1 Neutrino astronomy with MACRO

Besides high energy gamma ray production resulting from $\pi^{0}$ decay, astrophysical beam dump models predict neutrino emission from $\pi^{\pm}$ decay. Mesons are produced by accelerated protons interacting with matter or photons in an accretion disk (Gaisser [1995]). The discovery of TeV gamma-ray emissions has enhanced the potential possibilities of this mechanism and the possible existence of such sources, but energies are not high enough to exclude syncroton radiation or bremsstrahlung and inverse Compton production mechanisms. Neglecting photon absorption, it is expected that neutrino fluxes are almost equal to gamma ray ones and that the spectrum has the typical form due to the Fermi acceleration mechanism: $\frac{{\rm d}N}{{\rm d}E} \propto E^{-2.0 \div 2.2}$. Neutrinos produced in atmospheric cascades are background to the search for astrophysical neutrinos for energies $\lesssim 10$ TeV. In fact, atmospheric neutrinos have a softer spectrum than astrophysical neutrinos, since at energies $\gtrsim 100$ GeV the decay length of mesons in the atmosphere becomes longer than atmospheric depth and the spectrum steepens (differential spectral index $\gamma\simeq$ 3.7). GRBs are possible sources of high energy $\nu$s: in the fireball scenario the beam dump mechanism can lead to $\nu$ emission (Halzen [1996]; Waxman [1997]). The short time duration of the GRB emission allows to reduce the atmospheric neutrino background thanks to the additional requirement on time coincidence with the GRB emission besides the directional coincidence. Hence, although the $\nu$ expected fluxes from GRBs are much lower than the atmospheric $\nu$ flux, the time and directional association between GRBs and neutrinos can reach interesting sensitivities in underground detectors.

The MACRO detector, located in the Hall B of the Gran Sasso underground laboratories, with a surface of 76.6 12 m² and a height of 9 m, can indirectly detect neutrinos using a system of $\sim 600$ ton of liquid scintillator to measure the time of flight of particles (resolution $\sim 500$ ps) and $\sim 20000$ m² of streamer tubes for tracking (angular resolution better than 1$^{\circ}$ and pointing accuracy checked using moon shadow detection (Ambrosio [1998a]). The time of flight technique allows the discrimination between downward-going atmospheric muons and upward-going events produced in the rock below (average atmospheric neutrino energy $\langle E_{\nu} \rangle \sim 100$ GeV) and inside ( $\langle E_{\nu} \rangle \sim 4$ GeV) the detector by neutrinos which have crossed the Earth. Between $\sim 31~10^{6}$ atmospheric muons, a sample of 909 upward-going muons is selected with an automated analysis. The data taking has begun since March 1989 with the incomplete detector (Ahlen [1995]) and since April 1994 with the full detector (Ambrosio [1998b]). In our convention $1/\beta = \Delta T c/L$, calculated from the measured time of flight $\Delta t$ and the track length between the scintillator layers, is $\sim 1$ for downward-going muons and $\sim -1$ for upward-going muons. Events with $-1.25 < 1/\beta <-0.75$ are selected.

We look for a statistically significant excess of $\nu$ events in the direction of known $\gamma$ and X-ray sources (a list of 40 selected sources, 129 sources of the $2^{{\rm nd}}$ Egret Catalogue, 2233 Batse GRB, 220 SN remnants, 7 sources with $\gamma$ emission above 1 TeV) with respect to fluctuations of the atmospheric $\nu$ background. The expected background from atmospheric $\nu$s is calculated in declination bands of $\pm 5^{\circ}$ around the declination of the source mixing for 100 times local coordinates and times of upward-going $\mu$s. We calculate flux limits in half-cones of $3^{\circ}$taking into account reduction factors due to the signal fraction lost outside the cone (which depends on $\nu$ spectrum, kinematics of CC interaction, $\mu$ propagation in the rock, MACRO angular resolution). We do not find any signal evidence from known sources or of clustering of events (we measure 89 clusters of $\ge 3$ events and expect 81.2 of them in a cone of $3^{\circ}$). Muon flux limits for some sources are: 2.5 for Crab Nebula, 5.6 for MRK 421, 3.71 for Her X-1, 0.45 for Vela Pulsar, 0.65 for SN 1006 in units of $\times 10^{-14}$ cm^-2 s^-1. For most of the considered sources MACRO gives the best flux limits compared to other underground experiments.