Introduction

The search for astrophysical sources of high energy neutrinos is one of the central missions for the AMANDA detector. In this report, we concentrate on point sources with continuous (``DC'') output. Most theoretical models of potential astrophysical sources predict that the neutrino energy spectrum is very hard, approximately E^-2. Therefore, the most probable energy of a detected neutrino is well above 1 TeV. The analysis procedure described in this document utilizes two essential characteristics of the signal to reduce background. First, the sources are assumed to be points in the sky, so only events within a selected angular bin are considered. Second, we indirectly use the characteristics of neutrino signal with hard energy spectra to reject atmospheric neutrino and poorly reconstructed atmospheric muons, which have softer spectra. For example, we show that that the energy estimator based on hit probabilities provides a strong discriminant between signal and some important background topologies. The cut optimization presented is the first iteration in developing the procedures for the optimization of cut parameters. While optimized for E^-2<> spectra, as shown in section 3, these procedures produce reasonable acceptance for softer spectra.

The point source search differs from the atmospheric neutrino search in several ways. First, by concentrating on harder spectra, the effective area of the detector can be increased by relaxing the background rejection criteria and allowing relatively large background contamination of down-going muons. In the AMANDA architecture, there is an inverse correlation between rejection factor and effective area (or effective volume). The point source analysis tolerates a larger background rate than the atmospheric neutrino analysis, and therefore the effective area of the detector can be increased, especially for events with zenith angles more than 20 degrees from vertical. Due to the correlation between rejection and sensitivity, we have based the analysis on the optimization for Signal to Noise, rather than signal purity. Also because of the relaxed rejection requirements, our final selection cuts on analysis variables are weaker than previously published for atmospheric neutrino studies. Therefore, the statistical precision required by the background simulation is proportionality relaxed. Point searches require an assessment of the angular resolution. Since this analysis searches for point sources and not for diffuse sources (such as atmospheric neutrinos), sources are revealed by a statistically significant angular clustering of events. Hence the experimental data is reduced fewer than 10 events per angular bin. The probability distributions for random fluctuation of background counts are computed from Poisson statistics. Lastly, since the goal is to optimize on signal to noise rather than signal purity, the selection criteria used for this analysis may introduce different systematic errors than the atmospheric neutrino analysis.

An energy estimator is a valuable tool to reject atmospheric neutrino background. However at this stage, the energy estimator is primarily used as a topological cut, as explained in section 3. This analysis incorporates a selection criteria which significantly improves signal to noise for high energy neutrinos. Further work is planned to develop a reliable energy estimator.