MOA-2009-BLG-319Lb: A Sub-Saturn Planet Inside the Predicted Mass Desert.

We present an adaptive optics (AO) analysis of images from the Keck-II telescope NIRC2 instrument of the planetary microlensing event MOA-2009-BLG-319. The $\sim$10 year baseline between the event and the Keck observations allows the planetary host star to be detected at a separation of $66.5\pm 1.7\,$mas from the source star, consistent with the light curve model prediction. The combination of the host star brightness and light curve parameters yield host star and planet masses of M_host = 0.514 $\pm$ 0.063M_Sun and m_p = 66.0 $\pm$ 8.1M_Earth at a distance of $D_L = 7.0 \pm 0.7\,$kpc. The star-planet projected separation is $2.03 \pm 0.21\,$AU. The planet-star mass ratio of this system, $q = (3.857 \pm 0.029)\times 10^{-4}$, places it in the predicted "planet desert" at $10^{-4} < q < 4\times 10^{-4}$ according to the runaway gas accretion scenario of the core accretion theory. Seven of the 30 planets in the Suzuki et al. (2016) sample fall in this mass ratio range, and this is the third with a measured host mass. All three of these host stars have masses of 0.5 $\leq$ M_host/M_Sun $\leq$ 0.7, which implies that this predicted mass ratio gap is filled with planets that have host stars within a factor of two of 1M_Sun. This suggests that runaway gas accretion does not play a major role in determining giant planet masses for stars somewhat less massive than the Sun. Our analysis has been accomplished with a modified DAOPHOT code that has been designed to measure the brightness and positions of closely blended stars. This will aid in the development of the primary method that the Nancy Grace Roman Space Telescope mission will use to determine the masses of microlens planets and their hosts.