Person:

Rao, Ramprasad

The millimeter/submillimeter wavelength polarization of Sgr A* is known to be variable in both magnitude and position angle on time scales down to a few hours. The unstable polarization has prevented measurements made at different frequencies and different epochs from yielding convincing measurements of Faraday rotation in this source. Here we present observations made with the Submillimeter Array polarimeter at 227 and 343 GHz with sufficient sensitivity to determine the rotation measure at each band without comparing position angles measured at separate epochs. We find the 10-epoch mean rotation measure to be ( − 5 . 6 ± 0 .7) × 10 5 rad m − 2; the measurements are consistent with a constant value. We conservatively assign a 3 σ upper limit of 2 × 10 5 rad m − 2 to rotation measure changes, which limits accretion rate fluctuations to 25%. This rotation measure detection limits the accretion rate to less than 2 × 10 − 7 M⊙ yr − 1 if the magnetic field is near equipartition, ordered, and largely radial, while a lower limit of 2 ×10 − 9 M⊙ yr − 1 holds even for a sub-equipartition, disordered, or toroidal field. The mean intrinsic position angle is 167 ◦ ± 7 ◦ and we detect variations of 31+18 −9 degrees. These variations must originate in the submillimeter photosphere, rather than arising from rotation measure changes.

Using the Submillimeter Array, we have made the first high angular resolution measurements of the linear polarization of Sagittarius A* at submillimeter wavelengths, and the first detection of intra-day variability in its linear polarization. We detected linear polarization at 340 GHz (880 µm) at several epochs. At the typical resolution of 1 . ′′ 4 × 2. ′′2, the expected contamination from the surrounding (partially polarized) dust emission is negligible. We found that both the polarization fraction and position angle are variable, with the polarization fraction dropping from 8.5% to 2.3% over three days. This is the first significant measurement of variability in the linear polarization fraction in this source. We also found variability in the polarization and total intensity within single nights, although the relationship between the two is not clear from these data. The simultaneous 332 and 342 GHz position angles are the same, setting a one-sigma rotation measure (RM) upper limit of 7 ×10 5 rad m − 2 . From position angle variations and comparison of “quiescent” position angles observed here and at 230 GHz we infer that the RM is a few ×10 5 rad m − 2, a factor of a few below our direct detection limit. A generalized model of the RM produced in the accretion flow suggests that the accretion rate at small radii must be low, below 10 − 6 − 10 − 7 M⊙ yr − 1 depending on the radial density and temperature profiles, but in all cases below the gas capture rate inferred from X-ray observations.