Search

Over a quarter of white dwarfs have photospheric metal pollution, which is evidence for recent accretion of exoplanetary material. While a wide range of mechanisms have been proposed to account for this pollution, there are currently few observational constraints to differentiate between them. To investigate the driving mechanism, we observe a sample of polluted and non-polluted white dwarfs in wide binary systems with main-sequence stars. Using the companion stars' metallicities as a proxy for the white dwarfs' primordial metallicities, we compare the metallicities of polluted and non-polluted systems. Because there is a well-known correlation between giant planet occurrence and higher metallicity (with a stronger correlation for close-in and eccentric planets), these metallicity distributions can be used to probe the role of gas giants in white dwarf accretion. We find that the metallicity distributions of polluted and non-polluted systems are consistent with the hypothesis that both samples have the same underlying metallicity distribution. However, we note that this result is likely biased by several selection effects. Additionally, we find no significant trend between white dwarf accretion rates and metallicity. These findings suggest that giant planets are not the dominant cause of white dwarf accretion events in binary systems., Comment: 17 pages, 9 figures, accepted to MNRAS

This paper introduces the front-propagation algorithm, a novel eXplainable AI (XAI) technique designed to elucidate the decision-making logic of deep neural networks. Unlike other popular explainability algorithms such as Integrated Gradients or Shapley Values, the proposed algorithm is able to extract an accurate and consistent linear function explanation of the network in a single forward pass of the trained model. This nuance sets apart the time complexity of the front-propagation as it could be running real-time and in parallel with deployed models. We packaged this algorithm in a software called $\texttt{front-prop}$ and we demonstrate its efficacy in providing accurate linear functions with three different neural network architectures trained on publicly available benchmark datasets., Comment: 14 pages, 6 figures. Accepted for publication in: Barnabas Bede, Kelly Cohen, and Vladik Kreinovich (eds.), Proceedings of the NAFIPS International Conference on Fuzzy Systems, Soft Computing, and Explainable AI. NAFIPS'2024, South Padre Island, Texas, May 27-29, 2024

We present a catalogue of 62 polluted white dwarfs observed by the 9th Low-Resolution Data Release of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST LRS DR9v1; R$\approx$1,800) and the Early Data Release 3 (EDR3) of the Gaia Mission. Among these stellar remnants, 30 are new discoveries with previously unknown traces of calcium pollution. To generate our catalogue, we used a database of 4,324 unique Gaia EDR3 white dwarf candidates with LAMOST LRS DR9v1 observations, many of which have been spectroscopically confirmed by other telescopes. For these stars, we developed a quantitative method to detect calcium absorption in their spectra between 3,900-4,000$\mathring {\mathrm A}$, which we then validated through visual inspection and multiple literature cross-checks. Our catalogue provides the astrometric and photometric properties of the white dwarf candidates, incorporates supplementary data (e.g. Montreal White Dwarf Database, MWDD; PanSTARRS; the Hubble Space Telescope), and indicates the possibility of calcium pollution in their atmospheres. For our final sample of polluted white dwarfs, we also determine the main atmospheric properties of 23 sources with effective temperatures $T_{\rm eff}$$\leq$25,000K and no existing calcium abundances in the MWDD. Our analysis represents a first step towards measuring the full atmospheric composition of these stars and learning about the bulk properties of their accreted material. As we venture into the era of wide-field spectroscopic surveys, our work highlights the importance of combining large-scale databases for identifying and characterising new polluted white dwarfs., Comment: 29 pages, 14 figures (+3 in the Appendix), 5 tables (+5 in the Appendix). Accepted for publication in MNRAS

Advancements in space telescopes have opened new avenues for gathering vast amounts of data on exoplanet atmosphere spectra. However, accurately extracting chemical and physical properties from these spectra poses significant challenges due to the non-linear nature of the underlying physics. This paper presents novel machine learning models developed by the AstroAI team for the Ariel Data Challenge 2023, where one of the models secured the top position among 293 competitors. Leveraging Normalizing Flows, our models predict the posterior probability distribution of atmospheric parameters under different atmospheric assumptions. Moreover, we introduce an alternative model that exhibits higher performance potential than the winning model, despite scoring lower in the challenge. These findings highlight the need to reevaluate the evaluation metric and prompt further exploration of more efficient and accurate approaches for exoplanet atmosphere spectra analysis. Finally, we present recommendations to enhance the challenge and models, providing valuable insights for future applications on real observational data. These advancements pave the way for more effective and timely analysis of exoplanet atmospheric properties, advancing our understanding of these distant worlds., Comment: Conference proceeding for the ECML PKDD 2023