## 2. Correlation## 2.1. DataSince blazars are known to be strongly variable in
-rays, we use both maximum and average
-ray fluxes from Hartman et al.
(1999). For the maximum fluxes, we use only those with significance
level . For the averages, we use the
flux for the sum of all EGRET observation (denoted P1234 in Hartman et
al. 1999); for the cases in which P1234 has only an upper limit, half
of the (2) limit value was used. For
the emission line information, we used the data listed in the paper by
Cao & Jiang (1999) except for the marked items. The relevant data
are listed in Table 1, where Column 1 gives the name of the
source; Column 2, classification, FQ for flat spectrum radio quasar
and BL for BL Lacertae object; Column 3, the redshift; Column. 4 and
5, the maximum and the average -ray
flux in units of
photon cm
## 2.2. ResultThe observed photons are converted to flux densities at E GeV as follows. Let where is the normalization and
is the photon spectral index given
in Column 6. Integrating the above relation from 100 MeV to 10 GeV and
setting it equal the observed photon flux given in Column 4 or 5, we
obtain . We calculate the flux
density at 0.4 GeV, since that is about the average energy of the
photons. The flux density is k-corrected according to
, where
is the spectral index
( and
). Adopting H When the linear regression analysis is performed (excluding 3C 273) for the maximum -luminosities, a correlation is found, with a correlation coefficient and a chance probability . For the average -luminosities, a correlation is (excluding 3C 273 again) with a correlation coefficient and a chance probability . Fig. 1 shows the correlation for the average -luminosities; open circles are for flat spectrum radio quasars while the filled points for BL Lacertae objects. The solid line is the best fit.
© European Southern Observatory (ESO) 2000 Online publication: June 20, 2000 |