Variability and stability in blazar jets on time-scales of years: optical polarization monitoring of OJ 287 in 2005-2009
OJ 287 is a BL Lac object at redshift z = 0.306 that has shown double-peaked bursts at regular intervals of similar to 12 yr during the last similar to 40 yr. We analyse optical photopolarimetric monitoring data from 2005 to 2009, during which the latest double-peaked outburst occurred. The aim of this study is twofold: firstly, we aim to analyse variability patterns and statistical properties of the optical polarization light curve. We find a strong preferred position angle in optical polarization. The preferred position angle can be explained by separating the jet emission into two components: an optical polarization core and chaotic jet emission. The optical polarization core is stable on time-scales of years and can be explained as emission from an underlying quiescent jet component. The chaotic jet emission sometimes exhibits a circular movement in the Stokes plane. We find six such events, all on the time-scales of 10-20 d. We interpret these events as a shock front moving forwards and backwards in the jet, swiping through a helical magnetic field. Secondly, we use our data to assess different binary black hole models proposed to explain the regularly appearing double-peaked bursts in OJ 287. We compose a list of requirements a model has to fulfil to explain the mysterious behaviour observed in OJ 287. The list includes not only characteristics of the light curve but also other properties of OJ 287, such as the black hole mass and restrictions on accretion flow properties. We rate all existing models using this list and conclude that none of the models is able to explain all observations. We discuss possible new explanations and propose a new approach to understanding OJ 287. We suggest that both the double-peaked bursts and the evolution of the optical polarization position angle could be explained as a sign of resonant accretion of magnetic field lines, a 'magnetic breathing' of the disc.