 

538: bound on the diffuse neutrino flux due to UHECRs.}.  \citet{Murase} predict a similar 

539: spectrum to Waxman and Bahcall for long-duration bursts, though their simulations include a 

540: wider range of parameters, leading to a wider variation in predicted fluxes.  Inclusion of 

541: neutrino oscillations reduce these predictions by a factor of two\footnote{Oscillations 

542: modify the flavor ratio from 1:2:0 at the source to 1:1:1 at Earth.  However, see 

543: \citep{Kashti} for a discussion regarding different flavor ratios due to energy losses 

544: of the ${\pi}$ and ${\nu}$.}.  \citet{Razzaq} hypothesize a different scenario in which a 

545: supernova precedes a long-duration GRB by several days to a week.  In this ``supranova'' 

546: scenario, the supernova remnant provides target nucleons for pp interactions leading to 

547: precursor neutrinos with energy $E_{\nu}$${\sim}$10 TeV. Furthermore, the remnant will 

548: produce target photons for p${\gamma}$ interactions, which will also yield muon neutrinos 

549: up to $10^{16}$ eV, albeit with a different spectral shape than that predicted by the 

550: Waxman-Bahcall model\footnote{Though the supranova model is still within the realm of 

551: possibility, it is somewhat disfavored based on observations of GRB060218, in which the 

552: supernova preceded the GRB by at most a few hours---not long enough to provide an ideal 

553: circumburst environment for a significant neutrino flux.}.  This model also has 

554: implications for gamma-ray dark (or ``choked'') bursts, which are briefly discussed in 

555: Section 5.  Furthermore, though these models explicitly incorporate only long-duration 

556: bursts into their models, the GRB central engine is in principle independent of burst type.  

557: Thus, though the flux upper limits for these models include long bursts only, the models 

558: could potentially be expanded to include neutrinos from short bursts as well.  Within the 

559: AMANDA dataset long bursts dominate over short bursts; incorporating short bursts would 

560: have a small, though not insignificant, effect on the overall limit (see Section 4 for 

561: details).  Figure 1 shows the expected GRB neutrino flux based on four representative 

562: models.  The precursor model predicts a neutrino flux as early as several tens of seconds 

563: prior to the observed GRB photons, whereas the other models tested here predict a neutrino 

564: flux in coincidence with the GRB photons\footnote{Any time delay in observing the neutrinos 

565: due to the neutrino mass is assumed to be negligible compared to the time scale over which 

566: we search for the burst emission.}.  Other models of GRB emission also exist 

567: \citep[see, e.g.][]{Dermer1,Dermer2}; though we do not explicitly focus on such models 

568: here, a flux upper limit can be calculated for such models using the Green's Function 

569: method detailed in Section 4.

570: 

571: Many theoretical models (most notably, the Waxman-Bahcall model) are based on assumptions 

572: regarding the circumburst environment as well as the average properties of bursts (total 

573: emission energy, redshift, etc.) which do not correspond directly to the properties of 

574: specific bursts.  It is possible to estimate the muon neutrino flux for individual bursts, 

575: but these estimates vary substantially, and often bracket the predictions of the averaged 

576: properties \citep{Mike}.  For those bursts where redshift and spectral information is 

577: available, more accurate estimates of muon neutrino flux can be made on a burst-by-burst 

578: basis.  For extremely bright, nearby bursts (e.g. GRB030329), the predicted fluxes can be 

579: as much as two orders of magnitude greater than the mean burst flux \citep{MikeSwift}.  

580: Finally, our simulations assume a $\mathrm{\Phi_{\nu}}$:$\mathrm{\Phi_{\bar{\nu}}}$ ratio of 

581: 1:1.  AMANDA does not distinguish the muon charge; however, neutrino event rates are 

582: larger than anti-neutrino rates for an equal flux, since the neutrino cross section is 

583: larger up to energies of ${\sim}$$10^5$ GeV.  Thus, any models proposing a ratio other than 

584: unity will result in a different expected event rate and, ultimately, a different flux upper 

585: limit for this analysis.  

586: 

587: \section{Observation Procedure}

588: 

589: \subsection{Correlated Observations}

590: 

591: This AMANDA GRB search relies on spatial and temporal correlations with photon observations 

592: of other instruments including BATSE aboard the Compton Gamma-Ray Observatory (CGRO), as 

593: well as HETE-II, Ulysses, and other satellites of the Third Interplanetary Network (IPN) 

594: \citep{Hurley}.  As stated previously, our search is restricted to that half of the bursts 

595: occurring in the Northern Hemisphere.  Furthermore, because engineering and maintenance 

596: work is performed on the AMANDA detector during the austral summer (December-February), 

597: only a few bursts from these months can potentially be observed each year.  For each GRB in 

598: the dataset, we search for muon neutrino emission during the coincident phase of burst 

599: emission.  The coincident phase is determined by either the $\mathrm{T_{90}}$ start and end 

600: times of the burst, or the entire duration of emission in excess of the background rate 

601: (for bursts without well-defined $\mathrm{T_{90}}$). A period of time before and after each 

602: burst is added to the search in order to accommodate the timing errors of the photon 

603: observations (which vary from burst to burst).  Most bursts have prompt phases lasting from 

604: a few seconds up to to a few tens of seconds, though there are some exceptional bursts 

605: lasting hundreds of seconds.  To investigate different model predictions for the bursts 

606: occurring during 2001-2003, we also performed an extended search for precursor neutrinos 

607: from 110 seconds before the burst start time until the beginning of the coincident search 

608: window.  BATSE observations were the sole source of data for the AMANDA B-10 analysis for 

609: 1997--1999.  Other IPN-detected bursts were included beginning with the AMANDA-II dataset 

610: in 2000, and the analysis then relied exclusively on IPN data from other satellites once 

611: CGRO was decommissioned in May, 2000.  Additional bursts were also discovered in the BATSE 

612: archival data \citep{Kommers,Stern}; the relevant time periods of the AMANDA data were 

613: searched for muon neutrinos from these bursts as well.  We do not, however, include this 

614: particular subset of bursts in the flux or fluence upper limits for the models addressed in 

615: this work, because non-triggered bursts were not incorporated into the primary models of 

616: GRB neutrino emission.   The instruments participating in the Interplanetary Network 

617: through 2003 are given in Table 1 and the number of bursts searched in each year of AMANDA 

618: observations is listed in Table 2; information on the specific bursts included in this 

619: analysis is also available \citep{myurl}.

620: 

621: \subsection{Background and Detector Stability}

622: 

623: To determine the background rate and to establish data selection criteria for each burst, a 

624: larger period of one hour and 50 minutes of data is analyzed---from one hour before the 

625: burst to one hour after the burst, with the 10 minute period during and immediately 

626: surrounding the burst excluded to ensure that the data quality cuts are not determined in a 

627: biased fashion (a ``blind'' analysis).  Prior to determination of the data selection 

628: criteria, we study detector stability in this background period.  The specific stability 

629: criteria for AMANDA B-10 have been discussed previously \citep{Rellen}; here we describe 

630: the AMANDA-II stability criteria in more detail.

631: 

632: We perform two tests to identify non-statistical fluctuations in the data rate that could 

633: produce fake events (``false positives'') or unanticipated dead time (``false negatives'') 

634: in the detector.  The first test compares the observed event count per 10 second time 

635: bin to the expected, temporally uncorrelated, distribution of background events.  

636: This tests for any non-statistical fluctuations in data rate due to temporary instability 

637: in the detector.  Without this test, an upward fluctuation in the data rate not caused by 

638: neutrinos could potentially be misinterpreted as a signal event.  This test has three 

639: successive steps based on the P-value of the event rate distribution. The P-value of a data 

640: segment is defined as the percent difference between the RMS variation of the data event 

641: rate and the width of a Gaussian fit to the data rate distribution.  The first step 

642: identifies all those bursts with stable periods---those having a P-value of less than 6$\%$ 

643: (corresponding to variations of less than 1${\sigma}$ relative to the overall distribution 

644: of P-values).  The second step identifies bursts with marginally stable detector 

645: performance: 6${\%}$${\leq}$P${\leq}$12$\%$ (1--2${\sigma}$).  Additional tests are 

646: performed on these bursts; specifically, the data rate of the previously blinded 10-minute 

647: period is explored in a region of the sky far away from the GRB (the ``on-time, 

648: off-source'' region).  This maintains the blindness of the analysis, while allowing a more 

649: detailed exploration of detector stability.  Marginally stable burst periods are included 

650: in the analysis if they are also marginally stable in the on-time, off-source region 

651: (P-value less than 12$\%$), and if the event rate has only small (${\leq}3{\sigma}$) 

652: variations throughout the on-time, off-source region.  The vast majority of all burst time 

653: periods were stable according to these criteria.  The final step of this test is applied if 

654: the first two steps are inconclusive.  It requires any event rate variations greater than 

655: 3$\sigma$ to occur at a significant distance from the burst time.  Two bursts fall into 

656: this category; they had marginally stable off-time periods and insufficient statistics for 

657: an on-time/off-source stability test.  However they were included in this analysis because 

658: the largest event rate variations were separated in time from the burst by several minutes.  

659: Only one time period associated with a burst in the AMANDA dataset had off-time and 

660: on-time/off-source P-values greater than 12\%, and this burst was excluded from the 

661: analysis.  Figure 2 shows the data rate per 10 seconds for a sample GRB period, overlaid 

662: with the Gaussian fit.  They are in very good agreement, showing a stable data rate for 

663: this period of detector activity.

664: %% Poisson, not just Gauss...

665: 

666: The second test utilizes the time between subsequent events ($\mathrm{{\delta}t}$) to 

667: ensure that there is not an anomalously large amount of time between detector triggers.  

668: The amount of time between triggers can vary widely, but larger gaps occur with much less 

669: frequency than shorter gaps.  There is also unavoidable (but quantifiable) dead time 

670: between each trigger while the detector is being read out.  The overall effect of the 

671: expected dead time is to reduce the detector's signal acceptance by approximately 17\%, and 

672: this quantity has been incorporated into the expected neutrino observation rate for this 

673: analysis.  However, large unexpected gaps between triggers would indicate a period of 

674: unstable detector performance, and would mean that an otherwise detectable neutrino signal 

675: might not be observed during such a period.  We test the 1 hour and 50 minute time periods 

676: surrounding each burst to ensure that no such gaps occur.  An example of the temporal 

677: distribution of triggers compared with an exponentially decreasing fit to the 

678: $\mathrm{{\delta}t}$ distribution is shown in Figure 3.  The variations observed in the 

679: data for this time period are within 2${\sigma}$ of the observed fit for all values of 

680: $\mathrm{{\delta}t}$.  Thus there are no unexpected variations in the time between detector 

681: triggers, and we confirm that AMANDA is collecting data as expected occurring during the 

682: on-time window for this burst.  All data periods associated with GRBs that pass the first 

683: test also pass this second test for stable detector operation.

684: 

685: \subsection{Data Selection Criteria}

686: 

687: For those bursts determined to be stable by the above criteria, data quality cuts are then 

688: selected to separate the predicted signal from the observed background events.  This 

689: process relies primarily on the simulated signal events and the observed background events.

690: The simulation of the detector response to signal and background events is described in 

691: Ahrens et al. \citep{Reco}.  The simulation procedure uses the neutrino generation 

692: program NuSim \citep{Hill97} for signal event simulations.  Background events are simulated 

693: with CORSIKA \citep{Cors}, which implements the 2001 version of the QGSJET model of 

694: hadronic interactions \citep{Kalmykov}.  Once the neutrino or other cosmic-ray primaries 

695: are generated and propagated to their interaction vertex, we simulate the secondary 

696: propagation with the Muon Monte Carlo (MMC) package \citep{Dima}.  Finally, we simulate the 

697: AMANDA detector response with the software package AMASIM.  We then are able to compare 

698: simulated signal, simulated background, and observed background data.

699: 

700: In the case of the GRB search, the background rate is measured using the off-time window, 

701: where no signal is expected.  Thus, unlike other AMANDA analyses \citep{Diffuse,Point}, the 

702: background events do not need to be simulated, nor do the data events need to be scrambled 

703: in time or azimuth to retain a blind analysis procedure.  Exploring the variations between 

704: observed background events and simulated events does, however, ensure that we understand 

705: the systematic errors associated with the simulation process.  For example, Figure 4 shows 

706: excellent agreement in ${\mathrm{{\pounds}_{reco}}}$, the log(Likelihood) of the 

707: reconstructed tracks of the simulated and observed background events.  Given this level of 

708: agreement, the errors arising from discrepancies between the simulated and observed events 

709: are expected to be small.  Additionally, atmospheric neutrinos have previously been 

710: observed by AMANDA up to TeV energies, and studies show that neutrinos from this proven 

711: source can be reconstructed with a high degree of accuracy\citep{Nature}.  Likewise, 

712: studies have been performed which compare simulated signal events with high-quality 

713: downgoing muon events \citep{Hodges}.  Because these downgoing events have similar 

714: properties to the simulated signal events, this provides additional assurance that the 

715: simulated signal events will have similar properties to the actual signal events we are 

716: attempting to observe.  Section 3.4 gives a quantitative discussion of systematic errors.

717:  

718: %% EQUATION!

719: To determine the set of data selection criteria that will produce the optimal flux upper 

720: limit in the absence of a signal, we minimize the Model Rejection Factor (MRF)

721: \citep{Hill03}.  The MRF is based on the expected detector sensitivity prior to observations: 

722: \begin{equation}

723: MRF = \frac{\bar{{\mu}}_{90}({\mathrm{N_{BG,Exp}}})}{{\mathrm{N_{Sig}}}}

724: \end{equation}

725: \noindent

726: where $\bar{{\mu}}_{90}$ is the Feldman-Cousins $90\%$ average event upper limit 

727: \citep{FeldCous} derived from the expected number of background events 

728: (N$_{\mathrm{BG,Exp}}$) and N$_{\mathrm{Sig}}$ is the expected number of 

729: signal events.  N$_{\mathrm{Sig}}$ is determined by convolving the theoretical spectrum 

730: (${\Phi}$ = dN$_{\mathrm{\nu}}$/dE) with the detector's energy- and angle-dependent 

731: effective neutrino collecting area ($A_{eff,{\nu}}$) and integrating over the angular 

732: acceptance of the detector, the energy range of interest ($10^{2}$ to $10^{7}$ GeV), 

733: and the observation time (assuming 700 bursts contribute equally to the annual expected 

734: flux):

735: \begin{equation}

736: N_{\mathrm{Sig}} = {\int}{\int}{\Phi}(E,{\theta},{\phi}) A_{\mathrm{eff,{\nu}}}(E,{\theta}) {\mathrm{dE d{\Omega} dt}}.

737: \end{equation}

738: \noindent

739: As an intermediate step in the determination of the expected number of signal events, we 

740: therefore need to determine the detector effective collection area.  

741: $A_{\mathrm{eff,{\nu}}}$ is determined by the fraction of simulated neutrino events that 

742: are retained after all data selection criteria are applied.  This area also accounts for 

743: neutrinos that generate muons passing nearby (but not through) the detector and still cause 

744: the telescope to trigger.

745: 

746: In determining the optimal data selection criteria for the coincident search, we assume a 

747: Waxman-Bahcall neutrino spectrum \citep{Waxman}; for the precursor search, we assume a 

748: Razzaque spectrum \citep{Razzaq}.  In addition to temporal coincidence described 

749: previously, the most relevant selection criterion for this analysis is the angular mismatch 

750: (${\Delta}$${\Psi}_{\mathrm{i}}$) between the burst position and the reconstructed event 

751: track.  This mismatch is determined for each of four separate maximum-likelihood pattern 

752: recognition algorithms (i = 1 to 4) applied to the timing of the hit OMs (as described in 

753: Section 2).  The different algorithms are based on different initial seeds and apply a 

754: different number of iterations to the track reconstruction procedure, thus they are able to 

755: provide different measures used for discrimination between expected signal and background 

756: events.  Though they are not completely independent, they do offer improvements to the MRF 

757: when applied consecutively.  The inherent difference in the muon and neutrino paths, as 

758: well as the inaccuracies of the reconstruction algorithms, prevent perfect characterization 

759: of all signal and background events.  Nevertheless, the angular mismatch is quite effective 

760: as a selection criterion. For example, selecting events with a mismatch angle 

761: ${\Delta}$${\Psi}_{1}$ of less than $12^{\mathrm{o}}$ retains more than $90\%$ of the 

762: expected signal events, while reducing the background to less than $0.5\%$ (Figure 5).  

763: Depending upon the changes in the detector characteristics and the analysis tools from year 

764: to year, the MRF optimization procedure allowed for some variation in the specific track 

765: reconstruction algorithms applied, as well as the mismatch angle values selected for each 

766: algorithm (see Table 3).

767: 

768: Several secondary criteria were also used to improve the separation between signal and 

769: background events.  Included in the secondary criteria is the measured number of hit 

770: channels---that is, the number of OMs participating in the reconstruction of each event.  

771: The number of direct hits---hits that occur within -15/+75 ns of the arrival time for light 

772: propagating from the reconstructed muon track to the OM in question---also serves as a 

773: useful criterion for data selection.  Direct hits should be due to photons that do 

774: not scatter, or scatter minimally; their straight trajectories give them a well-defined 

775: behavior, making them most useful in determining the muon direction.  Additionally, 

776: the likelihood of a given reconstruction and the angular resolution (${\sigma}_{\Psi}$) of 

777: the alternate event track reconstruction (the ``paraboloid fit'') provide a useful event 

778: discriminator, since high quality signal events will have higher likelihoods and superior 

779: angular resolution compared to the background events.  One additional criterion used in 

780: this analysis is the uniformity of the spatial distribution of the hit OMs---events with 

781: hit OMs spread evenly along the track are more likely to be single high-energy 

782: neutrino-induced muons, whereas events with hit OMs clustered in time and space along the 

783: track are more likely to be background events.  Different combinations of these criteria 

784: were applied in the 1997-1999, 2000, and 2001-2003 timeframes, as new analysis tools were 

785: developed and applied to the GRB neutrino search (see Table 3).

786: 

787: This analysis procedure was applied to bursts with localization errors from the satellite 

788: observations that are relatively small (typically less than 1$^{\circ}$) and therefore 

789: inconsequential on the scale of the AMANDA search bin radius.  However, several hundred IPN 

790: bursts have large localization errors (${\gtrsim}$1/2 of the search bin radius), but still 

791: lie completely within the field of view of AMANDA.  These were either marginal detections 

792: near the edge of BATSE's field of view or they were detected by only two IPN 

793: satellites, which prevents triangulation of their position but allows localization to an 

794: annular segment.  Eleven of the bursts in the AMANDA dataset are only poorly localized; the 

795: increased search area for these bursts results in a corresponding increase in the expected 

796: background rate.  To ensure that this increase does not diminish the overall sensitivity of 

797: the GRB search, more restrictive selection criteria are applied to these bursts.  Whether 

798: well localized or poorly localized, each burst has an associated background expected during 

799: the burst time, calculated from the event rate of the off-time background region multiplied 

800: by the duration of the time window during which we search for signal events.

801: 

802: The initial criteria were independently selected to optimize the MRF and were then collectively 

803: optimized in an iterative fashion.  The optimal criteria depended on the zenith angle of 

804: the burst, due to the higher observed background rate for bursts closer to the horizon.  

805: The criteria for higher background rates (i.e. low zenith angle bursts) were also applied 

806: to bursts with large satellite localization errors, regardless of the actual zenith angle 

807: of the burst.  Table 3 lists all data selection criteria used for the year-by-year GRB 

808: analyses, as well as the selection criteria for the precursor search applied in 2001-2003.  

809: Though the data selection criteria are optimized for specific models of neutrino emission,  

810: other models can also be tested using the Green's Function Fluence Limit Method (see 

811: Results).  While the muon track reconstruction algorithm is very accurate, there is a 

812: small probability that a downgoing muon will be misreconstructed in the upgoing direction; 

813: such events are the primary background for the GRB search.  After the application of data 

814: selection criteria, background events have an observed rate of ${\sim}5 {\times} 10^{-5} 

815: {\mathrm{Hz}}$ (with some seasonal variation).  

816: Figure 6 show the effective area for neutrinos for the AMANDA-II detector after all data 

817: selection criteria are applied.

818: Due to the large instrumented area and modest background rejection requirements of this 

819: analysis, AMANDA-II has an $A_{eff}$ significantly larger than any other 

820: contemporaneously-operating neutrino detector (e.g. Baikal \citep{Baikal}, SuperKamiokande 

821: \citep{SuperK}, and SNO \citep{SNO}).  A determination of the relative MRF for a 

822: subset of bursts from the year 2000 analysis is shown in Figure 7 (the arrow indicates the 

823: MRF for the selected criteria).

824: 

825: Prior to ``unblinding'' the analysis and determining the number of events we observe, we 

826: determine the flux sensitivity to simulated GRB neutrinos.  Results from the 268 bursts 

827: observed from 1997 to 1999 have been presented previously \citep{Bay, Rellen}.  We combine 

828: these initial observations with the results from the analysis of 151 bursts in the data 

829: collected in 2000-2003.  The flux sensitivity for all 419 bursts is the MRF prior to 

830: observations (see equation 5) multiplied by the normalization of the input spectrum; that 

831: is, 

832: $E^{2}$${\Phi}_{\mathrm{\nu}}$ ${\leq}$2 ${\times}$ $10^{-8}$ ${\mathrm{GeV cm^{-2} s^{-1} 

833: sr^{-1}}}$ for a Waxman-Bahcall muon neutrino spectrum with 90\% of the events expected 

834: between $\sim$10 TeV and $\sim$3 PeV.  This sensitivity is calculated prior to the inclusion of 

835: systematic uncertainties.

836: 

837: \subsection{Uncertainties in Observation and Modeling}

838: 

839: There are several potential sources of systematic uncertainty in this analysis, including 

840: the Monte Carlo simulations of signal events, the modeling of the scattering and absorption 

841: lengths of the South Pole ice, and the OM response to incident photons.  For the flux upper 

842: limits incorporating IPN bursts, the potential for inclusion of bursts which do not fit 

843: models based upon BATSE triggered bursts contributes to the overall uncertainty as well.  

844: Additionally, some bursts are of unknown duration--for the purposes of this search, they 

845: were classified as long-duration bursts so that we would not needlessly exclude any 

846: possible signal events.  However, including all such bursts will potentially overestimate 

847: the signal event predictions for models based solely upon long-duration bursts.  Finally, 

848: previous results from 1997-1999 were applied only to the Waxman-Bahcall model; limitations 

849: in the simulation procedures in place at that time means that adapting these results to 

850: other models will introduce uncertainties in the expected neutrino event rate.

851: 

852: The scattering and absorption lengths of the ice were measured during the 1999-2000 austral 

853: summer with in situ lasers and LED flashers \citep{Kurt}.  While these measurements were 

854: extremely accurate, the limited precision with which they were implemented in our detector 

855: simulations contributes about $15\%$ to the overall uncertainty.  Furthermore, the quantum 

856: efficiency of the photomultiplier tubes is known to within $10\%$, while the transmission 

857: efficiency of the glass pressure housing and the optical gel is known to a comparable 

858: precision.  However, triggering depends on the detection of photons by 24 or more PMTs, so 

859: the uncertainty in a single OM does not translate directly into an uncertainty in the 

860: expected flux.  Detailed simulations show that the quantum and transmission efficiencies 

861: together contribute only about 7\% uncertainty in the expected neutrino flux \citep{Reco}.  

862: Though the GRB search implements a different methodology from other IceCube analyses 

863: \citep[e.g. the point source search detailed in][]{Elisa}, the values for the individual 

864: contributions to the uncertainty are consistent across these different analyses.

865: 

866: Additionally, a statistical correction is required when IPN bursts are incorporated into 

867: the flux upper limits for models initially based on BATSE observations.  In principle, 

868: BATSE has a sensitivity comparable to the suite of other IPN satellites treated 

869: collectively; observationally, their duration distribution seems qualitatively to be 

870: derived from the same bimodal population (Figure 8).  However, the characteristics of the 

871: bursts detected by satellites with different sensitivities are not completely identical.  

872: BATSE non-triggered bursts have on average less than 1/10 of the peak photon flux of their triggered 

873: counterparts, and if we assume that the neutrino flux scales as the photon flux, then 

874: including non-triggered bursts in the upper limit calculation would artificially 

875: increase the expected number of signal events, and thus lead to a flux upper limit that is 

876: too restrictive.  We calculate (see Appendix A) that 12\% of the IPN bursts should not be 

877: considered equivalent to BATSE triggered bursts, and thus should be excluded from the 

878: dataset.  This leads to a 3\% correction in the number of expected signal events.  

879: Furthermore, for models based solely on long-duration bursts such as \citep{Murase, 

880: Razzaq}, the inclusion of bursts of unknown duration may also lead to an overestimation of 

881: the number of expected signal events.  In Appendix A, we derive a statistical correction of 

882: 6\% to the expected number of signal events due to this effect.

883: 

884: Finally, we determine the uncertainty introduced when the previous results from 1997-1999 

885: are applied to theoretical predictions other than the Waxman-Bahcall model.  Though the 

886: uncertainties specifically for the Waxman-Bahcall model are well understood and are 

887: incorporated into the previous results, limitations in the simulation procedures at the 

888: time of the previous analysis lead to a further uncertainty in the neutrino event rate for 

889: the Murase-Nagataki and Razzaque et al. models of $\sim$20\%.  When we combine the results 

890: from the 268 bursts from 1997-1999 with the results from 151 bursts from 2000-2003 into a 

891: single flux upper limit, we assume conservatively that the neutrino event rate for the 

892: bursts from 1997-1999 is overestimated by 20\%.  

893: 

894: All significant sources of uncertainty for the GRB analysis, along with the correction 

895: factors, are summarized in Table 4.  While the reduction in the expected neutrino event 

896: rate for the 1997-199 bursts is not specifically enumerated in this Table, it is 

897: incorporated into the relevant flux upper limits discussed in the next section.  Assuming 

898: no correlation among the other uncertainties, we summed the different factors in quadrature 

899: and applied the other relevant corrections to obtain a total uncertainty of +16\%/-17\% 

900: (+15\%/-18\% for models based on long-duration bursts only) in the total detector exposure, 

901: and therefore in the number of signal events and the flux and fluence upper limits.  This 

902: is comparable to the uncertainty determined by \citep{Hodges}, who also characterized the 

903: agreement between the simulated signal events and high-quality downgoing muon events, which 

904: served as a proxy for the expected signal events for AMANDA analyses.

905: 

906: \section{Results} \label{bozomath}

907: 

908: We observe zero events from the 419 Northern Hemisphere bursts searched during the years 

909: 1997 to 2003, which is consistent with the background estimate of 1.74 events (Table 

910: 5)\footnote{We also searched for neutrino emission from 153 additional non-triggered bursts 

911: discovered in the BATSE archival data; we observed zero events from these bursts as well.  

912: We do not include these results in the flux upper limits or MRF determinations.}.  Since 

913: the observed number of events is less than the expected background, the flux upper limits 

914: for the coincident muon neutrino search is approximately a factor of three better than the 

915: expected sensitivity (i.e. the observed MRF for a Waxman-Bahcall flux is 1.3 compared to 

916: the expected value of 3.8).  Figure 9 shows the 90\% C.L. flux upper limits relative to the 

917: Waxman-Bahcall, Razzaque, and Murase-Nagataki models.  Though our analysis was restricted 

918: to bursts located in the Northern Hemisphere (2${\pi}$ sr), all flux upper limits are for 

919: the entire sky (4${\pi}$ sr).  Including the systematic uncertainties in the manner 

920: outlined by \citet{Conrad}, we calculate the coincident muon neutrino flux upper limit for 

921: the Waxman-Bahcall spectrum to have a normalization at 1 PeV of 

922: 

923: \noindent

924: $E^{2}$$\Phi_{\mathrm{\nu}}$ ${\leq}$ 6.0 $\times$ $10^{-9}$ $\mathrm{GeV cm^{-2} s^{-1} sr^{-1}}$,

925: 

926: \noindent

927: with 90\% of the events expected between $\sim$10 TeV and $\sim$3 PeV.

928: 

929: We place similar constraints on the model parameters of \citet{Murase}.  Based on our null 

930: result, Parameter Set C is highly disfavored for all variations in their parameters, though 

931: this particular set is disfavored on other grounds as well, and is only briefly described 

932: in their work.  Parameter Set A is ruled out (MRF=0.82) by the current AMANDA 

933: observations at the 90\% confidence level.  However, it is important to note that Parameter 

934: Set A uses a baryon loading factor that is fine-tuned to provide significant neutrino flux.  

935: Other, possibly more realistic, values for the baryon loading would significantly reduce 

936: the expected neutrino emission, and therefore result in an MRF that is higher by an order 

937: of magnitude or more.  The original model incorporates only long-duration bursts that 

938: follow the cosmic star-formation rate \citep{Nagataki}; incorporating all short bursts 

939: would yield flux upper limits that are better than those presented here by approximately 

940: 13\%, which includes removing the ``correction'' due to incorporating of bursts of unknown 

941: duration (see Section 3.4).

942: 

943: Our combined results from 1997-2003 also constrain the supranova model of Razzaque et al.  

944: We begin by considering the assumption that all GRBs are preceded by supernovae that 

945: produce a circumburst environment ideally suited for neutrino production.  The observed MRF 

946: for this case is 0.40, and thus we exclude the predicted neutrino flux at the 90\% level.  

947: Furthermore, the flux upper limit determined for this model is derived from observations of 

948: long bursts only.  As with the results of \citet{Murase}, if this model is expanded to 

949: include short-duration bursts, the flux upper limit improves by approximately 13\%.  

950: However, only a very small number of all bursts (${\sim}$4 out of many thousands) have been 

951: observed in association with SNe.  And, as described in Section 2, at least a fraction of 

952: these SNe did not occur at an ideal time relative to the burst.  Thus, AMANDA's results 

953: confirm previous observations that lead us to expect less than maximal emission from this 

954: model of GRB neutrino production.

955: 

956: Finally, we observe zero events (on an expected background of 0.2 events) from the 

957: precursor time period of the bursts from 2001-2003 (Table 6).  The precursor model of 

958: neutrino production was tested for only a small subset of the long-duration bursts, and the 

959: neutrino energy spectrum peaks at a level where the AMANDA-II sensitivity is greatly 

960: reduced.  Thus, the flux upper limit for the precursor model is significantly less 

961: restrictive.

962: 

963: The results of these analyses can also be applied to any other hypothesized spectrum by 

964: using the Green's Function Fluence Limit formula, in a method similar to that presented by 

965: the Super-Kamiokande Collaboration \citep{SuperK}.  By folding the energy-dependent 

966: sensitivity of the detector into a desired theoretical spectrum, one can straightforwardly 

967: calculate a flux upper limit for that specific spectrum.  The Green's Function fluence 

968: upper limit for AMANDA-II (Figure 10) extends several orders of magnitude in energy beyond 

969: the range of the Super-Kamiokande limit, and is approximately an order of magnitude lower 

970: than the Super-Kamiokande results in the region of overlap, primarily due to the much 

971: larger effective area of AMANDA-II.  For example, at 100 TeV we calculate 

972: $F_{\mathrm{\nu}}$ ${\leq}$ 1.7 $\times$ $10^{-7}$ $\mathrm{cm^{-2}}$ (see also Appendix 

973: B).  As this method does not rely on averaging burst properties (as many specific models 

974: do), it is particularly effective for incorporating large burst-to-burst variations in 

975: expected muon neutrino flux \citep[e.g. for GRB030329, see][]{MikeSwift}.

976: 

977: \section{Conclusion and Outlook}

978: 

979: The AMANDA dataset has been searched for muon neutrino emission from more than 400 GRBs 

980: based on temporal and spatial coincidence with photon detections from numerous other 

981: observatories.  We determined that the detector was operating in a stable fashion during 

982: all of these bursts, and we have shown that the application of a number of data selection 

983: criteria lead to an optimized value of the Model Rejection Factor for the Waxman-Bahcall 

984: neutrino spectrum.  After the application of these criteria, zero neutrino events were 

985: observed in coincidence with the bursts, resulting in the most stringent upper limit on the 

986: muon neutrino flux from GRBs to date.  We have compared this limit to the flux predictions 

987: from several prominent GRB models based on averaged burst properties.  We constrain the 

988: parameter space of a number of these models at the 90\% confidence level; in particular, 

989: our flux upper limit is more than a factor of 2 below the most optimistic predictions of 

990: Razzaque et al.  However, we do not yet rule out the predictions of the canonical Waxman \& 

991: Bahcall model.  Furthermore, because individual bursts vary significantly in their expected 

992: neutrino spectra, we have presented a spectrum-independent method for determining flux 

993: upper limits for these bursts.  The observations detailed in this work will play a 

994: significant role as future analyses seek to further constrain various theoretical models.

995: 

996: Finally, AMANDA's search for muon neutrinos from more recent GRBs will benefit greatly from 

997: the advanced capabilities of the Swift satellite \citep{Swift}, as will the GRB searches 

998: of other neutrino observatories currently in operation \citep{Baikal,Nestor,Antares}.  

999: While Swift's rate of GRB detections is lower than that of BATSE, the spatial localizations 

1000: of the bursts by Swift are much more precise, which will obviate the need for a special 

1001: analysis of poorly-localized bursts with its accompanying reduction in signal detection 

1002: efficiency.  Additionally, the InterPlanetary Network of satellites will continue to operate, 

1003: and will incorporate newer instruments as they come online.  In particular, future missions 

1004: such as the Gamma-Ray Large Area Space Telescope \citep{GLAST} will provide an even greater 

1005: number of GRB localizations for use in neutrino searches.  Furthermore, while analyses similar 

1006: to the one presented here will continue to search specifically for muon neutrino flux in 

1007: coincidence with photon observations of gamma-ray bursts, the method described here can be 

1008: expanded to search for neutrinos correlated with other transient point sources as well (see 

1009: Appendix C).  In the future, AMANDA and its successor, IceCube, will have many more 

1010: opportunities to detect neutrino emission from a host of astrophysical sources.  Construction 

1011: of IceCube is currently underway, and the instrumented volume for the partial detector is already 

1012: significantly larger than the final instrumented volume of AMANDA.  A fully-instrumented 

1013: IceCube detector should surpass AMANDA's flux upper limits within its first few years of 

1014: operation.

1015: 

1016: \acknowledgments

1017: 

1018: \section{Acknowledgements}

1019: 

1020: The authors are thankful to E. Waxman and K. Murase for productive and thought-provoking discussions, S. Desai for detailed discussions regarding the 

1021: Super-Kamiokande GRB analysis, and the additional members of the Interplanetary Network for providing data that were crucial to this work.  The IceCube 

1022: collaboration acknowledges the support from the following agencies: National Science Foundation-Office of Polar Program, National Science 

1023: Foundation-Physics Division, University of Wisconsin Alumni Research Foundation, Department of Energy, and National Energy Research Scientific 

1024: Computing Center (supported by the Office of Energy Research of the Department of Energy), the NSF-supported TeraGrid system at the San Diego 

1025: Supercomputer Center (SDSC), and the National Center for Supercomputing Applications (NCSA); Swedish Research Council, Swedish Polar Research 

1026: Secretariat, and Knut and Alice Wallenberg Foundation, Sweden; German Ministry for Education and Research, Deutsche Forschungsgemeinschaft (DFG), 

1027: Germany; Fund for Scientific Research (FNRS-FWO), Flanders Institute to encourage scientific and technological research in industry (IWT), Belgian 

1028: Federal Office for Scientific, Technical and Cultural affairs (OSTC); the Netherlands Organisation for Scientific Research (NWO); M. Ribordy 

1029: acknowledges the support of the SNF (Switzerland); J. D. Zornoza acknowledges the Marie Curie OIF Program (contract 007921).

1030: K. Hurley is grateful for IPN support under the following contractsand grants: Ulysses, JPL958056; Mars Odyssey, JPL 1282043; HETE, 

1031: MIT-SC-R-293291; the U.S. BeppoSAX Guest Investigator program; the U.S. INTEGRAL Guest 

1032: Investigator program.  Integration of NEAR into the Interplanetary Network was supported by 

1033: NASA's NEAR Participating Scientist program.  Integration of RHESSI was supported by NASA's 

1034: Long Term Space Astrophysics program under grant NAG5-13080.

1035: 

1036: 

1037: {\it Facilities:} \facility{AMANDA}

1038: 

1039: \appendix

1040: 

1041: \section{Model-Dependent Statistical Corrections to Flux Upper Limits}

1042: 

1043: Though the ${\nu}$ flux formulation of \citet{Waxman} explicitly links GRB neutrinos to the 

1044: UHECR flux, elsewhere a formulation based on BATSE observations is treated in a comparable 

1045: fashion, and is considered to arise from the same underlying phenomena \citep{WaxBah}.  

1046: Thus it is necessary to address the limitations introduced by AMANDA's reliance upon BATSE 

1047: observations.  As described in Section 3.4, models defined initially in terms of BATSE 

1048: observations were also applied to bursts detected by the other IPN satellites.  However, we 

1049: cannot assume that characteristics of bursts detected by satellites with different 

1050: sensitivities are completely identical.  Since BATSE was decommissioned in May of 2000, 

1051: there is no longer a way to cross-correlate the two datasets.  Non-triggered BATSE bursts 

1052: have on average less than 1/10 of the peak photon flux of the triggered bursts; assuming 

1053: that the energy of neutrinos scales with the energy carried by gamma rays, we expect only a small 

1054: fraction of the standard neutrino flux from these non-triggered bursts.  Thus, if 

1055: non-triggered bursts are inadvertently included in the flux upper limit, they will 

1056: artificially improve that limit, because the extra bursts are assumed to have a larger 

1057: neutrino flux than they would actually possess.   

1058: 

1059: During the period of simultaneous operation from 1991 to 2000, 1088 IPN bursts were 

1060: observed by BATSE, 953 of which were triggered.  Undoubtedly some of these bursts did not 

1061: trigger BATSE for reasons other than a lower flux.  For example, BATSE may have been 

1062: powered down, may have been in the vicinity of the South Atlantic Anomaly, or may have 

1063: experienced unrelated on-board performance problems.  However, we assume conservatively 

1064: that all such bursts did in fact exhibit the lower flux common to non-triggered bursts.  

1065: Therefore, ${\sim}$12$\%$ of the IPN bursts should not actually be a part of the dataset 

1066: that is compared with the models that are based upon BATSE's triggered GRB rate.  Because 

1067: IPN bursts are expected to contribute ${\sim}$25\% of our detectable signal, this effect 

1068: reduces the total expected neutrino flux by ${\sim}$3\%.  This correction is applied 

1069: asymmetrically to the overall uncertainty, because it can hinder, but not improve, the 

1070: effectiveness of the analysis (see Table 4).

1071: 

1072: For models based solely on long-duration bursts, such as \citep{Murase, Razzaq}, the inclusion 

1073: of bursts of unknown duration may also lead to an overestimation of the expected signal 

1074: events, and thus a flux upper limit that is too restrictive.  In order to ensure that we 

1075: would not exclude potentially detectable neutrino events, the 75 bursts of unknown duration 

1076: included in the dataset are assumed to last 100 s (for 1997-1999) or 50 s (for 2000-2003).  

1077: Thus, for purposes of data analysis, they are classified as long-duration bursts.  However, 

1078: this necessitates a statistical correction to the resulting flux limits.  We assume that up 

1079: to 1/3 of these bursts may in fact be short-duration, based upon the standard ratio of 

1080: short- to long-duration bursts observed by BATSE.  So, of the 389 bursts known (or assumed) 

1081: to be long-duration, 25 were excluded from the relevant limits, thus reducing the expected 

1082: number of signal events by 25/389, or ${\sim}$6\%.  This correction is likewise applied 

1083: asymmetrically to the overall uncertainty.

1084: 

1085: \section{Green's Function Fluence Upper Limit Calculation}

1086: 

1087: We show here sample calculations of the differential neutrino fluence upper limit, as well 

1088: as a procedure to determine the integrated fluence and flux upper limits, following the 

1089: Green's Function method set out in Section 3 of \citet{SuperK}.  The fluence upper limit 

1090: calculation assumes a monochromatic neutrino spectrum; the calculation is repeated at 

1091: different values of the neutrino energy.  The benefit of this method is that an integrated 

1092: fluence upper limit can then be determined for any input spectrum, whether it be based on 

1093: all of the bursts in this dataset or only on a subset of all bursts.

1094: 

1095: The fluence upper limit is defined as

1096: \begin{equation}

1097: F(E) \leq \frac{{\mathrm{N_{90}}}}{{\mathrm{A_{eff,{\nu}}}}({\mathrm{E_{\nu}}})}

1098: \end{equation}

1099: \noindent

1100: where ${\mathrm{N_{90}}}$ is ${\mathrm{{\mu}_{90}}}$/${\mathrm{N_{Bursts}}}$ and 

1101: ${\mathrm{A_{eff,{\nu}}}}$ is the energy-dependent neutrino effective collecting area (see 

1102: Section 3.3)\footnote{Instead of using the neutrino effective area, one could also use the 

1103: muon effective area multiplied by the neutrino to muon conversion probability \citep[as 

1104: in][]{SuperK}; in the case of AMANDA one must also account for attenuation of neutrinos in 

1105: the earth.}.

1106: 

1107: Figure 10 is determined by the results of AMANDA's 2000-2003 observations.  For example, 

1108: ${\mathrm{{\mu}_{90}}}$ = 1.30 and ${\mathrm{N_{Bursts}}}$ = 151; therefore, 

1109: ${\mathrm{N_{90}}}$ = 8.61 $\times$ 10$^{-3}$.

1110: 

1111: For $E_{\mathrm{{\nu}}}$ = 100 TeV (near the peak of the predicted neutrino flux), 

1112: $A_{\mathrm{{eff,{\nu}}}}$ = 5.0 $\times$ 10$^{4}$ cm$^2$ therefore F(100 TeV) ${\leq}$ 1.7 

1113: $\times$ 10$^{-7}$ cm$^{-2}$.

1114: 

1115: We now determine the integrated fluence upper limit explicitly for an E$^{-2}$ spectrum, as 

1116: well the Waxman-Bahcall spectrum.  First, the integrated fluence, ${\mathrm{F_{int}}}$ for 

1117: an E$^{-2}$ spectrum is 

1118: \begin{equation}

1119: F_{\mathrm{int}} \leq [\int_{250 {\mathrm{GeV}}}^{10^{7} {\mathrm{GeV}}} \frac{C 

1120: E_{\mathrm{\nu}}^{-2}}{F(E)} {\mathrm{dE_{\nu}}}]^{-1} =

1121: 1.4 {\times} 10^{-5} {\mathrm{cm^{-2}}},

1122: \end{equation}

1123: \noindent

1124: where C is the factor required to normalize the neutrino spectrum to unity---in this case, 

1125: C = 250 GeV.  This integrated fluence upper limit is significantly lower than the results 

1126: of similar calculations performed by \citet{SuperK} (we combine the ${\mathrm{{\nu}_{\mu}}}$ 

1127: and ${\mathrm{\bar{{\nu}}_{\mu}}}$ fluences into a single limit, while they present two 

1128: separate fluence upper limits).  However, a direct, quantitative comparison between these 

1129: two results cannot be made due to the vastly different energy ranges of the two instruments.  

1130: Note also the limits of integration employed here---though AMANDA is sensitive to neutrinos 

1131: at higher and lower energies, the vast majority of the flux from GRBs is expected to come 

1132: from neutrinos of a few hundred GeV to a few PeV.

1133: 

1134: Now we determine the integrated fluence upper limit for the Waxman-Bahcall spectrum, to 

1135: provide a further example of the wide applicability of the Green's Function method:

1136: \begin{equation}

1137: F_{\mathrm{int}} \leq [\int_{250 {\mathrm{GeV}}}^{10^{5} {\mathrm{GeV}}} 

1138: \frac{C E_{\mathrm{\nu}}^{-1} E_{\mathrm{Break}}^{-1}} {F(E)} {\mathrm{dE_{\nu}}}

1139: + \int_{10^{5}}^{10^{7}} \frac{C E_{\mathrm{\nu}}^{-2}}{F(E)} {\mathrm{dE_{\nu}}}]^{-1}

1140: = 5.3 {\times} 10^{-7} {\mathrm{cm^{-2}}},

1141: \end{equation}

1142: \noindent

1143: where C again is the constant required to normalize the overall spectrum to 

1144: unity; here C = $7.0 {\times} 10^{-5} {\mathrm{GeV}}$.

1145: 

1146: Finally, we compare this fluence upper limit to the flux upper limit derived for the 

1147: Waxman-Bahcall spectrum in Section 4.  To do this, we must convert the integrated 

1148: fluence upper limit into a differential all-sky flux upper limit per burst; that is, from 

1149: units of cm$^{-2}$ to units of $\mathrm{GeV^{-1} cm^{-2} s^{-1} sr^{-1}}$:

1150: 

1151: \begin{equation}

1152: \frac{F_{\mathrm{int}}}{{\Omega} {\mathrm{t}}} = \frac{5.3 {\times} 10^{-7}} {(4{\pi}) 

1153: (3.15 {\times} 10^{7}/700)}= 9.4 \times 10^{-13} 

1154: {\mathrm{cm^{-2} s^{-1} sr^{-1}}}.

1155: \end{equation}

1156: 

1157: Next, we multiply by the normalization of the energy spectrum and take the differential to 

1158: provide a flux upper limit of

1159: 

1160: \noindent

1161: E$^{2}$${\Phi}_{\mathrm{\nu}}$ ${\leq}$ 1.3 ${\times}$ 10$^{-8}$ $\mathrm{GeV cm^{-2} 

1162: s^{-1} sr^{-1}}$.

1163: 

1164: \noindent

1165: This is nearly identical to the flux upper limit derived in the manner described in Section 

1166: 3 for 151 bursts from 2000 to 2003 (see also Table 5, where an MRF of 2.5 yields a flux 

1167: upper limit of E$^{2}$${\Phi}_{\mathrm{\nu}}$ ${\leq}$ 1.1 ${\times}$ 10$^{-8}$ 

1168: $\mathrm{GeV cm^{-2} s^{-1} sr^{-1}}$, consistent with the result derived above to within 

1169: the applicable uncertainties).

1170: 

1171: Thus we show that the Green's Function Method agrees with calculations which explicitly 

1172: incorporated prior assumptions about the GRB neutrino spectrum.  Therefore, this alternate 

1173: method provides a powerful tool for determining the flux upper limit based on AMANDA 

1174: observations for any proposed neutrino spectrum.

1175: 

1176: \section{Expanding the GRB Search to Other Transient Point Sources}

1177: 

1178: While this work has provided the most stringent upper limit to date specifically for muon 

1179: neutrino flux for gamma-ray bursts in coincidence with photon observations, the method 

1180: described above can be expanded to search for other transient point sources as well.  X-ray 

1181: flares occurring minutes to hours after a GRB are thought to be caused by re-activation of 

1182: the GRB central engine, and are a natural candidate for correlated neutrino searches 

1183: \citep{Murase2}. Additionally, photon emission from supernovae could 

1184: be used as a key element in searches for neutrino emission from jet-driven supernovae and 

1185: ${\gamma}$-ray dark (``choked'') GRBs \citep{Razza2}.  Jet-driven supernovae are expected 

1186: to accelerate baryonic material to mildly relativistic energies (the Lorentz boost 

1187: ${\Gamma}$ ${\sim}$ a few), which may subsequently result in significant neutrino emission 

1188: \citep{Ando}.  Not all supernovae will be jet-driven, but population estimates vary between 

1189: 0.2\% and 25\% of all type Ib/c SNe \citep{Putten,Berger,Soder}.  Given the number of such 

1190: supernovae observed annually, it is reasonable to search for a neutrino signal from these 

1191: events.

1192: 

1193: Another reason to search for neutrino emission from supernovae becomes apparent when we 

1194: consider the recently-established SN-GRB connection.  Several supernovae (including 1998bw 

1195: and 2003dh) are known to be associated with GRBs.  Furthermore, Razzaque et al. (2003b) 

1196: describe a scenario where as many as $10^{3}$ times the standard number of GRBs occur, 

1197: though in these bursts the photon jet does not succeed in escaping the stellar envelope 

1198: (the ${\gamma}$-ray dark GRBs).  For these types of bursts, no gamma-rays will be observed.  

1199: However, if even a fraction of these GRBs are associated with SNe (the fraction for observed 

1200: GRBs has been calculated to be in the range of $10^{-2}$ to $10^{-3}$ \citep{Bissaldi}), 

1201: then it will be possible to search for neutrinos in the time period surrounding the SN 

1202: emission (provided the SN start time, the GRB time delay relative to the SN, and the 

1203: duration of the GRB can be estimated with sufficient precision).  Because these SNe are 

1204: localized transient phenomena, the primary selection criteria for the GRB analysis (spatial 

1205: and temporal correlation) are an excellent starting point for such a search, though it is 

1206: possible that not all of the other data quality cuts used in the GRB search would be 

1207: optimal for a supernova search.  Finally, it is also possible to complement any of the 

1208: transient point source searches described above by inverting the search algorithm, that is, 

1209: by implementing Target of Opportunity photon searches based on spatio-temporal localization 

1210: of potential neutrino events \citep{Marek}.  Any of these searches can potentially be of 

1211: great benefit to the long-term goals of multi-messenger astronomy. 

1212: 

1213: 

1214: \begin{thebibliography}{}

1215: \bibitem[Achterberg et al.~(2007)]{Casc} Achterberg, A. et al., ApJ, in press (astro-ph/0702265)

1216: \bibitem[Achterberg et al.~(2006b)]{Elisa} Achterberg, A. et al., submitted to PRD, November 2006 (astro-ph/0611063)

1217: \bibitem[Achterberg et al.~(2006a)]{IceCub} Achterberg, A. et al., Astropart. Phys. $\bf{26}$ (2006) 155 (astro-ph/0604450)

1218: \bibitem[Ackermann et al.~(2006)]{Kurt} Ackermann, M. et al., J. Geophys. Res. $\bf{111}$ (2006), D13203

1219: \bibitem[Aguilar et al.~(2006)]{Antares} Aguilar, A.J. et al., Astropart. Phys. $\bf{26}$ (2006) 314-324; see also the Antares 

1220: Neutrino Telescope Page: http://antares.in2p3.fr

1221: \bibitem[Aharmim et al.~(2000)]{SNO} Aharmim, B. et al., NIM $\bf{A449}$ (2000) 172-207

1222: \bibitem[Ahrens et al.~(2004)]{Reco} Ahrens, J. et al., NIM $\bf{A54}$ (2004) 169

1223: \bibitem[Ahrens et al.~(2003a)]{Diffuse} Ahrens, J. et al., PRL $\bf{90}$ (2003) 251101

1224: \bibitem[Ahrens et al.~(2003b)]{Point} Ahrens, J. et al., ApJ $\bf{583}$ (2003) 1040

1225: \bibitem[Ahrens et al.~(2002)]{AMANDA} Ahrens, J. et al., PRD $\bf{66}$  (2002) 012005

1226: %% \bibitem[Ambrosio et al. (2001)]{Macro} Ambrosio, M. et al., ApJ $\bf{546}$ (2001) 1038-1054

1227: \bibitem[Andr\'es et al.~(2001)]{Nature} Andr\'es, E. et al., Nature $\bf{410}$  (2001) 441-443

1228: \bibitem[Ando \& Beacom~(2005)]{Ando} Ando, S. and J. Beacom, PRL $\bf{95}$ (2005) 061103

1229: \bibitem[Bay~(2000)]{Bay} Bay, R.C. Ph.D. Thesis, University of California (2000) (astro-ph/0008255)

1230: \bibitem[Berger et al.~(2003)]{Berger} Berger, E. et al., ApJ $\bf{599}$ (2003) 408

1231: \bibitem[Bissaldi et al.~(2006)]{Bissaldi} Bissaldi, E. et al., A\&A, in press (astro-ph/0702652)

1232: \bibitem[Burrows et al.~(2005)]{Swift} Burrows, D. et al., Space Sci. Rev. $\bf{120}$ (2005) 165; 

1233: see also the Swift Gamma-Ray Burst Mission Page: http://swift.gsfc.nasa.gov/docs/swift/swiftsc.html

1234: \bibitem[Carson~(2006)]{GLAST} Carson, J. for the GLAST Collaboration, Proc. 2nd TeV Workshop, Madison, WI (2006) (astro-ph/0610960); see also 

1235: the GLAST Mission Page: http://glast.gsfc.nasa.gov

1236: \bibitem[Chirkin \& Rhode~(2004)]{Dima} Chirkin, D. and W. Rhode, hep-ph/0407075

1237: \bibitem[Conrad et al.~(2003)]{Conrad} Conrad, J. et al., PRD $\bf{67}$ (2003) 012002 

1238: \bibitem[Costa et al.~(2003)]{Beppo} Costa, E. et al., Nature $\bf{386}$ (1997) 686-688 

1239: \bibitem[Dermer \& Atoyan~(2006)]{Dermer1} Dermer, C. and A. Atoyan, New Journal of Physics $\bf{8}$ (2006) 122 (astro-ph/0606629)

1240: \bibitem[Dermer \& Atoyan~(2003)]{Dermer2} Dermer, C. and A. Atoyan, PRL $\bf{91}$ (2003) 1102

1241: \bibitem[Eicher et al.~(1989)]{Eicher} Eicher, D. et al., Nature $\bf{340}$ (1989) 126-128 

1242: \bibitem[Feldman \& Cousins~(1998)]{FeldCous} Feldman, G. and R. Cousins, PRD $\bf{57}$ (1998) 3873-3889

1243: \bibitem[Fishman et al.~(1993)]{BATSE} Fishman, G.J. et al., A\&A Supp $\bf{97}$ (1993) 17-20

1244: \bibitem[Frail et al.~(2001)]{Frail} Frail, D. et al., ApJ $\bf{562}$ $\bf{L}$55-58 (2001)

1245: \bibitem[Fryer \& M\'esz\'aros~(2003)]{Fryer} Fryer, C. and P. M\'esz\'aros, ApJ $\bf{588}$ $\bf{L}$25-28 (2003) 

1246: \bibitem[Fukuda et al.~(2002)]{SuperK} Fukuda, S. et al., ApJ $\bf{578}$ (2002) 317

1247: \bibitem[Gandhi et al.~(1998)]{Gandhi} Gandhi, R. et al., PRD $\bf{58}$ (1998) 093009

1248: \bibitem[Gandhi et al.~(1996)]{Gandhi2} Gandhi, R. et al., Astropart. Phys. $\bf{5}$ (1996) 81-110 

1249: \bibitem[Hardtke~(2002)]{Rellen} Hardtke, R., Ph.D Thesis, University of Wisconsin (2002)

1250: \bibitem[Heck et al.~(1998)]{Cors} Heck, D. et al., Report FZKA 6019 (1998).  An update of the original Corsika (6.030) was used in this 

1251: analysis.  See http://www-ik.fzk.de/corsika/ for additional information and more recent versions.

1252: \bibitem[Hill \& Rawlins~(2003)]{Hill03} Hill, G.C. and K. Rawlins, Astropart. Phys. $\bf{19}$ (2003) 393

1253: \bibitem[Hill~(1997)]{Hill97} Hill, G.C., Astropart. Phys. $\bf{6}$ (1997) 215

1254: \bibitem[Hodges~(2006)]{Hodges} Hodges, J. for the IceCube Collaboration, Proc. 2nd TeV Workshop, Madison, WI (2006) (astro-ph/0611597)

1255: %% \bibitem[Hughey \& Taboada (2005)]{Brennan} Hughey, B. and I. Taboada for the IceCube Collaboration, Proc. 29th ICRC, Pune, India, (2005) (astro-ph/0509330)

1256: \bibitem[Lewin et al.~(2006)]{Lewin} Hurley, K. et al., in ``Compact Stellar X-Ray Sources'', Lewin, W. et al., Eds. Cambridge University Press, Cambridge (2006)

1257: \bibitem[Hurley~(1998)]{Hurley} Hurley, K., Astron. Telegram \#19 (1998)

1258: \bibitem[Kalmykov \& Ostapchenko~(1993)]{Kalmykov} Kalmykov, N. and S. Ostapchenko, Phys. Atom. Nucl. $\bf{56}$ (1993) 346

1259: %% \bibitem[Karle (2003)]{IceCub} Karle, A. et al., Nuc. Phys. B Proc. Supp. $\bf{118}$ (2003) 388

1260: \bibitem[Kashti \& Waxman~(2005)]{Kashti} Kashti, T. and E. Waxman, PRL $\bf{95}$ (2005) 181101

1261: \bibitem[Kommers~(1998)]{Kommers} Kommers, J. Ph.D. Thesis, MIT (1998); see also http://space.mit.edu/BATSE

1262: \bibitem[Kowalski \& Mohr~(2007)]{Marek} Kowalski, M. and A. Mohr, astro-ph/0701618

1263: \bibitem[Kuehn~(2007)]{myurl} Kuehn, K., http://icecube.wisc.edu/$\sim$kuehn/grb9703.html (AMANDA-GRB Supplemental Information Page)

1264: \bibitem[Lipari \& Stanev~(1991)]{Lipari} Lipari, P. and T. Stanev, PRD $\bf{44}$ (1991) 113543

1265: \bibitem[Markwardt et al.~(2005)]{Mark} Markwardt, C.B. et al., ApJ $\bf{633}$ $\bf{L}$77-80 (2005)

1266: \bibitem[Murase \& Nagataki~(2006a)]{Murase} Murase, K. and S. Nagataki, PRD $\bf{73}$ (2006) 063002

1267: \bibitem[Murase \& Nagataki~(2006b)]{Murase2} Murase, K. and S. Nagataki, PRL $\bf{97}$ (2006) 051101 (astro-ph/0604437)

1268: \bibitem[Murase \& Nagataki~(2006c)]{Nagataki} Murase, K. and S. Nagataki, private communication

1269: %% \bibitem[M\'esz\'aros \& Waxman~(2001)]{Meszar} M\'esz\'aros, P. and E. Waxman, PRL $\bf{87}$ (2001) 1102 (and references therein).  

1270: \bibitem[Paczy\'nski~(1998)]{Pacz} Paczy\'nski, B., ApJ $\bf{494}$ $\bf{L}$45 (1998)

1271: \bibitem[Piran~(2004)]{Piran1} Piran, T., Rev. Mod. Phys. $\bf{76}$ (2004) 1143

1272: \bibitem[Piran~(2002)]{Piran2} Piran, T., Science $\bf{295}$ (2002) 986

1273: \bibitem[Razzaque et al.~(2003a)]{Razzaq} Razzaque, S. et al., PRD $\bf{68}$ (2003) 3001

1274: \bibitem[Razzaque et al.~(2003b)]{Razza2} Razzaque, S. et al., PRL $\bf{90}$ (2003) 241103

1275: \bibitem[Resvanis et al.~(2003)]{Nestor} Resvanis, L.K. et al., Nuc. Phys. B Proc. Supp. $\bf{122}$ (2003) 24-39; see also the NESTOR Institute for 

1276: Astroparticle Physics Page: http://www.nestor.org.gr

1277: \bibitem[Roming et al.~(2006)]{Roming} Roming, P.W.A. et al., ApJ $\bf{651}$ (2006) 985 (astro-ph/0605005)

1278: \bibitem[Soderberg~(2005)]{Soder} Soderberg, A. Nuovo Cim. 28C (2005) 563-573

1279: \bibitem[Spiering et al.~(2004)]{Baikal} Spiering, C. et al., astro-ph/0404096

1280: \bibitem[Stamatikos~(2005)]{Mike} Stamatikos, M. for the IceCube Collaboration, et al. Proc. 29th ICRC, Pune, India (2005) 471-474 (astro-ph/0510336)

1281: \bibitem[Stamatikos~(2006)]{MikeSwift} Stamatikos, M. for the IceCube Collaboration, and D. Band,  AIP Conf. Proc. {\bf 836} (2006) 599 (astro-ph/0604281)

1282: \bibitem[Stern et al.~(2001)]{Stern} Stern, B.E. et al., ApJ $\bf{563}$ (2001) 80

1283: \bibitem[van Putten~(2004)]{Putten} van Putten, M., ApJ $\bf{611}$ $\bf{L}$81-84 (2004)

1284: \bibitem[Waxman~(2003)]{Waxman} Waxman, E., Nuc. Phys. B Proc. Supp. $\bf{118}$ (2003) 353-362 (and references therein)

1285: \bibitem[Waxman \& Bahcall~(1997)]{WaxBah} Waxman, E. and J. Bahcall PRL $\bf{78}$ (1997) 2292  

1286: 

1287: \end{thebibliography}

1288: 

1289: \clearpage

1290: 

1291: \begin{figure}

1292: %% \epsscale{.80}

1293: \epsscale{1.0}

1294: \plotone{f1.eps}

1295: %% \plotone{models_final_070125.eps}

1296: \caption{Predicted differential muon neutrino flux as a function of energy for four 

1297: different models of GRB neutrino production: the precursor model (solid line), the 

1298: canonical Waxman-Bahcall model (thick dotted line), the Murase-Nagataki model (thin dotted 

1299: line), and the supranova model (dot-dashed line).  All models include the effect of 

1300: ${\nu}$ oscillations.  The diffuse neutrino bounds determined from cosmic ray observations 

1301: with (upper horizontal line) and without (lower horizontal line) z evolution are also shown 

1302: for reference.

1303: \label{fig1}}

1304: \end{figure}

1305: 

1306: \begin{figure}

1307: %% \epsscale{0.80}

1308: \epsscale{1.0}

1309: \plotone{f2.eps}

1310: %% \plotone{stability_070213.eps}

1311: \caption{A stable period of detector activity, shown by the nearly Gaussian random temporal 

1312: distribution of events in each 10-second bin during the off-time period of BATSE GRB 

1313: 6610.  Initial selection criteria have been applied to these data, but the GRB-specific criteria 

1314: have not yet been applied.

1315: \label{fig2}}

1316: \end{figure}

1317: 

1318: \begin{figure}

1319: \epsscale{1.0}

1320: %% \epsscale{0.80}

1321: \plotone{f3.eps}

1322: %% \plotone{deltat_digitized.eps}

1323: \caption{Time difference ($\mathrm{{\delta}t}$) between subsequent events during the 

1324: background time period of a representative GRB, after application of initial data quality 

1325: cuts.  There is no evidence for significant gaps in the data that could produce a ``false 

1326: negative'' result.

1327: \label{fig3}}

1328: \end{figure}

1329: 

1330: \begin{figure}

1331: \epsscale{1.0}

1332: %% \epsscale{0.80}

1333: \plotone{f4.eps}

1334: %% \plotone{likelihood_digitized.eps}

1335: \caption{A comparison of the likelihood of track reconstruction, 

1336: ${\mathrm{{\pounds_{reco}}}}$ for observed data (solid line) and simulated background 

1337: events (dashed line).  Both curves are normalized after preliminary data selection criteria 

1338: are applied.  The close agreement signifies that our simulations are properly modeling the 

1339: observed events, thus providing additional evidence for the trustworthiness of the 

1340: simulated signal events as well.

1341: \label{fig4}}

1342: \end{figure}

1343: 

1344: 

1345: \begin{figure}

1346: \epsscale{1.0}

1347: %% \epsscale{0.80}

1348: \plotone{f5.eps}

1349: %% \plotone{delang_070412.eps}

1350: \caption{The expected distribution of angular mismatch ${\Delta}$${\Psi}_{1}$ for a simulated 

1351: muon neutrino spectrum (shaded region) and observed background (open region).  

1352: ${\Delta}$${\Psi}_{1}$ = 0 is the position of the burst determined from photon 

1353: observations.  Selecting events with ${\Delta}$${\Psi}_{1}$${\leq}$12$^\circ$ 

1354: retains more than $90\%$ of the signal events.

1355: \label{fig5}}

1356: \end{figure}

1357: 

1358: \begin{figure}

1359: \epsscale{1.0}

1360: %% \epsscale{0.80}

1361: \plotone{f6.eps}

1362: %% \plotone{aeff_070418_new.eps}

1363: \caption{Angle-averaged muon neutrino effective area for the AMANDA-II (years 2001-2003) 

1364: coincident search algorithm, based upon Monte Carlo simulations of expected signal 

1365: events from the Northern hemisphere.

1366: \label{fig6}}

1367: \end{figure}

1368: 

1369: \begin{figure}

1370: \epsscale{1.0}

1371: %% \epsscale{0.80}

1372: \plotone{f7.eps}

1373: %% \plotone{mrf_070205.eps}

1374: \caption{Relative Model Rejection Factor (MRF) as a function of angular mismatch 

1375: (${\Delta}$${\Psi}_{1}$) between the burst position and the reconstructed track, for the subset 

1376: of bursts from 2000.  The arrow indicates the mismatch angle selected for this analysis.

1377: \label{fig7}}

1378: \end{figure}

1379: 

1380: \begin{figure}

1381: \epsscale{1.00}

1382: %% \epsscale{0.80}

1383: \plotone{f8.eps}

1384: %% \plotone{batseipn_070125.eps}

1385: \caption{Duration distribution of BATSE GRBs (upper, open histogram) and 

1386: IPN bursts for which durations have been determined (lower, shaded 

1387: histogram).  Both distributions appear to be drawn from the same 

1388: underlying population.

1389: \label{fig8}}

1390: \end{figure}

1391: 

1392: \begin{figure}

1393: %% \epsscale{1.0}

1394: \epsscale{0.95}

1395: \plotone{f9.eps}

1396: %% \plotone{fluxlimits_070418.eps}

1397: \caption{AMANDA flux upper limits (solid lines) for muon neutrino energy spectra 

1398: predicted by the Waxman-Bahcall spectrum \citep{Waxman} (thick dashed line), the Razzaque 

1399: et al. spectrum \citep{Razzaq} (dot-dashed line) and the Murase-Nagataki spectrum 

1400: \citep{Murase} (thin dotted line).  The central 90\% of the expected flux for each model is 

1401: shown.  For the Waxman-Bahcall model we include both long- and short-duration bursts; for 

1402: the other spectra, only long-duration bursts are included.  Including short-duration bursts 

1403: would improve the flux upper limits by approximately 13\%.  While our analysis was 

1404: restricted to bursts located in the Northern Hemisphere (2${\pi}$ sr), all flux upper 

1405: limits are for the entire sky (4${\pi}$ sr).  

1406: \label{fig9}}

1407: \end{figure}

1408: 

1409: \begin{figure}

1410: %% \epsscale{0.80}

1411: \epsscale{1.0}

1412: \plotone{f10.eps}

1413: %% \plotone{green070418_line.eps}

1414: \caption{Green's Function Fluence Upper Limit for AMANDA's GRB analysis from 2000 to 2003.  

1415: This fluence upper limit can be folded into any desired spectrum to provide a flux upper 

1416: limit for that particular spectrum.

1417: \label{fig10}}

1418: \end{figure}

1419: 

1420: \clearpage

1421: 

1422: \begin{table}

1423: \begin{center}

1424: \caption{Primary Instruments in the Third Interplanetary Network, 1997-2003 \label{tab:a}}

1425: %% \tabletypesize{\scriptsize}

1426: \begin{tabular}{ccc}

1427: \tableline

1428: Instrument & Energy Range (keV) & Mission Homepage \\

1429: \hline

1430: BATSE LAD & 30 - 190 & http://www.batse.msfc.nasa.gov \\

1431: BeppoSAX GRBM & 40 - 700 & http://www.asdc.asi.it/bepposax \\

1432: BeppoSAX WFC & 2 - 26 & http://www.asdc.asi.it/bepposax \\

1433: HETE-II FREGATE & 6 - 400 & http://space.mit.edu/HETE/fregate.html \\

1434: HETE-II WXM & 2 - 25 & http://space.mit.edu/HETE/wxm.html \\

1435: HETE-II SXC & 2 - 14 & http://space.mit.edu/HETE/sxc.html \\

1436: %% HETE-II SXC & 0.5 - 10 & http://space.mit.edu/HETE/sxc.html \\

1437: INTEGRAL & 15 - 10000 & http://integral.esac.int/ \\

1438: Konus WIND & 12 - 10000 & http://www-spof.gsfc.nasa.gov/istp/wind/ \\

1439: Mars Odyssey & $\sim$100 - 8000 & http://mars.jpl.nasa.gov/odyssey/ \\

1440: NEAR XGRS & 100 - 1000 & http://near.jhuapl.edu \\

1441: %% NEAR GRS & 300 - 10000 & http://near.jhuapl.edu \\

1442: RHESSI & $\sim$25 - $\sim$25000 & http://hesperia.gsfc.nasa.gov/hessi \\

1443: %% RHESSI & $\sim$3 - $\sim$17000 & http://hesperia.gsfc.nasa.gov/hessi \\

1444: Ulysses & 25 - 150 & http://ulysses.jpl.nasa.gov \\

1445: \hline

1446: \end{tabular}

1447: \end{center}

1448: \end{table}

1449: 

1450: \clearpage

1451: 

1452: \begin{table}

1453: \begin{center}

1454: \caption{BATSE Triggered and IPN Bursts Per Year in the AMANDA Analysis, by Duration \label{tab:b}}

1455: %% \tabletypesize{\scriptsize}

1456: \begin{tabular}{cccccccc}

1457: \tableline

1458: Year & 1997 & 1998 & 1999 & 2000 & 2001 & 2002 & 2003 \\

1459: \hline

1460: N$_{\mathrm{Short}}$ & 12 & 15 & 9 & 7 & 1 & 1 & 2 \\

1461: N$_{\mathrm{Long}}$ & 51 & 50 & 61 & 77 & 15 & 21 & 24 \\

1462: N$_{\mathrm{Unknown}}$ & 15 & 29 & 26 & 3 & 0 & 0 & 0 \\

1463: \hline

1464: N$_{\mathrm{Total}}$ & 78 & 94 & 96 & 87 & 16 & 22 & 26 \\

1465: \hline

1466: \end{tabular}

1467: \end{center}

1468: \end{table}

1469: 

1470: \clearpage

1471: 

1472: \begin{table}

1473: \begin{center}

1474: \caption{Data Selection Criteria, Year by Year \label{tab:c}}

1475: %% \tabletypesize{\scriptsize}

1476: \begin{tabular}{cccccc}

1477: \tableline

1478: Criterion & 97-99 & 00 & 01-03 & Precursor\tablenotemark{a} \\

1479: \hline

1480: ${\Delta}$${\Psi}_1$, ${\delta}$${\geq}$10$^\circ$ (${\delta}$$<$10$^\circ$) & $<$20$^\circ$ ($<$6.5$^\circ$) & 

1481: $<$12.5$^\circ$ ($<$7$^\circ$) & $<$12$^\circ$ ($<$8$^\circ$) & $<$12$^\circ$ ($<$5$^\circ$) \\

1482: ${\Delta}$${\Psi}_2$, ${\delta}$${\geq}$10$^\circ$ (${\delta}$$<$10$^\circ$) & N/A & N/A & $<$12$^\circ$ ($<$8$^\circ$) & 

1483: $<$12$^\circ$ ($<$6$^\circ$) \\

1484: ${\Delta}$${\Psi}_3$, ${\delta}$${\geq}$10$^\circ$ (${\delta}$$<$10$^\circ$) & N/A & N/A & $<$16$^\circ$ ($<$8$^\circ$) & 

1485: $<$16$^\circ$ ($<$8$^\circ$) \\

1486: ${\Delta}$${\Psi}_4$, ${\delta}$${\geq}$10$^\circ$ (${\delta}$$<$10$^\circ$) & N/A & N/A & N/A & $<$40$^\circ$ 

1487: ($<$40$^\circ$) \\

1488: ${\sigma}_{\Psi}$\tablenotemark{b} & N/A & N/A & $<$5$^\circ$ ($<$5$^\circ$) & $<$5$^\circ$ ($<$5$^\circ$) \\

1489: Track Uniformity & N/A & $<$0.29 ($<$0.29) & $<$0.55 ($<$0.55) & $<$0.55 ($<$0.55) \\

1490: ${\pounds}_{reco}$\tablenotemark{c} & N/A & $<$7.85 ($<$7.5) & N/A & N/A \\

1491: Direct Hits & $>$10 & N/A & N/A & N/A \\

1492: $N_{OMs}$ in Event & N/A & N/A ($>$24) & N/A & N/A \\

1493: \hline

1494: Signal Passing Rate & 0.35 (0.22)& 0.69 (0.54) & 0.68 (0.61) & 0.69 \\

1495: %% Signal Passing Rate from L3 & 0.35 (0.22)& 0.77 (0.60) & 0.75 (0.67) & 0.77 \\

1496: \hline

1497: \end{tabular}

1498: \tablenotetext{a}{{The precursor time period was searched only during the 2001-2003 dataset.}}

1499: \tablenotetext{b}{{The angular resolution of the paraboloid fit.}}

1500: \tablenotetext{c}{{The log(Likelihood) of the reconstructed track.}}

1501: \end{center}

1502: \end{table}

1503: 

1504: \clearpage

1505: 

1506: \begin{table}

1507: \begin{center}

1508: \caption{Uncertainties/Corrections in the GRB Analysis \label{tab:d}}

1509: %% \tabletypesize{\scriptsize}

1510: \begin{tabular}{ccc}

1511: \tableline

1512: Source of Uncertainty & Quantity & Reference \\

1513: \hline

1514: OM sensitivity & ${\pm}$7\% & \citep{Reco} \\

1515: Simulation parameters (incl. ice properties) & ${\pm}$15\% & Sections 3.3 \& 3.4 \\

1516: Neutrino-nucleon cross-section & ${\pm}$3\% & \citep{Gandhi} \\

1517: Correction for IPN bursts not modeled & -3\% & Appendix A \\

1518: \hline

1519: Uncertainties added in quadrature & +16/-17\% & \\

1520: \hline

1521: Correction for short bursts not modeled & -6\% & Appendix A \\

1522: \hline

1523: Total & +15\%/-18\% & \\

1524: \hline

1525: \end{tabular}

1526: \end{center}

1527: \end{table}

1528: 

1529: \clearpage

1530: 

1531: \begin{table}

1532: \begin{center}

1533: \caption{Results of GRB Analysis 1997-2003 \label{tab:e}}

1534: %% \tabletypesize{\scriptsize}

1535: \begin{tabular}{cccccc}

1536: \tableline

1537: Year & 1997-1999 & 2000 & 2001-2003 & 2000-2003 & 1997-2003 \\

1538: \hline

1539: $N_{\mathrm{Bursts}}$ & 268 & 87 & 64 & 151 & 419 \\

1540: $N_{\mathrm{BG,Exp}}$ & 0.46 & 1.02 & 0.27 & 1.29 & 1.74 \\

1541: $N_{\mathrm{Obs}}$ & 0 & 0 & 0 & 0 & 0 \\

1542: Event Upper Limit & 1.98 & 1.50 & 2.30 & 1.30 & 1.10 \\

1543: $MRF_{\mathrm{WB}}$\tablenotemark{a} & 6.6 & 5.5 & 11 & 2.5 & 1.3 \\

1544: $MRF_{\mathrm{MN}}$\tablenotemark{b} & 4.9 & 3.1 & 6.2 & 1.4 & 0.82 \\

1545: $MRF_{\mathrm{Razz}}$\tablenotemark{c} & 2.4 & 1.5 & 3.0 & 0.68 & 0.40 \\

1546: \hline

1547: \end{tabular}

1548: \tablenotetext{a}{{Based on the flux of \citet{Waxman}, corrected for neutrino oscillations.}}

1549: \tablenotetext{b}{{Based on the flux of \citet{Murase}.}}

1550: \tablenotetext{c}{{Based on the ``supranova'' flux of \citet{Razzaq}.}}

1551: \end{center}

1552: \end{table}

1553: 

1554: \clearpage

1555: 

1556: \begin{table}

1557: \begin{center}

1558: \caption{Results of Precursor Search 2001-2003 \label{tab:f}}

1559: %% \tabletypesize{\scriptsize}

1560: \begin{tabular}{ccccc}

1561: \tableline

1562: Year & $N_{\mathrm{Bursts}}$ & $N_{\mathrm{BG,Exp}}$ & $N_{\mathrm{Obs}}$ & Event U.L. \\

1563: \hline

1564: 2001 & 15 & 0.06 & 0 & 2.38 \\

1565: 2002 & 21 & 0.07 & 0 & 2.37 \\

1566: 2003 & 24 & 0.07 & 0 & 2.37 \\

1567: \hline

1568: 2001-2003 & 60 & 0.20 & 0 & 2.30 \\

1569: \hline

1570: \end{tabular}

1571: \end{center}

1572: \end{table}

1573: 

1574: \end{document}

1575: 

1576: %%

1577: %% End of file `apj_grb.tex'.

1578: