We present the analysis of high-resolution images of MOA-2013-BLG-220 , taken with the Keck adaptive optics system 6 years after the initial observation , and identify the lens star as a solar type star hosting a super-Jupiter mass planet .
The masses of planets and host stars discovered by microlensing are often not determined from the light curve data , while the star-planet mass ratio and projected separation in units of Einstein ring radius are well measured .
High resolution follow-up observations , after the lensing event is over , can resolve the source and lens , determining their fluxes as well as the amplitude and direction of relative proper motion .
This places strong constraints on the system ’ s physical parameters .
Due to the high relative proper motion of this event , { \mbox { \boldmath$ \bf \mu$ } } _ { rel,G } = 12.6 \pm 0.3 mas/yr , we were able to resolve the source and lens with a separation of 77.6 \pm 0.4 mas .
We measure the K band lens flux to be K _ { L } = 17.92 \pm 0.05 .
By combining constraints from the light curve , the AO flux , and relative source-lens proper motion , we find the lens star to have a mass of M _ { L } = 0.89 \pm 0.05 M _ { \odot } located at D _ { L } = 6.96 \pm 0.60 kpc ; this implies that it is very likely to be in the Galactic bar .
With a mass ratio of q = ( 3.26 \pm 0.04 ) \times 10 ^ { -3 } from reanalysis of the light-curve , the planet is found to have a mass of M _ { P } = 3.06 \pm 0.18 M _ { J } and a separation of r _ { \perp } = 2.69 \pm 0.23 AU .
This mass is much higher than the prediction from a Bayesian analysis with a uniform prior on the mass ratio distribution with host star mass , suggesting that planets with higher mass ratios , of order 10 ^ { -3 } or greater , are more common orbiting higher-mass stars .
This demonstrates the importance of follow-up high angular resolution observations .
It will be possible to measure the metallicity of the host star with VLT MUSE observations in the near future .