Abstract In this paper, the multiwavelength data from radio to X-ray bands for 2709 blazars in the 4FGL-DR3 catalog are compiled to calculate their spectral energy distributions using a parabolic equation log ( ν f ν ) = P 1 log ν − P 2 2 + P 3 . Some important parameters including spectral curvature ( P 1 ), synchrotron peak frequency ( P 2 , log ν p ), and peak luminosity ( log L p ) are obtained. Based on those parameters, we discussed the classification of blazars using the “Bayesian classification” and investigated some mutual correlations. We came to the following results. (1) Based on the Bayesian classification of synchrotron peak frequencies, the 2709 blazars can be classified into three subclasses, i.e., log ( ν p / Hz ) < 13.7 for low synchrotron peak blazars (LSPs), 13.7 < log ( ν p / Hz ) < 14.9 for intermediate synchrotron peak blazars (ISPs), and log ( ν p / Hz ) > 14.9 for high synchrotron peak blazars (HSPs), and there are 820 HSPs, 750 ISPs, and 1139 LSPs. (2) The γ -ray emission has the closest relationship with radio emission, followed by optical emission, while the weakest relationship is that with X-ray emission. The γ -ray luminosity is also correlated with the synchrotron peak luminosity. (3) There are strong positive correlations between the curvature (1/∣ P 1 ∣) and the peak frequency ( log ν p

	Year	Citations

About