Predictions of polarized dust emission from interstellar clouds: spatial variations in the efficiency of radiative torque alignment

Context. Polarization carries information about the magnetic fields in interstellar clouds. The observations of polarized dust emission are used to study the role of magnetic fields in the evolution of molecular clouds and the initial phases of star formation. Therefore, it is important to understand how different cloud regions contribute to the observed polarized signal. Aims. We study the grain alignment with realistic simulations, assuming the radiative torques to be the main mechanism. The aim is to study the efficiency of the grain alignment as a function of cloud position and to study the observable consequences of these spatial variations. Methods. Our results are based on the analysis of model clouds derived from MHD simulations of super-Alfvenic magnetized turbulent flows. The continuum radiative transfer problem is solved with Monte Carlo methods to estimate the three-dimensional distribution of dust emission and the radiation field strength. The anisotropy of the radiation field is taken into account explicitly. We also examine the effect of grain growth in cores both to the observed polarization and to the inferred magnetic field. Results. Using the assumptions of Cho & Lazarian (2005, ApJ, 631, 361), our findings are generally consistent with their results. However, the anisotropy factor is lower than their assumption of gamma = 0.7, and thus radiative torques are less efficient. Compared with our previous paper, P/I relations are steeper. Without grain growth, the magnetic field of the cores is poorly recovered above a few AV. If grain size is doubled, the polarized dust emission can trace the magnetic field lines possibly up to A(V) similar to 10 mag. However, many of the prestellar cores may be too young for grain coagulation to play a major role. The inclusion of direction-dependent radiative torque efficiency weakens the alignment. Even with doubled grain size, we would not expect to probe the magnetic field beyond a few magnitudes in AV.