In Stern et al. (2015), we calculated the expected spectroastrometric signal (variation of photon center with wavelength) of the quasar broad line region (BLR). The angular size of the BLR is no more than ~100μas, a factor of ~1000 below the resolution of diffraction-limited observations on 8m telescopes. We demonstrated that the signal for luminous quasars at redshift ~2 is detectable with existing telescopes. A detection of the BLR spectroastrometric signal would be the first direct verification that the BLR has an ordered velocity field, and would provide a new way to measure BLR sizes and black hole masses in luminous high-redshift quasars. I recieved the MPIA Patzer prize for this work.
Figure 1: A simulation of the photon centroid across the broad Hα emission line in a luminous z~2 quasar. The BLR is assumed to originate from a rotating disk, as suggested by indirect observations. The top panel shows the spectrum (y-axis is the photon flux). Prominent narrow emission lines are marked. The second panel shows the expected offset of the photon centroid relative to the centroid at continuum wavelengths. The large scale narrow lines are expected to show a large offset, and hence velocities with narrow line emission need to be masked in order to measure the BLR signal. The two bottom panels show the expected signal with 10 hours on an 8m telescope using LGS-AO. The panels differ in the assumed distribution of BLR gas as a function of distance from the black hole.
Figure 2: The expected strength of the BLR spectroastrometric signal for the most luminous SDSS quasar at each redshift, with 10 hours on an 8m telescope (small symbols) or 1 hour on a 30m telescope (large symbols). The different markers denote different assumptions on the distribution of gas as a function of distance from the black hole (see two bottom panels in Figure 1). The signal should be detectable at z ~ 2 with existing telescopes, and at 0 ≲ z ≲ 6 with next generation telescopes.
I have applied for and been awarded observing time on both Gemini and the VLT, for a pilot observation of two quasars using LGS-AO. Some of the observations have already been carried out, while some are scheduled for early 2017. I am currently analyzing the existing data.
My collaborators (J. F. Hennawi and J.-U. Pott) are members of the science team of the AO-spectrometer on the E-ELT (Davies et al. 2010), for which BLR spectroastrometry is a key science case. This project thus constitutes an important pathfinder that can influence the development and design of AO spectrometers on 30m-class telescopes.