Comparing optical and radio polarization angles of the quasar 1954+513
Left: Rotation Measure (RM) of 1954+513. The leftmost red dot indicates electric vector position angle in optical, whilst the black dots indicate radio measurements at 12, 15, 22, 24 and 45 GHz. Errors are smaller than the size of the dots. Right: Core shift (drift of the VLBI core position with frequency) of 1954+513.
As both radio an optical emission from AGN arises from synchrotron mechanisms, a correlation between the properties at these bands would be expected if we are probing the same emitting region. However, when studying the simultaneous polarization angles, it seems that this is not the case on all sources. One example of this is the flat spectrum quasar 1954+513, where the optical value for the polarization angle does not correspond to the expected one from extrapolation of radio rotation measure (RM) fit (left panel, straight blue line). Such apparent discrepancy might be explained if we assume that the RM increases with decreasing wavelengths, as possibly indicated by the outlier at 43 GHz (left panel, dashed blue line). This is expected if we take into account core-shift effects, which suggests that at higher frequencies the VLBI core actually probes into regions upstream the jet. Such core shift is indeed found in this source (right panel). Furthermore, the core shift suggests a value of K_r close to unity, compatible with equipartition regimes, and a magnetic field at one parsec from the central engine of the order of few Gauss