Catalog Construction

Once the best position for each candidate source was determined, a new analysis of each viewing period or summed map was made, calculating the and the flux or upper limit at the fixed position. In each case, a simultaneous analysis of all the sources in the list was done, so that a flux or upper limit could be calculated for each source for each observation. For the entries in the catalog, the E>100 MeV maps were used. This energy selection allows direct comparison with previous EGRET, COS-B and SAS-2 results. With a few exceptions, the E>100 MeV flux was calculated assuming a photon power law with a spectral index of 2.0 () for the source point spread function. For the three brightest sources, the measured power law was used because of possible effects on nearby sources (Vela: 1.7; Geminga 1.5, Crab 2.1). The energy spectra shown in the catalog are taken largely from previous EGRET publications (or those currently in the publication process), which discuss the details of source identifications, time variability, and energy spectra.

For each source, the excess above the modeled diffuse radiation was compared to the calibrated EGRET point spread function. Most of the sources were well fit by the PSF (reduced 1.5). Those with poorer fits (reduced between 2.5 and 3.5) are sources which may be confused by nearby sources or those which show possible indications of being extended emission or multiple sources. Such sources are identified in the catalog.

The highly structured, intense diffuse galactic emission along the galactic plane makes sources within about more subject to systematic uncertainties than those at higher latitudes. For this reason, we have adopted the same procedure used in the first EGRET catalog of including sources with at least one detection with 4 in the catalog for but demanding at least one detection with 5 for . The exact choice of the TS level and the latitude for the transition from one level to the other is somewhat arbitrary: the influence of the galactic diffuse radiation does not show an abrupt change.