We present first results on polarization swings in optical emission of blazars obtained by RoboPol, a monitoring programme of an unbiased sample of gamma-ray bright blazars specially designed for effective detection of such events. A possible connection of polarization swing events with periods of high activity in gamma rays is investigated using the data set obtained during the first season of operation. It was found that the brightest gamma-ray flares tend to be located closer in time to rotation events, which may be an indication of two separate mechanisms responsible for the rotations. Blazars with detected rotations during non-rotating periods have significantly larger amplitude and faster variations of polarization angle than blazars without rotations. Our simulations show that the full set of observed rotations is not a likely outcome (probability ≤ 1.5 × 10 −2 ) of a random walk of the polarization vector simulated by a multicell model. Furthermore, it is highly unlikely (∼ 5 × 10 −5 ) that none of our rotations is physically connected with an increase in gamma-ray activity.

We present measurements of rotations of the optical polarization of blazars during the second year of operation of RoboPol, a monitoring programme of an unbiased sample of gamma-ray bright blazars specially designed for effective detection of such events, and we analyse the large set of rotation events discovered in two years of observation. We investigate patterns of variability in the polarization parameters and total flux density during the rotation events and compare them to the behaviour in a non-rotating state. We have searched for possible correlations between average parameters of the polarization-plane rotations and average parameters of polarization, with the following results: (1) there is no statistical association of the rotations with contemporaneous optical flares; (2) the average fractional polarization during the rotations tends to be lower than that in a non-rotating state; (3) the average fractional polarization during rotations is correlated with the rotation rate of the polarization plane in the jet rest frame; (4) it is likely that distributions of amplitudes and durations of the rotations have physical upper bounds, so arbitrarily long rotations are not realised in nature.

We describe the data reduction pipeline and control system for the RoboPol project. The RoboPol project is monitoring the optical R-band magnitude and linear polarization of a large sample of active galactic nuclei that is dominated by blazars. The pipeline calibrates and reduces each exposure frame, producing a measurement of the magnitude and linear polarization of every source in the 13 × 13 field of view. The control system combines a dynamic scheduler, realtime data reduction, and telescope automation to allow high-efficiency unassisted observations.

We present first results from RoboPol, a novel-design optical polarimeter operating at the Skinakas Observatory in Crete. The data, taken during the May -June 2013 commissioning of the instrument, constitute a single-epoch linear polarization survey of a sample of gamma-ray-loud blazars, defined according to unbiased and objective selection criteria, easily reproducible in simulations, as well as a comparison sample of, otherwise similar, gamma-ray-quiet blazars. As such, the results of this survey are appropriate for both phenomenological population studies and for tests of theoretical population models. We have measured polarization fractions as low as 0.015 down to R magnitude of 17 and as low as 0.035 down to 18 magnitude. The hypothesis that the polarization fractions of gamma-ray-loud and gamma-ray-quiet blazars are drawn from the same distribution is rejected at the 3σ level. We therefore conclude that gamma-ray-loud and gamma-ray-quiet sources have different optical polarization properties. This is the first time this statistical difference is demonstrated in optical wavelengths. The polarization fraction distributions of both samples are well-described by exponential distributions with averages of p = 6.4 +0.9 −0.8 ×10 −2 for gamma-ray-loud blazars, and p = 3.2 +2.0 −1.1 × 10 −2 for gamma-ray-quiet blazars. The most probable value for the difference of the means is 3.4 +1.5 −2.0 × 10 −2 . The distribution of polarization angles is statistically consistent with being uniform.

We present the design and performance of RoboPol, a four-channel optical polarimeter operating at the Skinakas Observatory in Crete, Greece. RoboPol is capable of measuring both relative linear Stokes parameters q and u (and the total intensity I) in one sky exposure. Though primarily used to measure the polarization of point sources in the R-band, the instrument features additional filters (B, V and I), enabling multi-wavelength imaging polarimetry over a large field of view (13.6 ′ × 13.6 ′ ). We demonstrate the accuracy and stability of the instrument throughout its five years of operation. Best performance is achieved within the central region of the field of view and in the R band. For such measurements the systematic uncertainty is below 0.1% in fractional linear polarization, p (0.05% maximum likelihood). Throughout all observing seasons the instrumental polarization varies within 0.1% in p and within ∼1 • in polarization angle.