WASP-189b: Ultra-Hot Jupiter Benchmark

• WASP-189b is an ultra-hot Jupiter defined by its 2.72-day polar orbit, extreme irradiation, and status as a benchmark for high-resolution atmospheric research.
• Doppler tomography and gravity-darkened photometry reveal a near-polar orbit with significant spin-orbit misalignment around a rapidly rotating A-type star.
• Multi-wavelength studies, including optical occultations and transmission spectroscopy, uncover thermal inversions, metal-rich terminators, and dynamic day-to-night winds.

WASP-189b is an ultra-hot Jupiter transiting WASP-189 = HR 5599 = HD 133112 = HIP 73608, a bright rapidly rotating A-type star in Libra. Discovered as a 2.72-d transiting planet in WASP-South photometry and confirmed with HARPS/CORALIE Doppler tomography and radial velocities, it was initially characterized as a $2.13 \pm 0.28\,M_{\rm Jup}$, $1.374 \pm 0.082\,R_{\rm Jup}$ giant on an essentially polar orbit with $\lambda = 89.3 \pm 1.4^\circ$ and $\Psi = 90.0 \pm 5.8^\circ$ (Anderson et al., 2018). Because the system combines a $V=6.6$ host, extreme irradiation, and unusually accessible orbital geometry, it has become one of the main benchmark targets for high-resolution studies of ultra-hot-Jupiter atmospheres from the near-ultraviolet to the near-infrared (Prinoth et al., 2021).

1. Discovery, host star, and basic system properties

The discovery analysis identified WASP-189b in a circular orbit with $P = 2.7240330 \pm 0.0000042\ {\rm d}$, $a = 0.0497 \pm 0.0026\ {\rm AU}$, $a/R_* = 4.591 \pm 0.041$, $i = 84.321 \pm 0.097^\circ$, and $b = 0.4537 \pm 0.0072$, with $1.374 \pm 0.082\,R_{\rm Jup}$0 explicitly adopted as the impact-parameter definition. Under zero albedo and efficient redistribution, the quoted comparison equilibrium temperature was $1.374 \pm 0.082\,R_{\rm Jup}$1, placing the planet among the hottest known exoplanets at the time (Anderson et al., 2018).

The host star is consistently described as a hot, rapidly rotating, bright A star. In the discovery paper it is classified as A6IV–V, with $1.374 \pm 0.082\,R_{\rm Jup}$2, $1.374 \pm 0.082\,R_{\rm Jup}$3, $1.374 \pm 0.082\,R_{\rm Jup}$4, $1.374 \pm 0.082\,R_{\rm Jup}$5, $1.374 \pm 0.082\,R_{\rm Jup}$6, $1.374 \pm 0.082\,R_{\rm Jup}$7, and age $1.374 \pm 0.082\,R_{\rm Jup}$8 (Anderson et al., 2018). CHEOPS-based reanalysis later adopted $1.374 \pm 0.082\,R_{\rm Jup}$9, $\lambda = 89.3 \pm 1.4^\circ$0, $\lambda = 89.3 \pm 1.4^\circ$1, $\lambda = 89.3 \pm 1.4^\circ$2, and age $\lambda = 89.3 \pm 1.4^\circ$3 (Lendl et al., 2020). A later ultraviolet study described the host as a bright A4 star, illustrating minor classification differences across the literature while retaining the same basic physical picture of a hot early-type rapid rotator (Sreejith et al., 2023).

The planetary radius was substantially revised upward by CHEOPS. The discovery paper reported $\lambda = 89.3 \pm 1.4^\circ$4 (Anderson et al., 2018), whereas CHEOPS transit modeling found a $\lambda = 89.3 \pm 1.4^\circ$5 deeper transit than the discovery paper and updated the radius to $\lambda = 89.3 \pm 1.4^\circ$6, with $\lambda = 89.3 \pm 1.4^\circ$7, $\lambda = 89.3 \pm 1.4^\circ$8, and $\lambda = 89.3 \pm 1.4^\circ$9 (Lendl et al., 2020). This revision propagated into much of the later atmospheric literature.

2. Orbital architecture and the role of rapid stellar rotation

WASP-189b is one of the clearest nearly polar hot-Jupiter systems. Doppler tomography during discovery showed the transit signature on the receding stellar hemisphere at nearly constant stellar velocity, exactly the morphology expected for a polar transit chord across a rapidly rotating star; in the adopted solution the sky-projected and true obliquities were $\Psi = 90.0 \pm 5.8^\circ$0 and $\Psi = 90.0 \pm 5.8^\circ$1 (Anderson et al., 2018). Because the host is an A star with sparse, very broad lines, tomography was more constraining than a standard Rossiter–McLaughlin fit, and fitting rotationally broadened profiles to the HARPS/CORALIE CCFs reduced the weighted RMS of the orbital RVs from $\Psi = 90.0 \pm 5.8^\circ$2 to $\Psi = 90.0 \pm 5.8^\circ$3 (Anderson et al., 2018).

Gravity-darkened photometry subsequently reinforced the polar interpretation. CHEOPS found asymmetric transits attributable to gravity darkening and derived $\Psi = 90.0 \pm 5.8^\circ$4 deg and $\Psi = 90.0 \pm 5.8^\circ$5, consistent with the spectroscopic result while using the star’s brightness asymmetry as the geometric diagnostic (Lendl et al., 2020). A later TESS+CHEOPS gravity-darkening analysis again found the planet to be in a polar orbit, with a reported spin-orbit angle of $\Psi = 90.0 \pm 5.8^\circ$6 deg, and found no hint of orbital precession when compared with the literature (Patel et al., 4 Dec 2025).

The system is also important in population context. The discovery paper noted that among the eight known hot-Jupiter systems with host-star $\Psi = 90.0 \pm 5.8^\circ$7, seven were strongly misaligned, and among the three systems with $\Psi = 90.0 \pm 5.8^\circ$8, two were polar and one aligned (Anderson et al., 2018). WASP-189b therefore became an important datum in the broader empirical association between hot hosts and dynamically extreme close-in giant-planet architectures.

3. Photometric emission, occultations, and phase-curve constraints

CHEOPS provided the first precision optical occultation constraints. From four occultations, the measured depth was $\Psi = 90.0 \pm 5.8^\circ$9 ppm, and comparison with PHOENIX+HELIOS modeling yielded a dayside temperature of $V=6.6$0 when assuming an unreflective atmosphere and inefficient heat redistribution; the same analysis also reported an optical brightness temperature of $V=6.6$1 (Lendl et al., 2020). The same study found no significant evidence for dayside variability across the 19-day CHEOPS occultation baseline (Lendl et al., 2020).

Later phase-curve work with TESS and archival CHEOPS extended this picture. The TESS phase curve yielded an occultation depth of $V=6.6$2 ppm, while the nightside flux, $V=6.6$3 ppm, was consistent with zero at $V=6.6$4. Inverting the phase curve to a temperature map led to expected median ranges of $V=6.6$5-$V=6.6$6 for the Bond albedo and $V=6.6$7-$V=6.6$8 for the heat redistribution efficiency, and the authors concluded that the dayside emission in the TESS and CHEOPS bandpasses is dominated by thermal emission with negligible reflected-light contribution (Patel et al., 4 Dec 2025).

This optical behavior is consistent with the wider ultra-hot-Jupiter picture developed for the system. The TESS/CHEOPS analysis explicitly associated the low optical reflectivity with strong $V=6.6$9 continuum opacity and interpreted the slight bandpass-dependent dayside temperatures as consistent with an inverted atmosphere in which the CHEOPS band probes somewhat higher, hotter layers than TESS (Patel et al., 4 Dec 2025). This suggests that WASP-189b’s optical broadband flux is set primarily by thermal structure rather than by reflective condensates.

4. Transmission spectroscopy and the terminator atmosphere

The transmission-spectroscopy record of WASP-189b is unusually rich but also methodologically complex. An early PEPSI/LBT partial-transit study reported no consistent atmospheric absorption in optical atomic tracers and argued that the apparent signals in H$P = 2.7240330 \pm 0.0000042\ {\rm d}$0, Fe I, and Mg I were driven by highly variable in-transit spectra more plausibly explained by stellar-surface inhomogeneity and/or stellar variability. That work concluded that the data showed no evidence for a highly extended atmosphere in the observed optical tracers and placed an upper limit of $P = 2.7240330 \pm 0.0000042\ {\rm d}$1 above the optical radius for the excited-hydrogen thermosphere (Cauley et al., 2020).

Subsequent multi-transit HARPS/HARPS-N studies, however, established a far more detailed transmission inventory. Using five transits, high-resolution cross-correlation detected Fe, Fe$P = 2.7240330 \pm 0.0000042\ {\rm d}$2, Ti, Ti$P = 2.7240330 \pm 0.0000042\ {\rm d}$3, Cr, Mg, V, Mn, and TiO robustly, with tentative Na, Ca, Sc$P = 2.7240330 \pm 0.0000042\ {\rm d}$4, Cr$P = 2.7240330 \pm 0.0000042\ {\rm d}$5, and Ni. The headline TiO result was a $P = 2.7240330 \pm 0.0000042\ {\rm d}$6 detection, while Fe and Fe$P = 2.7240330 \pm 0.0000042\ {\rm d}$7 were detected at $P = 2.7240330 \pm 0.0000042\ {\rm d}$8 and $P = 2.7240330 \pm 0.0000042\ {\rm d}$9, respectively (Prinoth et al., 2021). The same work emphasized that the line positions and apparent $a = 0.0497 \pm 0.0026\ {\rm AU}$0 values differ by species and interpreted this as direct observational evidence for three-dimensional thermo-chemical stratification: Fe, Cr, Mn, TiO, and V were blueshifted by several $a = 0.0497 \pm 0.0026\ {\rm AU}$1, whereas Ti and Mg were much closer to systemic velocity (Prinoth et al., 2021).

A separate three-transit HARPS/HARPS-N survey recovered Fe I, Fe II, and Ti I at $a = 0.0497 \pm 0.0026\ {\rm AU}$2, $a = 0.0497 \pm 0.0026\ {\rm AU}$3, and $a = 0.0497 \pm 0.0026\ {\rm AU}$4, respectively, and also reported an H$a = 0.0497 \pm 0.0026\ {\rm AU}$5 residual with depth $a = 0.0497 \pm 0.0026\ {\rm AU}$6, FWHM $a = 0.0497 \pm 0.0026\ {\rm AU}$7, and formal significance of roughly $a = 0.0497 \pm 0.0026\ {\rm AU}$8. That paper treated the Balmer and Ca II features as tentative because gravity darkening likely deforms the Rossiter–McLaughlin signal and the stellar correction was not yet sufficiently complete (Stangret et al., 2021). The methodological point is important: WASP-189b’s transmission spectrum is inseparable from line-profile distortions produced by a rapidly rotating, gravity-darkened A star.

Near-infrared transmission spectroscopy sharpened the role of continuum opacity. Simultaneous HARPS+NIRPS transit observations detected only atomic iron in HARPS, with S/N $a = 0.0497 \pm 0.0026\ {\rm AU}$9 and a measured centroid offset of $a/R_* = 4.591 \pm 0.041$0, but no corresponding Fe signal in NIRPS. Injection-recovery tests showed that Fe would have been detectable in NIRPS in the absence of continuum damping, and the joint optical+NIR retrievals therefore argued that $a/R_* = 4.591 \pm 0.041$1 suppresses near-infrared line contrast; the hydride-to-Fe ratio was inferred to exceed equilibrium predictions by about 0.5 dex (Vaulato et al., 28 Jul 2025). A later single-transit GIANO-B study found tentative molecular transmission signals for H$a/R_* = 4.591 \pm 0.041$2O and CO at $a/R_* = 4.591 \pm 0.041$3 and $a/R_* = 4.591 \pm 0.041$4, respectively, while reporting no secure detections for FeH, CO$a/R_* = 4.591 \pm 0.041$5, CH$a/R_* = 4.591 \pm 0.041$6, HCN, TiO, VO, or OH (Meni-Gallardo et al., 25 Jun 2026). Taken together, these results suggest a terminator rich in metals and structured by strong longitudinal and vertical gradients, with near-infrared molecular diagnostics partially obscured by continuum processes.

5. Dayside emission, thermal inversions, and atmospheric dynamics

WASP-189b’s dayside atmosphere is one of the clearest line-resolved cases for a thermal inversion in an ultra-hot Jupiter. HARPS-N dayside spectroscopy detected Fe I at $a/R_* = 4.591 \pm 0.041$7 using an emission-line template, which is direct evidence that the line-forming layers are hotter than the deeper continuum-forming layers. A two-point retrieval yielded $a/R_* = 4.591 \pm 0.041$8 at $a/R_* = 4.591 \pm 0.041$9 and $i = 84.321 \pm 0.097^\circ$0 at $i = 84.321 \pm 0.097^\circ$1, establishing a strong inversion over the relevant pressure range (Yan et al., 2020).

Near-infrared dayside emission spectroscopy then extended the picture to CO. GIANO-B detected CO lines in emission at $i = 84.321 \pm 0.097^\circ$2, with $i = 84.321 \pm 0.097^\circ$3 and $i = 84.321 \pm 0.097^\circ$4, fully consistent with the earlier Fe-emission inference of an inverted dayside (Yan et al., 2022). A later GHOST/Gemini South study independently recovered Fe I in emission at $i = 84.321 \pm 0.097^\circ$5 in the red arm, again verifying the thermal inversion from a different facility and reduction chain (Deibert et al., 2024).

High-resolution K-band spectroscopy with CRIRES$i = 84.321 \pm 0.097^\circ$6 connected these emission signals to atmospheric dynamics. That study detected CO and Fe at S/N $i = 84.321 \pm 0.097^\circ$7 and $i = 84.321 \pm 0.097^\circ$8, respectively, and found that the line profile is best fitted by a day-to-night wind of $i = 84.321 \pm 0.097^\circ$9, while the retrieved equatorial jet velocity of $b = 0.4537 \pm 0.0072$0 is consistent with the absence of such a jet (Lesjak et al., 2024). This interpretation is consistent with the earlier redshifted CO and Fe emission detections and provides a dynamical framework in which the dayside redshifts arise primarily from planet-wide day-to-night flow rather than from strong superrotation.

The dayside refractory inventory has also become a compositional probe. Simultaneous HARPS+NIRPS dayside thermal-emission retrievals detected Fe and Ti in cross-correlation and inferred a sub-solar Ti/Fe ratio, concluding that the planetary Ti/Fe lies between $b = 0.4537 \pm 0.0072$1 and $b = 0.4537 \pm 0.0072$2 of the stellar value at $b = 0.4537 \pm 0.0072$3 (Pelletier et al., 14 Jun 2026). The paper interpreted this slight underabundance as evidence that some titanium is missing from the gas phase, potentially because of a partial nightside cold trap rather than because titanium is globally absent (Pelletier et al., 14 Jun 2026). This suggests that the dayside inversion and the gas-phase titanium budget are related but not identical diagnostics.

6. Upper atmosphere, escape, and broader theoretical significance

The upper atmosphere of WASP-189b is markedly more extended in the near-ultraviolet than in the optical. CUTE observed three consecutive transits and measured a broadband NUV transit depth of $b = 0.4537 \pm 0.0072$4, about twice the optical/visual depth of $b = 0.4537 \pm 0.0072$5. In a 10 $b = 0.4537 \pm 0.0072$6 bin centered on the Mg II doublet, the reported radius ratio was $b = 0.4537 \pm 0.0072$7, with a quoted significance of $b = 0.4537 \pm 0.0072$8, interpreted as Mg II absorption beyond the Roche lobe. The same study inferred an upper-atmosphere temperature of about $b = 0.4537 \pm 0.0072$9 and a moderate mass-loss rate of about $1.374 \pm 0.082\,R_{\rm Jup}$00 (Sreejith et al., 2023). It also argued that the observed thermosphere is hotter than predicted by its self-consistent hydrodynamic reference model, pointing either to missing upper-atmospheric heating or to overestimated radiative cooling (Sreejith et al., 2023).

Possible NUV transit asymmetry has added a star–planet-interaction dimension. Reanalysis of the same CUTE dataset identified an approximately 31.5-minute phase offset during Visit 3 that is consistent with an early ingress, corresponding to an absorbing extension of about $1.374 \pm 0.082\,R_{\rm Jup}$01 ahead of the planet. Ideal-MHD simulations indicated that with a wind speed of $1.374 \pm 0.082\,R_{\rm Jup}$02 and an upper-atmospheric density of about $1.374 \pm 0.082\,R_{\rm Jup}$03, a higher-density compressed zone can form ahead of the planet within five planetary radii where the fast-mode Mach number falls below $1.374 \pm 0.082\,R_{\rm Jup}$04, even without a classical bow shock (Duann et al., 6 Nov 2025). Because the observational evidence is confined to one of the three visits, that paper explicitly treated the signal as tentative and used it as an upper bound for physical plausibility rather than as a definitive detection (Duann et al., 6 Nov 2025).

On the theoretical side, three-dimensional GCM modeling has emphasized that UV/optical absorbers strongly reshape the atmosphere of WASP-189b at low pressure. A dedicated Exo-FMS study found that including UV opacity sources such as TiO, VO, SiO, Fe, and Fe$1.374 \pm 0.082\,R_{\rm Jup}$05 produces sharp upper-atmosphere inversions and a dynamically distinct high-altitude region; the authors argued that optical high-resolution observations therefore probe an upper atmospheric layer rather than the deeper jet-forming levels (Lee et al., 2022). In a population framework, the later Ti/Fe study placed WASP-189b in a progressive “onset of titanium” sequence, arguing that the appearance of titanium in ultra-hot-Jupiter atmospheres coincides not with the first emergence of thermal inversions but with the nightside vaporization threshold of titanium condensates (Pelletier et al., 14 Jun 2026).

The system also has an explicit stellar-evolution timescale. Because the star is already somewhat evolved, the discovery paper estimated that it will expand to the planet’s current orbit in about $1.374 \pm 0.082\,R_{\rm Jup}$06 Myr, implying eventual engulfment of the planet as the star ascends the giant branch (Anderson et al., 2018). In that sense, WASP-189b is not only a laboratory for present-day atmospheric physics and orbital dynamics, but also a short-lived example of a close-in giant planet around a hot, evolving intermediate-mass star.