arXiv subjects

João Faria

Publications and source records attributed to João Faria.

16 recordsLinked to original sources

arXiv2026

Super-Earth masses and stellar abundances from NIRPS reveal tentative evidence for water-rich formation around M dwarfs

Tracing the compositional link between terrestrial super-Earths and their host stars provides clues to their dominant formation pathway. By constraining the stellar abundances of refractory elements, we can predict the core mass fractions (CMFs) of their super-Earths. The level of agreement between this prediction and the planetary CMF derived from their masses and radii can reveal past formation processes, like mantle stripping and water-rich formation plus sequestration in the planet's core. Here, we present the first results from the Near Infrared Planet Searcher (NIRPS) GTO CMF subprogram: an intensive radial velocity campaign to refine masses and compute host stellar abundances of three hot super- Earths around M dwarfs (GJ 1132 b, GJ 1252 b, and LTT 3780 b), calculating masses of $1.69 \pm 0.15M_\oplus$, $1.54 \pm 0.18M_\oplus$, and $2.34 \pm 0.10M_\oplus$ respectively. We measure the CMFs of these and six further hot super-Earths with precise masses already available in the literature to 10-15% precision. We compare these to CMF predictions made from measuring the Fe, Mg, and Si abundances of their host stars measured from the NIRPS spectra. We find that the CMFs of these planets are smaller than expected from their host stellar abundances, to a statistically significant degree. This discrepancy is suggestive of significant reservoirs of water, and while these planets are too hot to harbor surface water, they likely have interior water mass fractions of $\sim$1%.

astro-ph.EP↗

arXiv2026

NIRPS tightens the mass estimate of GJ 3090 b and detects a planet near the stellar rotation period

We present an updated characterization of the planetary system orbiting the nearby M2 dwarf GJ 3090 (TOI-177; $d = 22$ pc), based on new high-precision radial velocity (RV) observations from NIRPS and HARPS. With an orbital period of 2.85 d, the transiting sub-Neptune GJ 3090 b has a mass we refine to $4.52 \pm 0.47 M_{\oplus}$, which, combined with our derived radius of $2.18 \pm 0.06 R_{\oplus}$, yields a density of $2.40^{+0.33}_{-0.30}$ g cm$^{-3}$. The combined interior structure and atmospheric constraints indicate that GJ 3090 b is a compelling water-world candidate, with a volatile-rich envelope in which water likely represents a significant fraction. We also confirm the presence of a second planet, GJ 3090 c, a sub-Neptune with a 15.9 d orbit and a minimum mass of $10.0 \pm 1.3 M_{\oplus}$, which does not transit. Despite its proximity to the star's 18 d rotation period, our joint analysis using a multidimensional Gaussian process (GP) model that incorporates TESS photometry and differential stellar temperature measurements distinguishes this planetary signal from activity-induced variability. In addition, we place new constraints on a non-transiting planet candidate with a period of 12.7 d, suggested in earlier RV analyses. This candidate remains a compelling target for future monitoring. These results highlight the crucial role of multidimensional GP modelling in disentangling planetary signals from stellar activity, enabling the detection of a planet near the stellar rotation period that could have remained undetected with traditional approaches.

astro-ph.EP↗

arXiv2025

TOI-3288 b and TOI-4666 b: two gas giants transiting low-mass stars characterised by NIRPS

Gas giant planets orbiting low-mass stars are uncommon outcomes of planet formation. Increasing the sample of well-characterised giants around early M dwarfs will enable population-level studies of their properties, offering valuable insights into their formation and evolutionary histories. We aim to characterise giant exoplanets transiting M dwarfs identified by TESS. High-resolution spectroscopic data are obtained in the optical and nIR, combining HARPS and NIRPS. We derive RVs via the cross-correlation function and implement a novel post-processing procedure to further mitigate telluric contamination in the nIR. The resulting RVs are jointly fit with TESS and ground-based photometry to derive the orbital and physical parameters of the systems. We confirm two gas giants transiting the low-mass stars TOI-3288 A (K9V) and TOI-4666 (M2.5V). TOI-3288 A hosts a Hot Jupiter with a mass of $2.11\pm0.08~M_{\rm Jup}$ and a radius of $1.00 \pm 0.03~R_{\rm Jup}$, with an orbital period of 1.43 days ($T_{\rm eq} = 1059 \pm 20~{\rm K}$). TOI-4666 hosts a $0.70_{-0.06}^{+0.05}~M_{\rm Jup}$ warm Jupiter ($T_{\rm eq} = 713 \pm 14~{\rm K}$) with a radius of $1.11 \pm 0.04~R_{\rm Jup}$, and an orbital period of 2.91 days. We identify a decrease in planetary mass with spectral type, where late M dwarfs host less massive giant planets than early M dwarfs. More massive gas giants that deviate from this trend are preferentially hosted by more metal-rich stars. Furthermore, we find an increased binarity fraction among low-mass stars hosting gas giants, which may play a role in enhancing giant planet formation around low-mass stars. The observed population trends agree with theoretical expectations, where higher metallicity can compensate for lower disk masses, and wide binary systems may influence planet formation and migration through Kozai-Lidov cycles or disk instabilities.

astro-ph.EP↗

arXiv2025

NIRPS and TESS reveal a peculiar system around the M dwarf TOI-756: A transiting sub-Neptune and a cold eccentric giant

The Near InfraRed Planet Searcher (NIRPS) joined HARPS on the 3.6-m ESO telescope at La Silla Observatory in April 2023, dedicating part of its Guaranteed Time Observations (GTO) program to the radial velocity follow-up of TESS planet candidates to confirm and characterize transiting planets around M dwarfs. We report the first results of this program with the characterization of the TOI-756 system, which consists of TOI-756 b, a transiting sub-Neptune candidate detected by TESS, as well as TOI-756 c, an additional non-transiting planet discovered by NIRPS and HARPS. TOI-756 b is a 1.24-day period sub-Neptune with a radius of 2.81 $\pm$ 0.10 $R_\oplus$ and a mass of 9.8$^{+1.8}_{-1.6}$ $M_\oplus$. TOI-756 c is a cold eccentric (e$_c$ = 0.45 $\pm$ 0.01) giant planet orbiting with a period of 149.6 days around its star with a minimum mass of 4.05 $\pm$ 0.11 $M_\mathrm{jup}$. Additionally, a linear trend of 146$~\mathrm{m\,s}^{-1}\,\mathrm{yr}^{-1}$ is visible in the radial velocities, hinting at a third component, possibly in the planetary or brown dwarf regime. This system is unique in the exoplanet landscape, standing as the first confirmed example of such a planetary architecture around an M dwarf. With a density of 2.42 $\pm$ 0.49 g cm$^{-3}$, the inner planet, TOI-756 b, is a volatile-rich sub-Neptune. Assuming a pure H/He envelope, we inferred an atmospheric mass fraction of 0.023 and a core mass fraction of 0.27, which is well constrained by stellar refractory abundances derived from NIRPS spectra. It falls within the still poorly explored radius cliff and at the lower boundary of the Neptune desert, making it a prime target for a future atmospheric characterization with JWST to improve our understanding of this population.

astro-ph.EP↗

arXiv2025

Diving into the planetary system of Proxima with NIRPS -- Breaking the metre per second barrier in the infrared

We obtained 420 high-resolution spectra of Proxima, over 159 nights, using the Near Infra Red Planet Searcher (NIRPS). We derived 149 nightly binned radial velocity measurements with a standard deviation of 1.69 m/s and a median uncertainty of 55 cm/s, and performed a joint analysis combining radial velocities, spectroscopic activity indicators, and ground-based photometry, to model the planetary and stellar signals present in the data, applying multi-dimensional Gaussian process regression to model the activity signals. We detect the radial velocity signal of Proxima b in the NIRPS data. All planetary characteristics are consistent with those previously derived using visible light spectrographs. In addition, we find evidence of the presence of the sub-Earth Proxima d in the NIRPS data. When combining the data with the HARPS observations taken simultaneous to NIRPS, we obtain a tentative detection of Proxima d and parameters consistent with those measured with ESPRESSO. By combining the NIRPS data with simultaneously obtained HARPS observations and archival data, we confirm the existence of Proxima d, and demonstrate that its parameters are stable over time and against change of instrument. We refine the planetary parameters of Proxima b and d, and find inconclusive evidence of the signal attributed to Proxima c (P = 1900 d) being present in the data. We measure Proxima b and d to have minimum masses of 1.055 $\pm$ 0.055 Me, and 0.260 $\pm$ 0.038 Me, respectively. Our results show that, in the case of Proxima, NIRPS provides more precise radial velocity data than HARPS, and a more significant detection of the planetary signals. The standard deviation of the residuals of NIRPS after the fit is 80 cm/s, showcasing the potential of NIRPS to measure precise radial velocities in the near-infrared.

astro-ph.EP↗

arXiv2025

BEBOP VII. SOPHIE discovery of BEBOP-3b, a circumbinary giant planet on an eccentric orbit

Planetary systems orbiting close binaries are valuable testing grounds for planet formation and migration models. More detections with good mass measurements are needed. We present a new planet discovered during the BEBOP survey for circumbinary exoplanets using radial velocities. We use data taken with the SOPHIE spectrograph at the Observatoire de Haute-Provence, and perform a spectroscopic analysis to obtain high precision radial velocities. This planet is the first radial velocity detection of a previously unknown circumbinary system. The planet has a mass of $0.56$ $M_{Jup}$ and orbits its host binary in 550 days with an eccentricity of 0.25. Compared to most of the previously known circumbinary planets, BEBOP-3b has a long period (relative to the binary) and a high eccentricity. There also is a candidate outer planet with a $\sim1400$ day orbital period. We test the stability of potential further candidate signals inside the orbit of BEBOP-3b, and demonstrate that there are stable orbital solutions for planets near the instability region which is where the Kepler circumbinary planets are located. We also use our data to obtain independent dynamical masses for the two stellar components of the eclipsing binary using High Resolution Cross-Correlation Spectroscopy (HRCCS), and compare those results to a more traditional approach, finding them compatible with one another.

astro-ph.EP↗

arXiv2024

Resonant sub-Neptunes are puffier

A systematic, population-level discrepancy exists between the densities of exoplanets whose masses have been measured with transit timing variations (TTVs) versus those measured with radial velocities (RVs). Since the TTV planets are predominantly nearly resonant, it is still unclear whether the discrepancy is attributed to detection biases or to astrophysical differences between the nearly resonant and non resonant planet populations. We defined a controlled, unbiased sample of 36 sub-Neptunes characterised by Kepler, TESS, HARPS, and ESPRESSO. We found that their density depends mostly on the resonant state of the system, with a low probability (of $0.002_{-0.001}^{+0.010}$) that the mass of (nearly) resonant planets is drawn from the same underlying population as the bulk of sub-Neptunes. Increasing the sample to 133 sub-Neptunes reveals finer details: the densities of resonant planets are similar and lower than non-resonant planets, and both the mean and spread in density increase for planets that are away from resonance. This trend is also present in RV-characterised planets alone. In addition, TTVs and RVs have consistent density distributions for a given distance to resonance. We also show that systems closer to resonances tend to be more co-planar than their spread-out counterparts. These observational trends are also found in synthetic populations, where planets that survived in their original resonant configuration retain a lower density; whereas less compact systems have undergone post-disc giant collisions that increased the planet's density, while expanding their orbits. Our findings reinforce the claim that resonant systems are archetypes of planetary systems at their birth.

astro-ph.EP↗

arXiv2024

NIRPS first light and early science: breaking the 1 m/s RV precision barrier at infrared wavelengths

The Near-InfraRed Planet Searcher or NIRPS is a precision radial velocity spectrograph developed through collaborative efforts among laboratories in Switzerland, Canada, Brazil, France, Portugal and Spain. NIRPS extends to the 0.98-1.8 $μ$m domain of the pioneering HARPS instrument at the La Silla 3.6-m telescope in Chile and it has achieved unparalleled precision, measuring stellar radial velocities in the infrared with accuracy better than 1 m/s. NIRPS can be used either stand-alone or simultaneously with HARPS. Commissioned in late 2022 and early 2023, NIRPS embarked on a 5-year Guaranteed Time Observation (GTO) program in April 2023, spanning 720 observing nights. This program focuses on planetary systems around M dwarfs, encompassing both the immediate solar vicinity and transit follow-ups, alongside transit and emission spectroscopy observations. We highlight NIRPS's current performances and the insights gained during its deployment at the telescope. The lessons learned and successes achieved contribute to the ongoing advancement of precision radial velocity measurements and high spectral fidelity, further solidifying NIRPS' role in the forefront of the field of exoplanets.

astro-ph.IM↗

arXiv2023

Improving circumbinary planet detections by fitting their binary's apsidal precession

Apsidal precession in stellar binaries is the main non-Keplerian dynamical effect impacting the radial-velocities of a binary star system. Its presence can notably hide the presence of orbiting circumbinary planets because many fitting algorithms assume perfectly Keplerian motion. To first order, apsidal precession ($\dotω$) can be accounted for by adding a linear term to the usual Keplerian model. We include apsidal precession in the kima package, an orbital fitter designed to detect and characterise planets from radial velocity data. In this paper, we detail this and other additions to kima that improve fitting for stellar binaries and circumbinary planets including corrections from general relativity. We then demonstrate that fitting for $\dotω$ can improve the detection sensitivity to circumbinary exoplanets by up to an order of magnitude in some circumstances, particularly in the case of multi-planetary systems. In addition, we apply the algorithm to several real systems, producing a new measurement of aspidal precession in KOI-126 (a tight triple system), and a detection of $\dotω$ in the Kepler-16 circumbinary system. Although apsidal precession is detected for Kepler-16, it does not have a large effect on the detection limit or the planetary parameters. We also derive an expression for the precession an outer planet would induce on the inner binary and compare the value this predicts with the one we detect.

astro-ph.EP↗

arXiv2022

The AEROS ocean observation mission and its CubeSat pathfinder

AEROS aims to develop a nanosatellite as a precursor of a future system of systems, which will include assets and capabilities of both new and existing platforms operating in the Ocean and Space, equipped with state-of-the-art sensors and technologies, all connected through a communication network linked to a data gathering, processing and dissemination system. This constellation leverages scientific and economic synergies emerging from New Space and the opportunities in prospecting, monitoring, and valuing the Ocean in a sustainable manner, addressing the demand for improved spatial, temporal, and spectral coverage in areas such as coastal ecosystems management and climate change assessment and mitigation. Currently, novel sensors and systems, including a miniaturized hyperspectral imager and a flexible software-defined communication system, are being developed and integrated into a new versatile satellite structure, supported by an innovative on-board software. Additional sensors, like the LoRaWAN protocol and a wider field of view RGB camera, are under study. To cope with data needs, a Data Analysis Centre, including a cloud-based data and telemetry dashboard and a back-end layer, to receive and process acquired and ingested data, is being implemented to provide tailored-to-use remote sensing products for a wide range of applications for private and institutional stakeholders.

physics.ao-ph↗

arXiv2022

AEROS: Oceanographic Hyperspectral Imaging and Argos-Tracking CubeSat

AEROS is a 3U CubeSat pathfinder toward a future ocean-observing constellation, targeting the Portuguese Atlantic region. AEROS features a miniaturized, high-resolution Hyperspectral Imager (HSI), a 5MP RGB camera, and a Software Defined Radio (SDR). The sensor generated data will be processed and aggregated for end-users in a new web-based Data Analysis Center (DAC). The HSI has 150 spectrally contiguous bands covering visible to near-infrared with 10 nm bandwidth. The HSI collects ocean color data to support studies of oceanographic characteristics known to influence the spatio-temporal distribution and movement behavior of marine organisms. Usage of an SDR expands AEROS's operational and communication range and allows for remote reconfiguration. The SDR receives, demodulates, and retransmits short duration messages, from sources including tagged marine organisms, autonomous vehicles, subsurface floats, and buoys. The future DAC will collect, store, process, and analyze acquired data, taking advantage of its ability to disseminate data across the stakeholders and the scientific network. Correlation of animal-borne Argos platform locations and oceanographic data will advance fisheries management, ecosystem-based management, monitoring of marine protected areas, and bio-oceanographic research in the face of a rapidly changing environment. For example, correlation of oceanographic data collected by the HSI, geolocated with supplementary images from the RGB camera and fish locations, will provide researchers with near real-time estimates of essential oceanographic variables within areas selected by species of interest.

physics.ao-ph↗

arXiv2022

Enriching a Fashion Knowledge Graph from Product Textual Descriptions

Knowledge Graphs offer a very useful and powerful structure for representing information, consequently, they have been adopted as the backbone for many applications in e-commerce scenarios. In this paper, we describe an application of existing techniques for enriching thelarge-scale Fashion Knowledge Graph (FKG) that we build at Farfetch. In particular, we apply techniques for named entity recognition (NER) and entity linking (EL) in order to extract and link rich metadata from product textual descriptions to entities in the FKG. Having a complete and enriched FKG as an e-commerce backbone can have a highly valuable impact on downstream applications such as search and recommendations. However, enriching a Knowledge Graph in the fashion domain has its own challenges. Data representation is different from a more generic KG, like Wikidata and Yago, as entities (e.g. product attributes) are too specific to the domain, and long textual descriptions are not readily available. Data itself is also scarce, as labelling datasets to train supervised models is a very laborious task. Even more, fashion products display a high variability and require an intricate ontology of attributes to link to. We use a transfer learning based approach to train an NER module on a small amount of manually labeled data, followed by an EL module that links the previously identified named entities to the appropriate entities within the FKG. Experiments using a pre-trained model show that it is possible to achieve 89.75% accuracy in NER even with a small manually labeled dataset. Moreover, the EL module, despite relying on simple rule-based or ML models (due to lack of training data), is able to link relevant attributes to products, thus automatically enriching the FKG.

cs.IR↗

arXiv2021

BEBOP III. Observations and an independent mass measurement of Kepler-16 (AB) b -- the first circumbinary planet detected with radial velocities

The radial velocity method is amongst the most robust and most established means of detecting exoplanets. Yet, it has so far failed to detect circumbinary planets despite their relatively high occurrence rates. Here, we report velocimetric measurements of Kepler-16A, obtained with the SOPHIE spectrograph, at the Observatoire de Haute-Provence's 193cm telescope, collected during the BEBOP survey for circumbinary planets. Our measurements mark the first radial velocity detection of a circumbinary planet, independently determining the mass of Kepler-16~(AB)~b to be $0.313 \pm 0.039\,{\rm M}_{\rm Jup}$, a value in agreement with eclipse timing variations. Our observations demonstrate the capability to achieve photon-noise precision and accuracy on single-lined binaries, with our final precision reaching $\rm 1.5~m\,s^{-1}$ on the binary and planetary signals. Our analysis paves the way for more circumbinary planet detections using radial velocities which will increase the relatively small sample of currently known systems to statistically relevant numbers, using a method that also provides weaker detection biases. Our data also contain a long-term radial velocity signal, which we associate with the magnetic cycle of the primary star.

astro-ph.EP↗

arXiv2021

Scheduling strategies for the ESPRESSO follow-up of TESS targets

Radial-velocity follow-up of stars harbouring transiting planets detected by TESS is expected to require very large amounts of expensive telescope time in the next few years. Therefore, scheduling strategies should be implemented to maximize the amount of information gathered about the target planetary systems. We consider myopic and non-myopic versions of a novel uniform-in-phase scheduler, as well as a random scheduler, and compare these scheduling strategies with respect to the bias, accuracy and precision achieved in recovering the mass and orbital parameters of transiting and non-transiting planets. This comparison is carried out based on realistic simulations of radial-velocity follow-up with ESPRESSO of a sample of 50 TESS target stars, with simulated planetary systems containing at least one transiting planet with a radius below $4R_{\oplus}$. Radial-velocity datasets were generated under reasonable assumptions about their noise component, including that resulting from stellar activity, and analysed using a fully Bayesian methodology. We find the random scheduler leads to a more biased, less accurate, and less precise, estimation of the mass of the transiting exoplanets. No significant differences are found between the results of the myopic and non-myopic implementations of the uniform-in-phase scheduler. With only about 22 radial velocity measurements per dataset, our novel uniform-in-phase scheduler enables an unbiased (at the level of 1\%) measurement of the masses of the transiting planets, while keeping the average relative accuracy and precision around 16\% and 23\% respectively. The number of non-transiting planets detected is similar for all the scheduling strategies considered, as well as the bias, accuracy and precision with which their masses and orbital parameters are recovered.

astro-ph.EP↗

arXiv2018

A Unified Model with Structured Output for Fashion Images Classification

A picture is worth a thousand words. Albeit a cliché, for the fashion industry, an image of a clothing piece allows one to perceive its category (e.g., dress), sub-category (e.g., day dress) and properties (e.g., white colour with floral patterns). The seasonal nature of the fashion industry creates a highly dynamic and creative domain with evermore data, making it unpractical to manually describe a large set of images (of products). In this paper, we explore the concept of visual recognition for fashion images through an end-to-end architecture embedding the hierarchical nature of the annotations directly into the model. Towards that goal, and inspired by the work of [7], we have modified and adapted the original architecture proposal. Namely, we have removed the message passing layer symmetry to cope with Farfetch category tree, added extra layers for hierarchy level specificity, and moved the message passing layer into an enriched latent space. We compare the proposed unified architecture against state-of-the-art models and demonstrate the performance advantage of our model for structured multi-level categorization on a dataset of about 350k fashion product images.

cs.CV↗

arXiv2018

A Hierarchical Deep Learning Natural Language Parser for Fashion

This work presents a hierarchical deep learning natural language parser for fashion. Our proposal intends not only to recognize fashion-domain entities but also to expose syntactic and morphologic insights. We leverage the usage of an architecture of specialist models, each one for a different task (from parsing to entity recognition). Such architecture renders a hierarchical model able to capture the nuances of the fashion language. The natural language parser is able to deal with textual ambiguities which are left unresolved by our currently existing solution. Our empirical results establish a robust baseline, which justifies the use of hierarchical architectures of deep learning models while opening new research avenues to explore.

cs.IR↗