A Likely Detection of a Two-planet System in a Low-magnification Microlensing Event

We report on the analysis of a microlensing event OGLE-2014-BLG-1722 that showed two distinct short term anomalies. The best fit model to the observed light curves shows that the two anomalies are explained with two planetary mass ratio companions to the primary lens. Although a binary source model is also able to explain the second anomaly, it is marginally ruled out by 3.1 $\sigma$. The 2-planet model indicates that the first anomaly was caused by planet "b" with a mass ratio of $q = (4.5_{-0.6}^{+0.7}) \times 10^{-4}$ and projected separation in unit of the Einstein radius, $s = 0.753 \pm 0.004$. The second anomaly reveals planet "c" with a mass ratio of $q_{2} = (7.0_{-1.7}^{+2.3}) \times 10^{-4}$ with $\Delta \chi^{2} \sim 170$ compared to the single planet model. Its separation has a so-called close-wide degeneracy. We estimated the physical parameters of the lens system from Bayesian analysis. This gives that the masses of planet b and c are $m_{\rm b} = 56_{-33}^{+51}\,M_{\oplus}$ and $m_{\rm c} = 85_{-51}^{+86}\,M_{\oplus}$, respectively, and they orbit a late type star with a mass of $M_{\rm host} = 0.40_{-0.24}^{+0.36}\,M_{\odot}$ located at $D_{\rm L} = 6.4_{-1.8}^{+1.3}\,\rm kpc$ from us. If the 2-planet model is true, this is the third multiple planet system detected by using the microlensing method, and the first multiple planet system detected in the low magnification events, which are dominant in the microlensing survey data. The occurrence rate of multiple cold gas giant systems is estimated using the two such detections and a simple extrapolation of the survey sensitivity of 6 year MOA microlensing survey (Suzuki et al. 2016) combined with the 4 year $\mu$FUN detection efficiency (Gould et al. 2010). It is estimated that $6 \pm 2\,\%$ of stars host two cold giant planets.