A free-floating-planet microlensing event caused by a Saturn-mass object

Abstract

A population of free-floating planets is known from gravitational microlensing surveys. None have a directly measured mass, owing to a degeneracy with the distance, but the population statistics indicate that many are less massive than Jupiter. We report a microlensing event -- KMT-2024-BLG-0792/OGLE-2024-BLG-0516, which was observed from both ground- and space-based telescopes -- that breaks the mass-distance degeneracy. The event was caused by an object with 0.219^{+0.075}_{-0.046} Jupiter masses that is either gravitationally unbound or on a very wide orbit. Through comparison with the statistical properties of other observed microlensing events and predictions from simulations, we infer that this object likely formed in a protoplanetary disk (like a planet), not in isolation (like a brown dwarf), and dynamical processes then ejected it from its birth place, producing a free-floating object.

Figures (10)

• Figure 1: Light curves of the KMT-2024-BLG-0792/OGLE-2024-BLG-0516 microlensing event compared with the best-fitting model.(A) Circles (with $1\,\sigma$ error bars) indicate observations from the OGLE telescope in Chile (black open circle), a KMTNet telescope in South Africa (blue square), a KMTNet telescope in Australia (orange diamond), and the Gaia spacecraft at L2 (red solid circle). Their source magnification is plotted as a function of time (t), relative to the time of peak magnification determined from the ground-based data ($t_{0, \rm ground}=2460434.323$ barycentric Julian date) (table \ref{['tab:parameters']}). The solid lines indicate the best-fitting microlensing models of the ground-based data (black) and Gaia data (red); the latter reaches its peak approximately 1.9 hours later than the former. (B) Same as (A), but showing an expanded view spanning 10 days. The magenta box indicates the region shown in A.
• Figure 2: Illustration of the space-based microlens parallax effect.(A) Three-dimensional schematic diagram (not to scale) illustrating the geometric configuration of the source, lens and observers. The lens (blue dot) and its Einstein ring (blue circle) are located on the lens plane (gray surface), which we treat as fixed in space. The source (orange dot) moves relative to the lens, in the direction of the orange arrow. Observers located at Earth (black dot) and Gaia (red dot) see different apparent source trajectories (black and red solid lines) projected on the lens plane, with dashed lines indicating their sightlines at two illustrative epochs. The Earth-Gaia vector (magenta arrow) projected on the observer plane (beige surface) causes a corresponding shift (cyan arrow) between the apparent source trajectories on the lens plane. (B) The corresponding model light curves for Earth (black line) and Gaia (red line), as in Fig. \ref{['fig:lc']}. Numbered dots on both curves indicate the Gaia observation epochs. (C) Model source trajectory with respect to the lens (blue dot) projected onto the lens plane as seen from Earth (black line) and Gaia (red line) for this event. The source angular position vectors $\hbox{\boldmath$u$}$ are normalized to the radius of Einstein ring (blue circle), $\theta_{\rm E}$. The axes $u_\parallel$ and $u_\perp$ are defined as parallel and perpendicular, respectively, to the source motion's direction (orange arrow). The compass indicates the north and east directions. The numbered positions (solid dots) correspond to the Gaia observation epochs. (D) Zoomed view of the region in the black box in (C), with cyan arrows indicating the shift attributable to the microlens parallax effect.
• Figure 3: Cumulative distribution of Einstein radius $(\theta_{\rm E})$ and the Einstein desert.(A) Histogram (dark green) of the cumulative distribution of $\theta_{\rm E}$ from previous microlensing observations [the KMTNet FSPL sample Gould22]. The apparent lack of events between 9 and 25 $\upmu{\rm as}$ is the Einstein desert (blue shaded region). (B) The distribution of all published microlensing FFP events with measured $\theta_{\rm E}$ob121323ob190551ob161540ob161928Ryu21kb192073Koshimoto23kb232669, which are all located below the desert. KMT-2024-BLG-0792/OGLE-2024-BLG-0516 has $\theta_{\rm E}=18.6\pm0.9\,\,\upmu{\rm as}$ (red vertical line in both panels, with red-shaded $1\,\sigma$ uncertainty region), within the desert.
• Figure S1: Systematic corrections to the Gaia Science Alerts data.A Difference in G-band magnitudes with $1\,\sigma$ error bars as a function of $\zeta$ after subtracting the best-fitting $\psi$-trends for FoVP (black) and FoVF (red). The solid lines are linear models fitted to the data, incorporating an additional Gaussian component to capture the dip feature in FoVP. B Difference in G-band magnitudes as a function of $\psi$ after subtracting the best-fitting models in panel A. C & D Residuals between the data and the corresponding models in panels A and B, respectively. E & F Same panel as A, but plotted separately for FoVP and FoVF, with numbered shaded bands indicating the CCD row numbers of the data.
• Figure S2: OGLE-Gaia color-magnitude diagram and color-color relation.A Gaia $G$ vs. OGLE-III $(V-I)$ CMD for all stars with $G<18$ within $150^{\prime\prime}$ of the microlensing event. The magenta circle indicates the source star, with its G magnitude taken from the Gaia catalog GaiaDR3 and its OGLE color derived from our FSPL model. B Color-color plot for all stars (black points) shown in panel A. C Same as panel B, but restricted to $G<16$ (blue points). In panels B and C, the red line is a color-color linear relation fitted to the $2.4<(V-I)<2.8$ region. The red dot is the value of this relation at the source $(V-I)$ color. The cyan crosses are outliers rejected at $>3\,\sigma$. We find that the stars with $G<16$ (including the microlensed source) do not follow the same color-color relation as those with $16<G<18$.