Radio observational constraints on Galactic 3D-emission models

Context. Our position inside the Galaxy requires 3D-modelling to obtain the distribution of the Galactic magnetic field, cosmic-ray (CR) electrons and thermal electrons. Aims. Our intention is to find a Galactic 3D-model which agrees best with available radio observations. Methods. We constrain simulated all-sky maps in total intensity, linear polarization, and rotation measure (RM) by observations. For the simulated maps as a function of frequency we integrate in 15' wide cones the emission along the line of sight calculated from Galactic 3D-models. We test a number of large-scale magnetic field configurations and take the properties of the warm interstellar medium into account. Results. From a comparison of simulated and observed maps we are able to constrain the regular large-scale Galactic magnetic field in the disk and the halo of the Galaxy. The local regular field is 2 mu G and the average random field is about 3 mu G. The known local excess of synchrotron emission originating either from enhanced CR electrons or random magnetic fields is able to explain the observed high-latitude synchrotron emission. The thermal electron model (NE2001) in conjunction with a proper filling factor accounts for the observed optically thin thermal emission and low frequency absorption by optically thick emission. A coupling factor between thermal electrons and the random magnetic field component is proposed, which in addition to the small filling factor of thermal electrons increases small-scale RM fluctuations and thus accounts for the observed depolarization at 1.4 GHz. Conclusions. We conclude that an axisymmetric magnetic disk field configuration with reversals inside the solar circle fits available observations best. Out of the plane a strong toroidal magnetic field with different signs above and below the plane is needed to account for the observed high-latitude RMs. The large field strength is a consequence of the small thermal electron scale height of 1 kpc, which also limits the CR electron extent up to a height of I kpc not to contradict with the observed synchrotron emission out of the plane. Our preferred 3D-model fits the observed Galactic total intensity and polarized emission better than other models over a wide frequency range and also agrees with the observed RM from extragalactic sources.