betaFeedback

WorksPhysics and Astronomy

Signatures of CH4_4 and NH3_3 in the atmosphere of ββ Pictoris b

Abstract

ββ Pictoris is a nearby (19.7 pc), young (∼\sim 24 Myr), bright (V=3.85V = 3.85) planetary system hosting three directly imaged planets. Previous high-resolution spectroscopy of ββ Pic b has detected H2_2O and CO, but not CH4_4 or NH3_3. Combining observations across a broad wavelength range has been shown to maximise sensitivity to such weak molecular signals. We searched for previously undetected molecular species in the atmosphere of ββ Pic b using high-resolution near-infrared spectra. We combined 44 CRIRES+ spectral settings covering the H, K, L, and M bands and cross-correlated the extracted planetary spectra with molecular templates. For species with uncertain detectability, we weighted the observations by their molecular information content and data quality, independent of the measured candidate-species cross-correlation functions. We recover H2_2O and CO at cross-correlation S/N values of approximately 200 and 50, respectively. We additionally identify tentative CH4_4 and NH3_3 signals near the expected radial velocity of ββ Pic b, with S/N values of approximately 5.3 and 5.7. Equivalent extractions at the same projected separation on the opposite side of the star show no comparable peaks. We classify the CH4_4 and NH3_3 signals as tentative because their cross-correlation S/N values do not fully account for possible residual systematics. If confirmed, these would be the first detections of both species in the atmosphere of ββ Pic b.

Introduction

High-resolution spectroscopy (HRS) has become a powerful tool for characterising the atmospheres of exoplanets. HRS resolves molecular bands into tens to hundreds of individual lines, and cross-correlating the observed spectra with a model template enhances the signal for a given species. HRS is highly complementary to low-resolution spectroscopy because, in addition to probing abundant atmospheric species, it is sensitive to species present at relatively low abundances. It has been applied to transiting and close-in planets, where the planet’s changing radial velocity shifts its lines relative to the stellar and telluric ones (e.g. [11, 37, 16]),

and it can also be combined with high-contrast imaging (HCI) to study giant planets at wide separations (>10> 10 au) that are spatially resolved from their host stars (e.g. [38, 43, 45, 14, 6]). This synergy enables robust and precise measurements of elemental abundance ratios, rotational and radial velocities, and orbital trajectories, all of which are key diagnostics for investigating planet formation pathways, interactions with circumplanetary discs, and the dynamical and migration histories of planetary systems (e.g. [17, 39] (Mair et al. 2021)).

β\beta Pictoris is one of the most remarkable and extensively studied planetary systems. It is a nearby (19.7 pc; [8]), young (24.7−0.7+0.924.7^{+0.9}_{-0.7} Myr; [24]), bright (V=3.85V = 3.85) A6V star surrounded by three known giant planets discovered through direct imaging or interferometry ([19, 20, 29, 9, 41]). β\beta Pic b is the second planet in order of increasing orbital separation from the host star. Its mass is estimated to be 7–12 MJupM_{\mathrm{Jup}}, and it has a semi-major axis of ∼\sim10 au ([18, 2]), making it a widely separated super-Jupiter. Owing to its suitability for high-contrast HRS, β\beta Pic b has been the subject of numerous atmospheric studies. Using CRyogenic high-resolution InfraRed Echelle Spectrograph (CRIRES+, R∼50 000R \sim50\,000–100 000; [15, 7]) K-band observations, Landman et al. [21] reported detections of H2_2O and CO. A study in the M-band led by Parker et al. [31] later confirmed the presence of the species that were claimed to have been detected in the K-band, and reported a tentative detection of SiO. Janson et al. [14] investigated the system in the L and M bands, confirming the presence of H2_2O but not that of SiO. Finally, the recent K-band study by González Picos et al. [10] detected 13^{13}CO and measured a 12^{12}CO/13^{13}CO ratio consistent with that of the local interstellar medium and lower than the solar photospheric value.

In this paper, we present a panchromatic analysis of β\beta Pic b with CRIRES+, covering the near-infrared (NIR) wavelength range from 1.5 to 5.5 μ\mum. During ESO Period P115, we acquired 17 new observations in the K, L, and M bands and complemented them with additional NIR spectra from previous publications (see Sect. 2). Janson et al. [14] demonstrated that combining datasets with different spectral settings increases the signal-to-noise ratio (S/N) more than combining datasets with equal wavelength coverage. The newly acquired data, covering several previously unexplored regions of the spectrum, allow us to improve sensitivity to unexplored molecular species (e.g. CH4_4 and NH3_3) and strengthen detections of molecules reported in earlier studies (i.e. H2_2O and CO).

Observations

We observed the β\beta Pic planetary system with the CRIRES+ spectrograph at the VLT between 25 September and 1 October 2025 during ESO Period P115 (Programme ID: 115.2820.001; PI: Janson). We acquired 20 different spectral settings, 17 of which provided complete coverage of the K, L, and M bands; whereas the remaining three, covering the H and J bands, are not included in the analysis (see Table 1). Twelve of the spectral settings acquired during ESO Period P115 and used in this analysis cover a wavelength range which was never observed before in HRS.

DateSpectral SettingDITNDITNfr†N_{\mathrm{fr}}^{\dagger}
25 Sep 2025L326230186
25 Sep 2025L334030186
25 Sep 2025M418751242
27 Sep 2025L342630186
27 Sep 2025M431851242
27 Sep 2025M446151242
28 Sep 2025L324430186
28 Sep 2025L330230186
28 Sep 2025M421151242
29 Sep 2025L337730186
29 Sep 2025M436851242
29 Sep 2025M450451242
30 Sep 2025L341230186
30 Sep 2025K2192101222
30 Sep 2025K2148101222
30 Sep 2025M426651242
01 Oct 2025M451951242
01 Oct 2025H1567*101222
01 Oct 2025H1582*101222
01 Oct 2025J1226*101022

Table 1. β\beta Pic observations acquired with CRIRES+ in September 2025 (Programme ID: 115.2820.001).

†: Number of frames in the run. *: These data sets were not included in this analysis.

A fixed instrumental setup was used for all epochs, with a slit width of 0.2″, yielding a spectral resolving power of R∼100 000R \sim100\,000. The slit was aligned with the edge-on plane of the planetary system (position angle, PA = 31.6°) to encompass the host star and all the planets, β\beta Pic b, c and d. We are currently analysing planets c and d and will present the results in a subsequent paper. The observations were performed using the MACAO adaptive optics system ([33]) and an ABBA nodding pattern to enable pairwise subtraction of the high sky background in the L and M bands. For all L-band settings, we adopted a direct integration time (DIT) of 30 s per exposure, with four exposures per nodding position over 11 nodding cycles. For the M-band settings, the DIT was 5 s, and the number of detector integrations (NDIT) was 12, with two exposures per nodding position over 11 nodding cycles. The two K-band observing blocks (OBs) used a DIT of 10 s and an NDIT of 12, with two exposures per nodding position over six nodding cycles. The J- and H-band settings acquired during this programme

are not included in the present analysis and will instead be used in a dedicated study of the circumstellar disc.

In addition to the newly acquired observations, we included 13 CRIRES+ L- and M-band datasets from Janson et al. [14], two CRIRES+ K-band datasets from Landman et al. [21], one CRIRES+ K-band dataset from González Picos et al. [10], two CRIRES+ M-band datasets from Parker et al. [31], two additional CRIRES+ M-band datasets from the CRIMSON survey (Programme ID: 114.27LL.002; PI: Parker; Parker et al. [30]), and seven H-band spectra from the High-Resolution Imaging and Spectroscopy of Exoplanets (HiRISE; Vigan et al. [42]) survey (Programme ID: 114.27C6.001; PI: Hoy).

Combining these observations yielded a dataset comprising 44 spectral settings and covering nearly the entire near-infrared wavelength range (1.5–5.5 μ\mum). This broad wavelength coverage enables a panchromatic analysis, which has been shown to maximise the S/N (Janson et al. [14]). To our knowledge, this represents the broadest high-resolution wavelength coverage obtained for a directly imaged exoplanet to date (see Figure 1).

Spectral ranges covered by detector orders in the H, K, L and M bands

Figure 1. From top to bottom: spectral ranges covered by detector orders in the grating settings used for this analysis, divided into H, K, L and M bands, respectively. The observations cover almost all of the four bands. The overplotted black line is the fractional atmospheric transmission. These plots were generated using CRIRESViz2

Data reduction

Given the broad wavelength coverage of our dataset, different reduction and extraction strategies were adopted depending on the spectral band.

L- and M-band

We reduced the L- and M-band data with the custom pipeline2 described in Janson et al. [14]. After bias, dark-current, and

flat-field correction, the pipeline derived the slit geometry and obtained the wavelength solution from telluric absorption and emission features.

The point spread function (PSF) subtraction scheme performed here is a variant of the STARK3^{3} routine previously implemented on VLT/UVES [36] and JWST/NIRCAM [32] data. Stellar PSF subtraction was performed on the pairwise background-subtracted images. First, a non-uniform grid was constructed in wavelength (dispersion direction) and spatial separation. A two-dimensional spline was then fitted in wavelength-separation space using a non-uniform grid of knots. The knot positions were chosen in normalised-flux space, ensuring that the spline model adequately samples regions containing both continuum and telluric features. Consequently, this naturally ensures that deep telluric features are represented in the spline model, and these features are removed together with the stellar contribution during the PSF subtraction. This procedure yielded a two-dimensional model of the stellar PSF, which was subsequently subtracted from the data.

To extract the planetary signal, we derived its position using the whereistheplanet package [44] and extracted its spectrum using the same spatial PSF profile derived during the construction of the stellar PSF model. We performed the same extraction at the same projected separation on the opposite side of the star, where no planetary signal is expected, to provide a reference spectrum for estimating the noise properties in subsequent analyses. We divided the obtained spectra by the stellar one to obtain the data in contrast space.

K-band

Together with the newly observed K-band data and the data from Landman et al. [21], which were all taken with the slit aligned with the system’s orbital plane, we also collected a K-band spectrum from González Picos et al. [10], where the slit was set perpendicularly to the orbital plane to increase the integration time. The two sets of observations have been reduced separately.

Slit-aligned data

We reduced the slit-aligned K-band observations with pycrires4, following Landman et al. [21]. After the standard calibrations and wavelength refinement, we fitted a two-dimensional spline to the stellar PSF. The stellar flux was then measured through a weighted extraction within a 5-pixel aperture centred on the PSF. We then subtracted the stellar and residual systematic signals, and extracted the planetary spectrum using the fitted spatial profile at the predicted planet position. We divided the result by the stellar spectrum to obtain planet-to-star contrast.

Slit-perpendicular data

For a complete description of these K-band data, refer to González Picos et al. [10]. Briefly, we used the excalibur pipeline [47], which reduced the data using standard calibration, sky-subtraction, and frame-stacking procedures. The planetary spectrum was then extracted using a method optimised for faint companions against a strong slit background, with spatial rows weighted according to their cross-correlation strength.

H-band

We reduced the HiRISE observations with the pipeline described by Costes et al. [4] and Denis et al. [6], which combines the official CRIRES+ pipeline with custom routines for frame combination, extraction, filtering, and wavelength recalibration. The stellar and planetary spectra were extracted using a Gaussian-like profile and employing optimal extraction, as described in Denis et al. (2026). We then divided the planet’s spectra by the stellar spectrum to obtain planet-to-star contrast.

Analysis

To analyse β\beta Pic b data, we generated model templates and cross-correlated the reduced spectra with them to sum up and enhance the emission-line signal. In this section, we report the details of the model generation, the cross-correlation and the results.

Model generation

A detailed description of the model generation is provided in Janson et al. [14]. We used petitRADTRANS models (Mollière et al. [28, 27]) as templates, adopting the relatively well-constrained physical properties of planet b. The model atmosphere consisted of hydrogen and helium, which contribute to the atmospheric pressure but have no strong lines in the wavelength range considered here, together with each molecule searched for in the extracted spectra. We adopted the temperature-pressure profile presented in Guillot [12], assuming a negligible contribution from irradiation. For the β\beta Pic b model, we set the temperature to 1700 K, following Parker et al. [31], the surface gravity log⁡g\log g to 4.18, and the rotational velocity to 20 km s−1^{-1}. Finally, we fitted and subtracted the continuum from the model spectra to retain only the line information and rotationally broadened them to match the expected planetary spectra.

Cross-correlation

Each extracted spectrum is naturally divided into CRIRES+ spectral orders. Thus, each observation comprises 4–7 spectral chunks, which we cross-correlated with the models described in Sect. 4.1. The model templates span the full wavelength coverage of the data and were masked to match the wavelength range of each chunk. They were then Doppler-shifted to the planet’s radial velocity at the time of the observation. This procedure yielded a set of ∼250\sim250 cross-correlation functions (CCFs) per model. In the presence of a significant signal, the corresponding CCFs are expected to exhibit a peak centred at zero velocity. A coherent set of CCFs is required to combine the individual functions, increase the overall S/N, and recover the often hidden signals of more exotic species. A simple average over all spectral regions is not optimal because it does not account for variations in line density among the spectral settings or differences in atmospheric conditions between observing dates. We therefore divided the merging procedure into two steps.

First, we obtained the CCF for each observation (for example, L3262, acquired on 25 September 2025) by combining the CCFs from its spectral orders (for example, five chunks). In this first step, each chunk was weighted according to its expected information content. To estimate this information content, we defined a threshold (i.e. 0.1) for the model spectral gradient and counted the data points at which the derivative of the model spectrum exceeded this threshold (see Figure 2). Consequently, orders containing more spectral information contributed more strongly to the CCF of the corresponding spectral setting. This left us with 44 CCFs per model, one per observation. To merge the observations’ CCFs, we adopted two different methods depending on the molecule we were searching for.

Two-panel simulated planetary flux and contrast model

Figure 2. The upper panel shows, in blue, the simulated planetary flux containing CH4\mathrm{CH}_{4} and produced using petitRadTRANS. The model is baseline-subtracted. The bottom panel shows the planetary contrast, in yellow, derived from the upper panel and used in the cross-correlation procedure. The black dashed line represents the threshold (i.e. 0.1 in contrast space) used to select the regions containing information. The green areas denote where the model’s gradient exceeds the threshold (selected regions), while the red ones are discarded regions.

For molecular species expected to be detected, such as H2_2O and CO, we weighted each observation ii according to its measured S/N:

S/Ni=max⁡(CCFi[peak region])std⁡(antiCCFi[out-of-peak region])(1)\mathrm{S/N}_i = \frac{\max\left(\mathrm{CCF}_i[\text{peak region}]\right)}{\operatorname{std}\left(\text{antiCCF}_i[\text{out-of-peak region}]\right)} \tag*{(1)}

Here, the “peak region” corresponds to the expected width of the CCF peak, based on the rotational broadening of the planet and the peak biasing (∼70\sim70 km s−1^{-1}; Parker et al. [31]), while the antiCCFi_i “out-of-peak region” comprises the complementary range in velocity space, at the same projected separation from the star as the planet but on the opposite side (as in Landman et al. [21]), and therefore provides a reference estimate of the systematic noise. Thus, we assigned higher S/N observations greater weight when combining the CCFs. After merging all CCFs, we evaluated the total S/N using Eq. (1).

For molecular species whose detectability was uncertain, such as NH3_3 and CH4_4, we adopted a more conservative weighting scheme to avoid artificially enhancing the combined S/N. Using the H2_2O analysis, we derived a correction factor for each observation:

Ci=S/NH2O,i⋅σH2O,i∑infoH2O,i(2)C_i = \mathrm{S/N}_{\mathrm{H_2O},i} \cdot\frac{\sigma_{\mathrm{H_2O},i}}{\sum\mathrm{info}_{\mathrm{H_2O},i}} \tag*{(2)}

where σH2O,i\sigma_{\mathrm{H_2O},i} is the median of the pixel-by-pixel uncertainties and ∑infoH2O,i\sum\mathrm{info}_{\mathrm{H_2O},i} is the total information content of the observation’s wavelength range, as defined above. The correction factor therefore corresponds to the ratio between the measured H2_2O S/N and that expected from the information content and estimated uncertainties. For a new molecular species, we multiplied CiC_i by the ratio between its information content and estimated uncertainty:

wX,i=Ci⋅∑infoX,iσX,i(3)w_{X,i} = C_i \cdot\frac{\sum\mathrm{info}_{X,i}}{\sigma_{X,i}} \tag*{(3)}

where XX denotes the molecule of interest. The resulting quantity provides an observation-specific estimate of the expected S/N and was used as the weight when combining the CCFs. This approach assigns the weights independently of the measured CCF of the molecule being searched for, thereby reducing the risk of artificially increasing the significance of a noise fluctuation. We adopted H2_2O as the reference species because of its exceptionally strong detection in our data (see Sect. 4.3). When searching for molecular species whose detectability was uncertain, we found a higher noise level at the planetary position than at the corresponding position on the opposite side of the star. This excess may arise from residual systematics introduced by the model’s imperfections or by degeneracies within the molecular lines. We emphasise that the model was constructed specifically for molecular detection and is not intended to provide a physically or chemically self-consistent representation of the planetary atmosphere, which will be the goal of an upcoming in-depth analysis. We therefore adopted a more conservative noise estimate. We autocorrelated the model, normalised the resulting autocorrelation function to the observed CCF, and subtracted it from the observed CCF. For CH4_4 and NH3_3, we therefore replaced the denominator of Eq. (1) with the standard deviation of the residuals obtained after subtracting the scaled model autocorrelation, excluding the peak region.

Results

The analysis yielded strong molecular signals. Consistent with previous studies (e.g. Landman et al. [21]; Janson et al. [14]), H2_2O produced the strongest signal. Thanks to the broad wavelength coverage of our dataset, we detected H2_2O at an S/N of ∼200\sim200 (upper-left panel of Figure 3). The HiRISE H-band data provide the largest contribution, accounting for most of the total signal. Nevertheless, the K-, L-, and M-band data alone yield an S/N of ∼50\sim50, demonstrating the substantial gain obtained by combining datasets covering different spectral settings rather than repeatedly observing the same settings. Similarly, we detected CO at an S/N of ∼50\sim50 (upper-right panel of Figure 3). For this molecule, the slit-perpendicular K-band data contributed the most to the final S/N, also more than the slit-aligned data at the same wavelengths. This finding is in agreement with what is presented in González Picos et al. [10], which demonstrated that orienting the slit perpendicular to the orbital plane, using it as a pseudo-coronagraph, allows for a higher integration time, resulting in higher S/N.

Cross-correlation functions of β Pic b with H2O, CO, CH4, and NH3

Figure 3. Cross-correlation functions of β\beta Pic b with an H2O\mathrm{H}_{2}\mathrm{O} (upper-left panel), CO (upper-right panel), CH4\mathrm{CH}_{4} (bottom-left panel), and NH3\mathrm{NH}_{3} (bottom-right panel). The black lines are the cross-correlation functions at β\beta Pic b’s location; the grey lines are the corresponding functions at the same projected separation on the opposite side of the star; meanwhile, the brown lines show the residuals after subtraction of the scaled model autocorrelation from the planetary CCF.

Before searching for additional molecular species, we removed the contribution of H2_2O from the H-band data because its strong spectral features affected the cross-correlation results for less abundant species, including by generating spurious CCF peaks. We removed the H2_2O contribution using an iterative subtraction procedure. Specifically, from each H-band chunk (as defined in Sect. 4.2), we subtracted a scaled, normalised H2_2O model shifted to the expected planetary radial velocity for that observation and repeated the cross-correlation analysis, varying the scaling factor to minimise the residual H2_2O CCF signal. The resulting H-band residuals, from which the dominant H2_2O contribution had been removed, improved the sensitivity of our searches for molecular species other than H2_2O and CO.

We searched for a broad range of molecular species (e.g. CO2_2, SiO, PH3_3, HCN, O3_3, etc.), including both those expected and those not expected to be present, to test our detection framework. Most species were not detected, with the exception of CH4_4 and NH3_3. We identify candidate CH4_4 and NH3_3 signals with CCF S/N values of approximately 5.3 and 5.7, respectively (bottom-left and bottom-right panels of Figure 3). The NH3_3 CCF peaks at the expected velocity of β Pic b, while the CH4_4 peak is offset by approximately 15 km s−1^{-1}, which is small compared with the width of the CCF peak (∼70\sim70 km s−1^{-1}). Applying the same weighting and combination procedure to spectra extracted at the corresponding position on the opposite side of the star produces no comparable peak, with absolute CCF S/N values below 2.6 for CH4_4 and 1.6 for NH3_3. Before H2_2O subtraction, the CH4_4 CCF contained an additional spurious peak near 80 km s−1^{-1}, which we traced to the strong H-band H2_2O signal. This emphasises the importance of the water-subtraction procedure. The HiRISE H-band and slit-perpendicular K-band observations received the largest weights and are primarily responsible for the detection of both species. This is due to the significant difference in S/N between the acquisition methods (i.e. slit-aligned spectroscopy vs fibre-fed/slit-perpendicular spectroscopy), showing that, when looking for lower-abundance species, it may be preferable to prioritise the S/N (e.g. observing with HiRISE) more than observing at the wavelength of the fundamental band (e.g. 3.3 μm for CH4_4). However, the contrast with the L- and M-band weights is less pronounced than for H2_2O and CO. This highlights the strength of a multi-band analysis when searching for weak molecular signals. To our knowledge, these are the first reported candidate CH4_4 and NH3_3 signals in the atmosphere of β Pic b, and one of the first tentative detections of these species using HRS in an exoplanet’s atmosphere. Under the assumption of chemical equilibrium, theoretical models (e.g. Burrows et al. [3]; Sudarsky et al. [40]; Lodders & Fegley [23]) predict that, at the pressure and temperature expected for β\beta Pic b, the atmosphere should be dominated by H2\mathrm{H}_{2} and He, with H2O\mathrm{H}_{2}\mathrm{O}, CO, and N2\mathrm{N}_{2} as the main molecular species. This scenario is consistent with the high-significance detections of H2O\mathrm{H}_{2}\mathrm{O} and CO, while also allowing NH3\mathrm{NH}_{3} and CH4\mathrm{CH}_{4} to be present at lower abundances. Non-equilibrium chemistry models predict low CH4\mathrm{CH}_{4} and NH3\mathrm{NH}_{3} abundances in low-surface-gravity planets. However, the high mass, temperature and pressure of β\beta Pic b are comparable to those of some brown dwarfs in which CH4\mathrm{CH}_{4} and NH3\mathrm{NH}_{3} are predicted to be present (e.g. Zahnle & Marley [46]; Phillips et al. [34]). Moreover, methane has also been detected in some of these objects. For instance, in the atmospheres of PSO318 (Liu et al. [22]) and LRL 11001, which have a similar mass to β\beta Pic b, Beiler et al. [1] and Luhman et al. [25] found CH4\mathrm{CH}_{4} absorptions and CH4\mathrm{CH}_{4} emissions, respectively. Furthermore, a CH4\mathrm{CH}_{4} detection would also be consistent with recent findings in the atmosphere of self-luminous exoplanets. For example, Denis et al. (2026) found methane signatures in 51 Eri b, a planet in the β\beta Pic moving group (BPMG) with similar mass and age to those of β\beta Pic b.

Although both signals have CCF S/N values exceeding 5, we classify them as tentative because this statistic does not fully account for possible residual systematics. A more comprehensive analysis, including atmospheric retrievals and additional tests of the robustness of the signals, is currently underway and will be used to assess whether the detections can be confirmed.

Conclusions

We combined 44 CRIRES+ settings spanning the H to M bands to conduct a panchromatic molecular search in β\beta Pic b. We recover H2_2O and CO at CCF S/N values of approximately 200 and 50 and identify tentative CH4_4 and NH3_3 signals near the expected planetary velocity, with CCF S/N values of 5.3 and 5.7. Equivalent extractions on the opposite side of the star show no comparable peaks. These results demonstrate the molecular sensitivity gained by combining high-resolution observations across multiple near-infrared bands. A low/medium-resolution analysis using, for instance, JWST/NIRSpec [13] or MIRI [35], would probably not be sensitive enough to detect lower-abundance species, such as CH4_4 and NH3_3. However, a low-resolution observational campaign might still be worthwhile to cover a larger wavelength range and complement the high-resolution observations. CH4_4 and NH3_3’s detections are in agreement with theoretical models and recent detections in brown dwarfs and self-luminous exoplanets. A forthcoming analysis will investigate the robustness and atmospheric implications of the candidate signals using additional validation tests and physically self-consistent atmospheric modelling.

Data availability

The reduced spectra, together with the models used for the cross-correlation, and the CCFs shown in Figure 3 are only available in electronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/.

Acknowledgements. We thank the referee for the comments and suggestions that helped us to improve the quality of our paper. The authors thank D. González Picos for lending part of the K-band data. A.Ba. thanks L. Parker for the insightful discussion. This project has received funding from Agence Nationale de la Recherche (ANR) under grant ANR-23-CE31-0006-01 (MIRAGES). The HiRISE instrument has been developed with funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme, grant agreements No. 757561 (HiRISE) and 678777 (ICARUS), from the Commission Spécialisée Astronomie-Astrophysique (CSAA) of CNRS/INSU, and from the Action Spécifique Haute Résolution Angulaire (ASHRA) of CNRS/INSU co-funded by CNES, and from Région Provence-Alpes-Côte d’Azur under grant agreement 2014-0276 (ASOREX).

References

  1. [1]Beiler, S. A., Allers, K. N., Cushing, M., et al. 2023, MNRAS, 518, 4870
  2. [2]Brandt, G. M., Brandt, T. D., Dupuy, T. J., Li, Y., & Michalik, D. 2021, AJ, 161, 179
  3. [3]Burrows, A., Marley, M., Hubbard, W. B., et al. 1997, ApJ, 491, 856
  4. [4]Costes, J., Denis, A., & Vigan, A. 2024, hipipe: VLT/HiRISE reduction pipeline, Astrophysics Source Code Library, record ascl:2407.019
  5. [6]Denis, A., Vigan, A., Costes, J., et al. 2025, A&A, 696, A6
  6. [7]Dorn, R. J., Bristow, P., Smoker, J. V., et al. 2023, A&A, 671, A24
  7. [8]Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al. 2016, A&A, 595, A1
  8. [9]Gibbs, A., Ruffio, J.-B., Bidot, A., et al. 2026, ApJ, 1006, L11
  9. [10]González Picos, D., Snellen, I. A. G., Landman, R., et al. 2026, A&A, 711, A87
  10. [11]Greene, T. P., Line, M. R., Montero, C., et al. 2016, ApJ, 817, 17
  11. [12]Guillot, T. 2010, A&A, 520, A27
  12. [13]Jakobsen, P., Ferruit, P., Alves de Oliveira, C., et al. 2022, A&A, 661, A80
  13. [14]Janson, M., Wehrung-Montpezat, J., Wehrhan, A., et al. 2025, A&A, 694, A63
  14. [15]Kaeufl, H.-U., Ballester, P., Biereichel, P., et al. 2004, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 5492, Ground-based Instrumentation for Astronomy, ed. A. F. M. Moorwood & M. M. Iye, 1218–1227
  15. [16]Kempton, E. M.-R. & Knutson, H. A. 2024, Reviews in Mineralogy and Geochemistry, 90, 411
  16. [17]Konopacky, Q. M., Barman, T. S., Macintosh, B. A., & Marois, C. 2013, Science, 339, 1398
  17. [18]Lacour, S., Wang, J. J., Rodet, L., et al. 2021, A&A, 654, L2
  18. [19]Lagrange, A.-M., Bonnefoy, M., Chauvin, G., et al. 2010, Science, 329, 57
  19. [20]Lagrange, A.-M., Meunier, N., Rubini, P., et al. 2019, Nature Astronomy, 3, 1135
  20. [21]Landman, R., Stolker, T., Snellen, I. A. G., et al. 2024, A&A, 682, A48
  21. [22]Liu, M. C., Magnier, E. A., Deacon, N. R., et al. 2013, ApJ, 777, L20
  22. [23]Lodders, K. & Fegley, B. 2002, Icarus, 155, 393
  23. [24]Luhman, K. 2024, The Astronomical Journal, 168, 159
  24. [25]Luhman, K. L., Alves de Oliveira, C., Faherty, J. K., & Smith, B. C. 2026, A&A, 714, L15
  25. [27]Mollière, P., Stolker, T., Lacour, S., et al. 2020, A&A, 640, A131
  26. [28]Mollière, P., Wardenier, J. P., van Boekel, R., et al. 2019, A&A, 627, A67
  27. [29]Nowak, M., Lacour, S., Lagrange, A.-M., et al. 2020, A&A, 642, L2
  28. [30]Parker, L. T., Birkby, J. L., Gandhi, S., et al. 2026, MNRAS[arXiv:2606.03824]arxiv.org/abs/2606.03824
  29. [31]Parker, L. T., Birkby, J. L., Landman, R., et al. 2024, MNRAS, 531, 2356
  30. [32]Patel, J. A., Brandeker, A., Kitzmann, R., et al. 2024, A&A, 690, A159
  31. [33]Paufique, J., Biereichel, P., Donaldson, R., et al. 2004, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 5490, Advancements in Adaptive Optics, ed. D. Bonaccini Calia, B. L. Ellerbroek, & R. Ragazzoni, 216–227
  32. [34]Phillips, M. W., Tremblin, P., Baraffe, I., et al. 2020, A&A, 637, A38
  33. [35]Rieke, G. H., Wright, G. S., Böker, T., et al. 2015, PASP, 127, 584
  34. [36]Ringqvist, S. C., Viswanath, G., Aoyama, Y., et al. 2023, A&A, 669, L12
  35. [37]Sing, D. K. 2018, in Astrophysics of Exoplanetary Atmospheres: 2nd Advanced School on Exoplanetary Science (Springer), 3–48
  36. [38]Snellen, I., de Kok, R., Birkby, J. L., et al. 2015, A&A, 576, A59
  37. [39]Snellen, I. A. G., Brandl, B. R., de Kok, R. J., et al. 2014, Nature, 509, 63
  38. [40]Sudarsky, D., Burrows, A., & Pinto, P. 2000, ApJ, 538, 885
  39. [41]Sutlieff, B. J., Bonse, M. J., Christiaens, V., et al. 2026, ApJ, 1006, L10
  40. [42]Vigan, A., El Morsy, M., Lopez, M., et al. 2024, A&A, 682, A16
  41. [43]Wang, J., Mawet, D., Hu, R., & Benneke, B. 2016, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 9911, Modeling, Systems Engineering, and Project Management for Astronomy VI, ed. G. Z. Angeli & P. Dierickx, 99112T
  42. [44]Wang, J. J., Kulkarni, M., & Blunt, S. 2021a, whereistheplanet: Predicting positions of directly imaged companions, Astrophysics Source Code Library, record ascl:2101.003
  43. [45]Wang, J. J., Ruffio, J.-B., Morris, E., et al. 2021b, AJ, 162, 148
  44. [46]Zahnle, K. J. & Marley, M. S. 2014, ApJ, 797, 41
  45. [47]Zhang, Y., González Picos, D., de Regt, S., et al. 2024, AJ, 168, 246

Paper details

vlnc:W_umdthrxk
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents