In the EGRET energy range, localized sources are seen against a background of diffuse galactic and isotropic radiation, with the galactic radiation dominating for all except the highest galactic latitudes. The first EGRET catalog ([Fichtel et al. 1994b]) used the galactic diffuse gamma radiation model of Bertsch et al. (1993b). Since that publication, several improvements have been made to this model. These are outlined here and described in detail by Hunter et al (1995):
1. For the results presented by Bertsch et al. (1993b), the HI and W
radio data were
deconvolved by integrating the radio data over a velocity interval determined
by the endpoints of each distance interval (
).
While this technique works well in general, it becomes degenerate near the
tangent point of each line-of-sight in the inner Galaxy where the velocity
interval approaches zero. This degeneracy has been eliminated by determining
the line-of-sight distance, or near/far distances in the inner Galaxy, which
correspond to the mean velocity of each velocity interval of the radio data.
2.
Within the longitude range 
, the HI or W
with 
(quadrant I) or 
, (quadrant IV), which lies kinematically just
outside the solar circle, is treated in the same way as material which lies
just inside the solar circle i.e. it is distributed partially at the
kinematic distance and partially near the sun.
3. The Galactic radius limit used in the model has been increased to 30 kpc.
4.
In Bertsch et al. (1993b), the entire velocity range of the composite CO survey of the Galaxy
(Dame et al. 1987) contained in the FITS files was included in the
deconvolution. This range of integration has been restricted to match the
velocity limits used in the preparation of the total CO column density maps.
5.
Hunter et al (1995) reevaluated the relative normalization of the three HI surveys used for
this calculation: the northern HI survey of Weaver and Williams (1973), WW;
the Maryland-Parks southern survey (Kerr et al. 1986), MP; and the
Leiden-Green Bank, LGB, survey (Burton & Liszt 1983; Burton 1985). The
normalization values were determined by calculating the integrated antenna
temperatures in overlapping regions of the surveys, 
,
and 
.
6.
In the Galactic center, 
, the HI radio spectra,
from the LGB survey, show evidence of absorption due to intervening material.
In particular, the spectrum at 
shows substantial
absorption. An approximate unabsorbed spectrum for 
is obtained by averaging the spectra at 
.
7. The CR electron spectrum of Fichtel et al. (1991) was used by Bertsch et al. (1993). Hunter et al (1995) replaced this electron spectrum with the spectrum calculated by Skibo (1993) which is substantially flatter than the Fichtel et al. spectrum below about 100 MeV. The electron bremsstrahlung production function for CR electrons interacting with the HII has been modified to reflect the lack of screening.
The diffuse gamma radiation consists of two components, one assumed to be isotropic and one which is dependent on direction, described by the Hunter et al (1995) model. Although this approach provides a good representation to the large-scale EGRET data, it is not necessarily ideal for every point in the sky. For this reason, the catalog analysis for a specific region (typically 15
radius about the point being considered) allows the two components to be independently scaled from their nominal values.