Integrated Polarization of Sources at lambda ~1m and New Rotation Measure Ambiguities

We present an analysis of the polarization of compact radio sources from six pointings of the Westerbork Synthesis Radio Telescope (WSRT) at 350 MHz with 35% coverage in lambda^2. After correcting for the off-axis instrumental polarization with a simple analytical model, only a small number of 585 strong sources have significant polarizations at these wavelengths. The median depolarization ratio from 1.4 GHz for the strongest sources is <0.2, reinforcing the likelihood that radio galaxies are found in magnetized environments, even outside of rich clusters. Seven sources with significant 350 MHz polarization were selected for a more in-depth Faraday structure analysis. We fit the observed values Q/I and U/I as a function of lambda^2 using both a depolarizing screen and two component models. We also performed RM Synthesis/Clean and standard fitting of polarization angle vs. lambda^2. We find that a single rotation measure (RM), as found using polarization angle fitting or simple screen models, commonly provides a poor fit when the solutions are translated back into Q, U space. Thus, although a single "characteristic" rotation measure may be found using these techniques, the Faraday structure of the source may not be adequately represented. We also demonstrate that RM Synthesis may yield an erroneous Faraday structure in the presence of multiple, interfering RM components, even when cleaning of the Faraday spectrum is performed. We briefly explore the conditions under which rotation measures and Faraday structure results can be reliable. Many measurements in the literature do not meet these criteria; we discuss how these influence the resulting scientific conclusions and offer a prescription for obtaining reliable RMs.