{"id":"37b4611b-ba7c-47f6-99df-aab02172c598","arxiv_id":"2412.03651","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"OGLE-2012-BLG-0563Lb is a 1.1 Jupiter-mass planet orbiting a 0.8 solar-mass K dwarf at 5.5 kpc, with a host mass 2.4 times larger than originally reported.","lead":"By combining new Hubble and Keck images with the light curve, this paper finds that a previously characterized microlensing planet is actually a Jupiter-mass planet around a K dwarf at about 5.5 kpc, not a low-mass star at 1.3 kpc. The re-analysis exposes systematic errors in some ground-based photometry and demonstrates an image-constrained modeling approach for future Roman Space Telescope surveys.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central revision rests on a two-star decomposition of images with the lens and source only 37% of the Keck PSF FWHM apart; the paper itself notes a third star could hide between them, and such a star would bias the source magnitude (hence θE) and the lens magnitude (hence mass/distance).","rationale":"The reader's weakest assumption identifies the correct load-bearing point. The paper's central claim is that the host is a 0.801 M_sun K dwarf at 5.49 kpc, a factor of 2.4 more massive than the F15 value. What has to be true for this claim to hold is that the high-resolution image constraints are unbiased, and the least secure part of that chain is the two-star decomposition of partially resolved images. The paper explicitly acknowledges in Section 2.1 that a third star between the two components would be very difficult to detect, and the two-star fit has strongly correlated magnitude errors. Since the fitted source magnitude sets θ* and hence θE, and the fitted lens magnitude sets the mass-luminosity distance, a hidden third star can shift both θE and the lens brightness in a way that directly biases M_host and D_L. I do not think the data exclusion is the main issue, because Table 2 shows that including FTS/B&C data with the image constraints yields nearly the same parameters as excluding them; the driver is the image constraints themselves. The three-passband agreement is reassuring but does not independently validate the two-star assumption, because all passbands are analyzed with the same assumption. The concrete three-PSF test is decisive: it uses the same imaging data to place quantitative limits on a hidden third star and shows whether any acceptable third-star solution can move the parameters outside the quoted error bars. The verdict should remain CONDITIONAL: the claim is internally consistent and supported by multiple passbands, but the third-star possibility is an admitted, testable caveat rather than a fatal flaw.","tokens_in":21489,"tokens_out":13292,"duration_ms":131696,"concrete_test":"Perform a three-PSF-component fit to the Keck NIRC2 K, HST F814W, and F555W images, allowing a third component at a position along the lens-source line, with flux from 0% up to ~50% of the combined source-plus-lens light. For every third-component solution with Δχ² < 2 relative to the two-star fit and with colors consistent with the bulge/disc main-sequence locus of Figure 3, re-run the image-constrained MCMC with the corresponding source and lens magnitude constraints. If any such solution moves M_host outside 0.801 ± 0.08 M_sun or D_L outside 5.49 ± 0.5 kpc, the claimed revision is not robust; if none do, the third-star concern is quantitatively settled.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is not the post-hoc exclusion of the FTS/B&C data; Table 2 columns 4 and 5 show that the image-constrained parameters are nearly identical with or without those data. It is the assumption, in the two-star PSF decomposition of the Keck K, HST F814W, and F555W images (Section 2.1, Table 1), that all flux at the event position is divided between exactly two stars: the lens (star 1) and the source (star 2). The lens-source separation is 23.29 ± 0.98 mas, only 37% of the Keck PSF FWHM, and the paper states in Section 2.1 that a third star located between the two components would be very difficult to detect. The two-star fit yields strongly correlated magnitude uncertainties, so the individual lens and source magnitudes entering the Gaussian constraints are not independently pinned down; the fitted source magnitude controls θ* and hence θE = θ* tE/t*, while the fitted lens magnitude sets the mass-luminosity distance. If a third star contributes even roughly 10-20% of the combined flux, the source magnitude can be biased by ~0.1-0.2 mag, shifting θE by ~10-20%, and the lens magnitude bias propagates into M_host and D_L. Since all three passbands are reduced with the same two-component model, their agreement is not an independent check of that assumption. Thus the central claim is only as secure as the two-star assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript uses Keck NIRC2 adaptive-optics and HST/WFC3 images of the microlensing event OGLE-2012-BLG-0563 to constrain a light-curve model of the planetary system. A two-star PSF decomposition of the K-, I-, and V-band images gives positions, proper motions, and magnitudes for the candidate lens and source; these are imposed as Gaussian constraints in image-constrained modeling of the OGLE/MOA and follow-up photometry. The resulting solution has t* ≈ 0.043 d, θE ≈ 0.65 mas, host mass 0.801 ± 0.033 M_sun, planet mass 1.116 ± 0.087 M_Jup, and lens distance 5.49 ± 0.56 kpc, in strong disagreement with the Fukui et al. (2015) values. The authors attribute the difference to systematic errors in the FTS and B&C data and argue that image-constrained modeling will be valuable for the Roman Space Telescope's exoplanet survey.","tokens_in":21796,"tokens_out":7785,"duration_ms":76442,"significance":"Should the result stand, it is significant: it removes a long-standing low-distance outlier identified by Penny et al. (2016), demonstrates a method for catching systematic photometry errors, and provides a concrete case study for Roman. The paper's strengths include public data availability, independent Keck and HST proper-motion measurements that agree (Table 1), an internal consistency check in which the OGLE chi^2 improves when image constraints are added (Table 2), and a multi-passband mass-distance consistency check (Figure 8). The central claim is nevertheless only as secure as the assumption that the blended light is exactly two stars; the paper itself notes that a third star could hide between the two components, and this is not quantitatively excluded.","major_comments":[{"comment":"The load-bearing step of the analysis is the two-star PSF decomposition. The measured lens-source separation is 23.29 ± 0.98 mas, only 37% of the Keck FWHM, and Section 2.1 explicitly states that a third star located between the two components would be very difficult to detect. Such a star would bias the fitted source magnitude (and hence theta* and theta_E = theta* t_E/t*) and the lens magnitude (and hence M_host and D_L), and because all three passbands are reduced with the same two-component model, their cross-band agreement is not an independent validation of the assumption. Please perform explicit three-star fits or equivalent injection tests that place quantitative upper limits on possible third-star flux, and propagate any allowed bias into the reported masses, distance, and theta_E.","section":"Section 2.1, Table 1, Section 3"},{"comment":"The quoted host mass uncertainty, ±0.033 M_sun (about 4%), is small compared with the Table 1 lens magnitude uncertainties of roughly 0.1-0.2 mag, and the lens mass is obtained from the Bennett et al. (2018b, 2020) empirical mass-luminosity relations. It is not stated whether the scatter and systematic uncertainty of those relations, and the h_dust = 0.10 ± 0.02 kpc uncertainty in Eq. (6), are propagated in the MCMC. Please state explicitly which input uncertainties are included and quantify the sensitivity of M_host, D_L, and m_pl to the mass-luminosity calibration and to the assumed extinction scaling.","section":"Section 5 and Eq. (6)"}],"minor_comments":[{"comment":"In the sentence after Eq. (3), 'DS ≃ 1.3 kpc' should read 'DL ≃ 1.3 kpc', since it refers to the lens distance; also the notation v⊕E,N is confusing and should be defined more carefully.","section":"Section 2, Eq. (3)"},{"comment":"The caption labels Star 1 as the lens and Star 2 as the source, but this identification is established later in Section 3; please label them as 'candidate lens' and 'candidate source' in the table and caption.","section":"Table 1 caption"},{"comment":"The sentence 'F15 found K_LS = 17.071 ± 0.044, which is within 1 sigma of our value (if we combine the error bars)' should state the quadrature combination explicitly, since the two measurements have different systematics.","section":"Section 2.1"},{"comment":"The statement that 'B&C has also proved to be problematic for some previously analyzed events' lacks specific references or event names; please provide them so the reader can evaluate the precedent.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"The main technical barrier to acceptance is the unverified two-star assumption in the PSF decomposition; if the authors can supply three-star fits or convincing injection tests, I would support acceptance. Note also that much of the technical machinery lives in companion papers (Bhattacharya et al. 2024; Bennett et al. 2024), which makes independent verification harder; the revision should make the key PSF-fitting details sufficiently self-contained for a referee."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — the headline is that this paper fixes a known bad distance measurement in the microlensing planet sample, and it does it with a method that is clearly explained and mostly self-consistent. The new number is a factor-2.4 upward revision of the host mass for OGLE-2012-BLG-0563L, from roughly 0.34 Msun at 1.3 kpc to 0.80 Msun at 5.5 kpc, with a Jupiter-mass planet. That removes one of the six low-distance outliers flagged by Penny et al. (2016), leaving only OGLE-2006-BLG-109 plausibly in that category. The result matters for the demographic sample.\n\nWhat the paper does well: the image constraints are independent of the light curve, and the fact that OGLE chi^2 improves when they are applied is a genuine cross-check. The exclusion of the FTS and B&C data is post hoc, but Table 2 shows the final parameters are nearly identical with or without those two datasets, so that decision is not carrying the result. The authors also state plainly the possibility of a third star between the lens and source, which is the right thing to do.\n\nThe soft spot is exactly that two-star assumption. The lens and source are separated by only 23 mas, which is 37% of the Keck PSF FWHM and about two-thirds of a Hubble pixel. The paper itself acknowledges that a third star between the two components would be very difficult to detect. If a third star contributes even 10-20% of the blended flux, the derived source magnitude changes enough to shift theta_E by 10-20%, and the lens magnitude changes shift the mass and distance. Since all three passbands are fit with the same two-component model, their agreement does not independently validate the assumption. So the central revision is credible but not airtight; the paper frames it as such.\n\nMinor issue: the abstract lists D_L=5.46±0.56 while Table 4 gives 5.49±0.56; a small internal inconsistency that a referee would want cleaned up.\n\nWho this is for: anyone working on microlensing planet demographics or on Roman photometry systematics. It is an application of an existing method rather than a new technique, but it demonstrates the method catching a real error. I'd send it to review; it deserves referee time. My own verdict is conditional, but the paper is honest and the evidence is decent.\n\nRecommendation: engage with it, but make sure the referee pushes on the two-star decomposition and asks for a quantitative bound on a possible third star, either from the images or from a Galactic prior.","headline":"Solid application of image-constrained modeling that fixes a bad distance for a known microlensing planet, though the two-star decomposition of the blended images is the load-bearing assumption to watch.","tokens_in":22446,"tokens_out":3635,"would_cite":true,"duration_ms":36111,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Hubble and Keck imaging shows OGLE-2012-BLG-0563Lb's host is a 0.80 solar-mass K dwarf at 5.5 kpc, correcting the original mass estimate upward by a factor of 2.4.","keywords":["gravitational microlensing","exoplanets","image-constrained modeling","Hubble Space Telescope","adaptive optics","stellar mass measurement","OGLE-2012-BLG-0563","systematic photometry errors"],"falsifier":"Take a second-epoch image with higher resolution or at a later time when the lens and source have separated further, and fit a three-star PSF model; if a third star with significant flux lies between the two detected components, the two-star fit's magnitudes and proper motions are biased and the derived host mass and distance shift. A radial-velocity measurement of the resolved candidate host could also check whether it is a single K dwarf.","tokens_in":21259,"feed_emoji":"🔭","tokens_out":5547,"duration_ms":53917,"temperature":0.7,"pith_summary":"This paper uses high-resolution Hubble and Keck images to re-fit the microlensing light curve of OGLE-2012-BLG-0563, a Jupiter-mass planet candidate discovered in 2015. The authors claim that the host star is a K dwarf with mass 0.801±0.033 solar masses at a distance of 5.49±0.56 kpc, orbited by a planet of 1.116±0.087 Jupiter masses. This contradicts the discovery paper's host mass of about 0.34 solar masses and distance of about 1.3 kpc, which were wrong by factors of roughly 2.4 and 4. The cause is a systematic error in some ground-based light curve photometry that inflated the angular Einstein radius by about a factor of 2. The paper argues that image-constrained modeling—fitting the light curve with constraints from resolved images of the lens and source—is a general method for catching such systematics, including in the upcoming Roman Space Telescope survey.","feed_headline":"Host of microlensed Jupiter is a K dwarf at 5.5 kpc","feed_subtitle":"Hubble and Keck images overturn the 2015 mass estimate by a factor of 2.4.","key_machinery":"The central object is the image-constrained modeling method: a light-curve fitting code that adds Gaussian constraints on the lens and source magnitudes in K, I, and V, the combined lens-plus-source magnitudes, and the two components of the heliocentric relative proper motion. These constraints come from PSF fits to the partially resolved Keck and Hubble images of the blended source and lens. The method also uses a dust-extinction model that scales with lens distance and empirical mass-luminosity relations, so the measured brightness of the resolved lens can be converted into a host mass. The load-bearing quantity that the constraints correct is the source radius crossing time, which sets the angular Einstein radius; an inflated θE had pushed the mass-distance relation toward low masses and short distances.","core_discovery":"By combining HST WFC3 images in F814W and F555W with Keck adaptive-optics K-band images, the paper separates the microlensed source from the planetary host lens at a separation of about 23 milliarcseconds and measures their heliocentric relative proper motion (weighted mean 4.103±0.112 milliarcseconds per year) and magnitudes in three passbands. Feeding these measurements as Gaussian constraints into the light curve model reveals that the source radius crossing time is about twice as large as an unconstrained fit gives, so the angular Einstein radius is θE = 0.645±0.017 milliarcseconds instead of the previously claimed ~1.36 milliarcseconds. This changes the mass-distance intersection: the host is a 0.801±0.033 solar-mass K dwarf at 5.49±0.56 kpc, and the planet is 1.116±0.087 Jupiter masses. The close-wide degeneracy remains, giving projected star-planet separations of 1.50±0.16 AU (close model) or 8.41±0.87 AU (wide model). The discrepancy with the discovery paper is traced to systematic photometry errors in two ground-based follow-up datasets, which were excluded from the final fit.","pith_inferences":["If other microlensing planets originally reported at distances under 2 kpc suffered the same kind of blend-induced photometry systematics, some published host masses and distances in statistical samples may be biased; targeted re-imaging of those targets would settle it.","A third star hiding between the two partially resolved components would break the two-star assumption; a future epoch at larger lens-source separation or a space-based PSF decomposition could test this directly.","The image-constrained approach could be applied to archival Hubble and Keck data of other microlensing events with candidate host detections, potentially revising more mass measurements.","For Roman, image-constrained modeling could also help calibrate new infrared detector systematics by comparing resolved source brightnesses from high-resolution images to light-curve source magnitudes."],"forward_implications":["If correct, the planet orbits a relatively massive K dwarf in the far Galactic disk, not a nearby low-mass star, which changes the interpretation of the apparently nearby microlens planet population.","The angular Einstein radius is about 0.645 milliarcseconds, roughly half the original estimate, and the implied lens-source relative proper motion of about 3.7 milliarcseconds per year is consistent with the resolved images.","The close-wide degeneracy is not resolved, so the projected star-planet separation is either about 1.5 AU or about 8.4 AU; future astrometric epochs could distinguish these.","The systematic error was found only because image constraints contradicted the light curve fit, suggesting that similar image-constrained fitting will be valuable for validating Roman Space Telescope microlensing photometry."],"supporting_citations":[{"why":"Supplies the discovery light curve, Subaru adaptive-optics excess flux measurement, and the original low-mass, nearby host solution that this paper corrects.","marker":"Fukui et al. (2015)"},{"why":"Provides the Keck adaptive-optics two-star PSF analysis, magnitude difference, and proper-motion measurements used as constraints.","marker":"Bhattacharya et al. (2024)"},{"why":"Defines the image-constrained modeling method used in the light curve fits.","marker":"Bennett et al. (2024)"},{"why":"Provides the improved MOA photometry reduction and detrending used to re-reduce the light curve data.","marker":"Bond et al. (2017)"},{"why":"Supplies the Galactic model prior on source distance used in the Markov Chain Monte Carlo analysis.","marker":"Koshimoto et al. (2021a)"},{"why":"Provides the empirical mass-luminosity relations used to convert resolved lens magnitudes into host mass.","marker":"Bennett et al. (2018b, 2020)"}],"fun_headline_variants":["Hubble+Keck images reveal K dwarf host for Jupiter-mass planet","Image-constrained model corrects 2.4x mass error for planet host","K dwarf at 5.5 kpc hosts a Jupiter-mass planet, images show","Keck and Hubble imaging boosts host mass by 2.4x for microlensed planet"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The fit assumes exactly two stars contribute to the blended light at the event position—the source and the lens—so any third star between them would be missed and would bias the derived magnitudes, and therefore the host mass and distance.","fun_headline_variants_meta":{"raw":{"variants":["Hubble+Keck images reveal K dwarf host for Jupiter-mass planet","Image-constrained model corrects 2.4x mass error for planet host","K dwarf at 5.5 kpc hosts a Jupiter-mass planet, images show","Keck and Hubble imaging boosts host mass by 2.4x for microlensed planet"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000487,"raw_usage":{"total_tokens":2498,"prompt_tokens":1142,"completion_tokens":1356,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":758,"completion_tokens_details":{"reasoning_tokens":1268}},"tokens_in":758,"tokens_out":1356,"duration_ms":13327,"temperature":1.0,"reasoning_tokens":1268,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T22:14:59.934751+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a second-epoch image with higher resolution or at a later time when the lens and source have separated further, and fit a three-star PSF model; if a third star with significant flux lies between the two detected components, the two-star fit's magnitudes and proper motions are biased and the derived host mass and distance shift. A radial-velocity measurement of the resolved candidate host could also check whether it is a single K dwarf.","supporting_citations":[],"review_version":1}