Kepler-167e: Cold Jupiter Analog
- Kepler-167e is a long-period cold giant exoplanet with near-Jovian size, a ~1071-day orbit, and a 16-hour transit, marking it as the first validated transiting Jupiter analog.
- Its confirmation through combined Kepler photometry and radial-velocity measurements precisely determined its mass, radius, and low irradiation, affirming its benchmark status in comparative giant-planet science.
- The unique system architecture—featuring inner super-Earths alongside an outer cold giant—provides key insights into planetary formation theories and drives advanced JWST studies on oblateness and exomoon detection.
Kepler-167e is a long-period, cold giant exoplanet transiting the K-dwarf Kepler-167 (KIC-3239945), notable as the first validated transiting Jupiter analog and, after subsequent radial-velocity work, a directly confirmed one. With an orbital period near $1071.23$ days, a -hour transit, low irradiation, and a system architecture that combines an outer gas giant with three inner small transiting planets, it occupies parameter space that is largely inaccessible to standard transit surveys. That rarity has made it a benchmark target for validation methodology, ephemeris maintenance, comparative giant-planet science, formation-theory tests, and more recently for JWST studies of planetary oblateness and exomoons (Kipping et al., 2016, Chachan et al., 2021, Liu et al., 2024).
1. Discovery, validation, and system context
Kepler-167e was initially identified as KOI-490.02 and validated from Kepler archival photometry as the first transiting Jupiter analog. The original Kepler analysis established a planet with radius , low eccentricity , equilibrium temperature K, and orbital period d. The host star, Kepler-167, was characterized as an early K dwarf, specifically K3-K4 or K4, with spectroscopic parameters , , , and . Its optical faintness (0) contrasts with more favorable infrared magnitudes, including 1, which later became important for Spitzer and JWST follow-up (Kipping et al., 2016).
Validation was statistical rather than dynamical in the discovery paper. The BLENDER framework was used to test and reject background eclipsing binaries (BEB), background/foreground stars transited by a larger planet (BP), hierarchical triple systems where a bound companion is eclipsed by a star (HTS), and hierarchical triple systems where a bound companion is transited by a planet (HTP). Supporting observations included Keck/HIRES spectroscopy, which found no secondary spectral lines down to companions contributing about 2 of the primary’s flux if velocity-separated by more than 3, and Keck/NIRC2 adaptive-optics imaging, which resolved a companion at 4, 5, with 6 and 7. That nearby star had to be modeled as a dilution source and a potential false-positive host, but the validation analysis ruled it out for Kepler-167e (Kipping et al., 2016).
The planetary system is architecturally distinctive. Kepler-167 hosts three close-in small planets—Kepler-167b, c, and d—with periods 8, 9, and 0 d, and radii 1, 2, and 3, respectively, plus the outer giant Kepler-167e. Later Gaia-informed analysis revised the inner radii slightly upward to 4, 5, and 6, while keeping the broader system picture intact. The key point is the coexistence of a compact inner super-Earth system with an outer cold giant beyond 7 au; in the 2021 formation study, this was described as the only known system with multiple inner transiting super-Earths and a confirmed outer transiting gas giant companion beyond 8 au (Kipping et al., 2016, Chachan et al., 2021).
A recurrent classification issue in the early literature was whether radius alone could distinguish a giant planet from a brown dwarf. Because Kepler-167e sits in the nearly flat, radius-degenerate regime around Jupiter radius, the discovery paper could not measure mass directly and forecast a broad 9 mass range of 0. It nevertheless found a 1 probability that the object lies below the hydrogen-burning limit and, after applying an occurrence prior 2, odds of 3 in favor of a Jupiter-like planet over a brown dwarf. That ambiguity was later removed by radial velocities (Kipping et al., 2016).
2. Orbital configuration and physical properties
The original Kepler transit solution measured 4, 5, 6, 7, 8, 9, 0, and 1. The later RV-plus-transit solution gave closely related values, including 2, 3, and a transit depth 4. The resulting transit is both deep and long: the depth is about 5–6, and later work repeatedly emphasized the 7-hour duration as a defining observational feature (Kipping et al., 2016, Chachan et al., 2021).
| Parameter | Value | Source solution |
|---|---|---|
| Orbital period 8 | 9 d | RV + transit |
| Semimajor axis 0 | 1 au | RV + transit |
| Mass 2 | 3 | RV + transit |
| Radius 4 | 5 | RV + transit |
| Mean density | 6 | RV + transit |
| Equilibrium temperature | 7 K | RV + transit |
The “Jupiter analog” classification rests on several jointly measured properties rather than on radius alone. In the discovery paper, the argument was that Kepler-167e combines near-Jovian size, low eccentricity, and a cold orbit beyond the snow line. Its semimajor axis was measured as 8, with received insolation 9, equilibrium temperature 0 under the assumption of Jupiter-like Bond albedo 1, and “effective semimajor axis” around a Sun-like star 2. The later RV study confirmed the mass to be essentially Jovian and reported 3 K under assumptions of Bond albedo 4 and full heat redistribution, reinforcing the same classification (Kipping et al., 2016, Chachan et al., 2021).
The original eccentricity estimate was itself methodologically important. It was derived without radial velocities through multibody asterodensity profiling, using the inner three planets as a stellar-density anchor and the photo-eccentric effect for the outer planet. The key relation was the light-curve density estimator
5
for a transiting planet on a nearly Keplerian circular orbit. Comparison between the stellar density implied by the inner planets and the transit shape of Kepler-167e yielded 6. The later RV analysis did not significantly detect eccentricity and instead quoted a 7 upper limit 8, with 9 rad. These two results are consistent in the sense that both place Kepler-167e in the low-to-moderate eccentricity regime characteristic of dynamically quiet cold giants (Kipping et al., 2016, Chachan et al., 2021).
A common misconception is to treat Kepler-167e as relevant to habitable-zone studies because the system contains a large gap between the inner planets and the outer giant. The discovery paper explicitly distinguished system architecture from planetary habitability: the habitable zone lies in the “large cavity of transiting planets,” but Kepler-167e itself is not a habitable-zone world. With 0 and 1, it is a cold Jupiter-sized planet well beyond the habitable zone (Kipping et al., 2016).
3. Ephemeris control and timing measurements
Because Kepler observed only two transits, early follow-up was dominated by ephemeris control rather than by attempts at detailed atmospheric or dynamical characterization. Spitzer addressed the central risk: hidden transit timing variations (TTVs) could have rendered future observations with HST or JWST impractical. A 10-hour partial transit observation on 2018 December 14 with IRAC Channel 1 at 2 produced 3 and 4. The corresponding infrared transit depth,
5
was consistent with the Kepler optical depth 6, and the timing was consistent with a linear ephemeris rather than with order-hours to days TTVs (Dalba et al., 2019).
Using the two Kepler epochs and the new Spitzer epoch, the linear ephemeris
7
yielded 8 and 9. The same analysis stated that TTVs of 11, 34, and 57 minutes were ruled out at 0, 1, and 2, respectively, and forecast future transits through 2030 with uncertainties below six minutes. For a transit lasting roughly 16 hours, that level of timing precision converted Kepler-167e from a risky target into a schedulable one (Dalba et al., 2019).
The next major timing milestone was the first ground-based detection of a Kepler-167e transit. From 2021 November 18 to 20, the Unistellar network assembled 43 observations from nine countries, retaining 27 data sets for the final analysis. A dynamic-nested-sampling fit measured
3
equivalently UTC 2021 November 19 17:20:51 4 min. This was consistent with the predicted value 5, an observed-minus-calculated offset of roughly 6 minutes, and thus again consistent with the absence of large TTVs. The campaign was notable not because it improved the space-based ephemeris precision, but because it showed that globally distributed small telescopes could keep the timing of a 7-day transiting giant current without requiring expensive space observatories (Perrocheau et al., 2022).
The two follow-up studies therefore addressed different operational problems. Spitzer demonstrated that the two-transit Kepler ephemeris had not been invalidated by large unseen TTVs, while the Unistellar campaign demonstrated that a 8-hour transit of a faint 9 target could nevertheless be recovered from the ground. Together they established Kepler-167e as an unusually manageable long-period transit target despite its rarity and long period (Dalba et al., 2019, Perrocheau et al., 2022).
4. Mass determination, internal composition, and formation constraints
The 2021 radial-velocity study transformed Kepler-167e from a radius-validated Jupiter analog into a dynamically characterized one. Using 13 Keck/HIRES iodine-cell measurements obtained between 2017 August and 2020 December, together with the two Kepler transits, the authors measured 0 and 1, with the RV signal described as a 2 detection. The fitted long-term slope was 3, consistent with zero, and no covariance was found between the velocities and the 4 activity index. This measurement directly confirmed the “Jupiter analog” classification on mass grounds (Chachan et al., 2021).
The same study tightened the physical interpretation of the planet. Combining the RV mass with 5 yielded a mean density of 6. In the giant-planet structure framework of Thorngren et al., updated with the Chabrier et al. H/He equation of state, the inferred bulk metallicity was 7, corresponding to a heavy-element content of 8, summarized in the abstract as 9. The host star’s metallicity corresponds to 00, so the planet is substantially metal-enriched relative to the star. The analysis also emphasized a key limitation: the data constrain total heavy-element inventory, not its partition between a distinct core and a metal-rich envelope (Chachan et al., 2021).
These measurements matter because Kepler-167e is not merely an isolated cold giant. It resides in a system with three inner small planets whose radii lie at or below the radius valley. Their masses were not directly measured, but using the Ning et al. non-parametric mass-radius relation, the RV study estimated 01, 02, and 03 for planets b, c, and d, for a total 04. That made the total solid inventory of the system—roughly 05 in the giant plus 06 in the inner planets—a constraint on the primordial disk rather than on Kepler-167e alone (Chachan et al., 2021).
The formation analysis interpreted those constraints in a pebble-accretion framework. Using the twopoppy two-population dust model of Birnstiel et al., passively irradiated disks, and seed cores inserted at 07 yr, the study found that giant cores analogous to Kepler-167e require early seed formation: seeds introduced at 08 yr do not reach pebble isolation in any model, while successful giant-core formation generally requires 09 yr and typically reaches isolation by 10 Myr. When the model was specialized to Kepler-167, the authors concluded that the natal disk likely contained 11 of dust and had size 12 au. They further argued that disks capable of forming an outer cold giant can still deliver enough solids inward to make inner super-Earths, so the coexistence of the two populations in Kepler-167 need not be paradoxical (Chachan et al., 2021).
5. Kepler-167e as a transit-shape and oblateness target
A 2024 JWST-oriented study recast Kepler-167e as a shape-and-spin target rather than only a mass-and-orbit target. The central observable is planetary oblateness: rotation-induced flattening, which in true three-dimensional form is defined as
13
with 14 and 15 the equatorial and polar radii. Transit photometry, however, measures only the projected silhouette. The study therefore worked with projected oblateness 16 and projected spin obliquity 17, the angle between the projected spin direction and the orbital direction on the sky. The basic premise was that a rapidly rotating cold giant could show ingress/egress asymmetries encoding both flattening and projected spin orientation (Liu et al., 2024).
Kepler-167e was treated as an unusually favorable test case because the expected signal is large by exoplanet standards. The adopted parameters were 18, 19, 20, and a 21 hr transit around a quiet K dwarf with 22, 23, 24, and 25. The depth is about 26–27, and the long transit means that the oblateness signature at ingress and egress is resolved in time. In the absence of limb darkening, the maximum amplitude was approximated as
28
and realistic limb-darkened simulations for Kepler-167e-like parameters reached 29 ppm for 30, directly comparable to the assumed JWST precision of about 31 ppm per 138 s exposure (Liu et al., 2024).
That study also introduced the transit code JoJo, which models the planet as a projected ellipse transiting a limb-darkened stellar disk and uses Green’s theorem to reduce the blocked-flux calculation from a 2D area integral to a 1D boundary integral. For Kepler-167e-like parameters, about 30 integration points were sufficient to reach 32 ppm precision, and the method was reported to be about six times faster than the earlier polygon-based PYPPLUSS approach. This computational gain mattered because the inference problem involved Monte Carlo exploration of oblate-transit parameter space rather than a single forward model (Liu et al., 2024).
The detection forecasts were driven less by absolute signal amplitude than by symmetry breaking. Injection tests compared spherical transits with projected-oblate cases 33 at 34, 35, and 36. When 37, the oblate signal produced clear left-right asymmetry between ingress and egress, and the best oblate model was favored over the best spherical model by 38. By contrast, for 39, the signal remained theoretically large but became symmetric enough that changes in impact parameter, transit duration, stellar density, and limb-darkening profile could absorb it. The paper therefore emphasized a practical threshold of roughly 40: if the projected spin axis is modestly misaligned, JWST should detect Saturn-like oblateness from a single transit or place a strong upper limit; if nearly aligned, the inference becomes strongly degenerate (Liu et al., 2024).
The null-case forecast was equally important. For spherical injections, the oblate model did not generate robust false positives. Instead, if 41, the study found a stringent upper limit of roughly 42 at about the 43 level. For Kepler-167e, that would imply a spin period longer than about 11 hr through the Darwin-Radau relation, excluding Jupiter- and Saturn-like rotation rates. A major caveat was already explicit in this forecast paper: transit photometry constrains only projected shape and projected obliquity, not the full three-dimensional spin vector, and even a genuinely oblate planet can project to a smaller apparent signal (Liu et al., 2024).
6. JWST oblateness constraints from the 2024 transit
JWST observed a transit of Kepler-167e in October 2024 with NIRSpec/Prism-Clear over 44–45, obtaining a 46-hr time series with 134,208 integrations. Operational constraints split the visit into six exposures with short pauses, and exposures 3–5 contained the Kepler-167e transit. The program was intentionally designed for both oblateness and exomoon sensitivity, but the multi-exposure structure proved to be a central limitation for the former because each exposure showed its own decreasing, nonlinear trend (Cassese et al., 3 Nov 2025).
The oblateness analysis explicitly treated pipeline dependence as part of the inference problem. It explored 3 reduction pipelines with 6 settings each, 6 time-binning cadences, 3 parametric trend models, 3 Gaussian-process augmentations, and 5 limb-darkening models, for 4320 model/data combinations before nested sampling. The three pipelines were Eureka!, ExoTiC-JEDI, and the new independent detector-level reduction framework katahdin. Oblate transits were modeled with squishyplanet, and Bayesian model comparison used nautilus. This unusually broad exploration reflected the fact that the target signal exists almost entirely on ingress/egress timescales of 47–30 minutes, while the dominant systematics occur on exposure-long 48-hr timescales (Cassese et al., 3 Nov 2025).
The principal result was non-detection in the strict model-selection sense. Across the final nested-sampling comparisons, spherical and oblate models fit the data equally well, with Bayes factors below the adopted threshold for strong evidence (49, equivalently 50). All marginalized posteriors for projected oblateness peaked at 51. The conservative quoted bound was
52
at 95% credibility for the projected oblateness, while some end-to-end analyses produced tighter 95% limits down to
53
Under the explicit assumption that the projected flattening can be interpreted as true oblateness because the spin axis is aligned with the sky plane, the abstract quoted a rotation-period lower limit
54
The paper was careful to frame this as an orientation-dependent translation rather than as a model-independent spin measurement (Cassese et al., 3 Nov 2025).
Two physical distinctions were central to the interpretation. First, the transit constrains projected oblateness rather than the intrinsic flattening 55. The study noted that, under random orientations, projected 56 is typically only 57–58 of true 59, so a projected upper bound only weakly constrains intrinsic shape unless alignment assumptions are added. Second, the data quality was limited more by exposure-long trends and correlated noise than by photon noise. Preferred Gaussian-process length scales were often comparable to about half the ingress/egress duration, which means the GP had enough flexibility to absorb or mimic the very deviations that would encode oblateness. The robust conclusion was therefore an upper limit on projected shape, not a direct inference of Jupiter-like or Saturn-like rotational flattening (Cassese et al., 3 Nov 2025).
This outcome provides an objective counterpoint to the 2024 forecast study. The earlier analysis had identified Kepler-167e as perhaps the best known current target for single-transit JWST oblateness work, especially if 60. The actual 2024 JWST data instead showed that exposure-long trends, multi-exposure resets, and reduction dependence materially degrade the inference. The later paper therefore recommended future observing strategies that avoid splitting the transit across multiple exposures when possible and stressed that any eventual nonzero oblateness claim should survive multiple independent reduction pipelines (Cassese et al., 3 Nov 2025).
7. Exomoon search, revised JWST transit fit, and unresolved issues
The same 61-hr JWST/NIRSpec dataset was also used for the first dedicated exomoon search with JWST, again taking Kepler-167e as the flagship target because a long-period cold giant in a dynamically quiet system is an astrophysically plausible host for large regular moons. The analysis compared a planet-only model 62 with a planet-plus-moon model 63 across a 64 grid of 12 reduction/trend realizations: three reductions (a custom conservative pipeline, ExoTiC-JEDI, and katahdin) and four exposure-long trend models (quadratic, exponential-plus-linear, squared-exponential GP, and Matérn-65 GP). Seven of the twelve realizations formally favored an exomoon by the paper’s detection criterion, typically corresponding to a Roche-skimming moon with 66–0.10, that is, roughly 67 of the planet’s radius (Kipping et al., 19 Nov 2025).
The paper did not interpret those formal detections as secure evidence for a moon. The preference for 68 was dominated by a localized mid-transit feature described as syzygy-like, while a second post-transit dip was not robust across trend models. The authors concluded that the only likely real astrophysical feature actually driving the moon fits was the mid-transit anomaly, but that the same feature is highly ambiguous with a starspot crossing. They showed that a spot large enough to explain the signal is compatible with earlier Kepler data: a 69 ppm effective spot in the Kepler band would require a spot of about 70, which would translate to about 263 ppm in the JWST bandpass and, in the fully opaque limit, mimic an apparent moon with 71. That is larger than essentially all of the inferred moon sizes in the preferred non-outlier fits, making the spot interpretation entirely viable (Kipping et al., 19 Nov 2025).
The most conservative result was therefore a sensitivity limit rather than a detection. After masking the syzygy-like interval and using the custom reduction with a Matérn-72 GP, the analysis inferred
73
with a 95% upper limit
74
Using the measured planet radius, this corresponds to
75
or, in transit-depth units,
76
The main methodological lesson was that exposure-long NIRSpec trends dominate the uncertainty budget for phenomena on transit-duration timescales, so single-transit exomoon inference remains highly vulnerable to both systematics modeling and stellar heterogeneity (Kipping et al., 19 Nov 2025).
The exomoon paper also reported a revised planet-only JWST transit fit. Averaging over all 12 reduction/trend combinations gave 77, 78, 79, 80, 81, and 82. It further stated that the JWST transit occurred a bit over an hour later than expected from earlier ephemerides and gave a revised fit 83, 84, suggesting transit-timing variations at the 85-minute level and potentially hinting at another outer giant. A plausible implication is that the timing picture is no longer as settled as the earlier Spitzer and ground-based analyses had suggested, although the same paper repeatedly emphasized that its inference is strongly entangled with trend modeling and that only another high-precision transit can break the current degeneracies (Kipping et al., 19 Nov 2025).
Kepler-167e thus occupies a distinctive position in exoplanet research. It is simultaneously a rare cold transiting Jupiter analog, a fully confirmed giant with measured mass and metal enrichment, a key case study for the co-formation of outer giants and inner super-Earths, and a testbed for whether JWST can infer planetary shape or detect large moons from a single transit. The current observational record supports the planet’s long-period cold-giant status very strongly, but it does not yet support a robust detection of either oblateness or an exomoon. The near-term scientific agenda therefore remains defined by repetition and control: another transit, especially the one recommended for October 2027, would distinguish repeatable orbital phenomena from starspot structure and instrument-trend degeneracy, and would determine whether Kepler-167e becomes primarily a benchmark in comparative giant-planet physics or a cautionary benchmark in long-baseline JWST time-series analysis (Cassese et al., 3 Nov 2025, Kipping et al., 19 Nov 2025).