cecilia: A Machine Learning-Based Pipeline for Measuring Metal Abundances of Helium-rich Polluted White Dwarfs (2402.05176v1)

Abstract: Over the past several decades, conventional spectral analysis techniques of polluted white dwarfs have become powerful tools to learn about the geology and chemistry of extrasolar bodies. Despite their proven capabilities and extensive legacy of scientific discoveries, these techniques are however still limited by their manual, time-intensive, and iterative nature. As a result, they are susceptible to human errors and are difficult to scale up to population-wide studies of metal pollution. This paper seeks to address this problem by presenting cecilia, the first Machine Learning (ML)-powered spectral modeling code designed to measure the metal abundances of intermediate-temperature (10,000$\leq T_{\rm eff} \leq$20,000 K), Helium-rich polluted white dwarfs. Trained with more than 22,000 randomly drawn atmosphere models and stellar parameters, our pipeline aims to overcome the limitations of classical methods by replacing the generation of synthetic spectra from computationally expensive codes and uniformly spaced model grids, with a fast, automated, and efficient neural-network-based interpolator. More specifically, cecilia combines state-of-the-art atmosphere models, powerful artificial intelligence tools, and robust statistical techniques to rapidly generate synthetic spectra of polluted white dwarfs in high-dimensional space, and enable accurate ($\lesssim$0.1 dex) and simultaneous measurements of 14 stellar parameters -- including 11 elemental abundances -- from real spectroscopic observations. As massively multiplexed astronomical surveys begin scientific operations, cecilia's performance has the potential to unlock large-scale studies of extrasolar geochemistry and propel the field of white dwarf science into the era of Big Data. In doing so, we aspire to uncover new statistical insights that were previously impractical with traditional white dwarf characterisation techniques.

• The paper introduces cecilia, a machine learning pipeline that automates metal abundance measurements in helium-rich polluted white dwarfs with error margins below 0.1 dex.
• It leverages over 22,000 synthetic models and 13 stellar parameters to generate high-resolution spectra, matching the accuracy of traditional manual methods.
• cecilia's scalable, rapid approach enables statistically robust analysis of large white dwarf datasets from upcoming wide-field astronomical surveys.

Unveiling the Geochemistry of Extrasolar Worlds with cecilia: A ML Approach to Polluted White Dwarf Spectra

Polluted white dwarfs, those which show signs of metal accretion in their atmospheres, are unique cosmic laboratories for studying the geochemistry of extrasolar planetary material. By examining these remnants of once-active planetary systems, scientists can glean insights into the composition and evolution of planets orbiting stars other than our Sun. However, the manual and time-intensive nature of traditional spectroscopic analysis methods has hampered efforts to scale studies to the large populations of such stars revealed by astronomical surveys.

Addressing this bottleneck, a new ML pipeline, named cecilia, is introduced as a powerful tool for measuring the metal abundances in the atmospheres of intermediate-temperature, Helium-rich polluted white dwarfs directly from their spectral observations. This novel approach combines state-of-the-art atmosphere models with advanced ML techniques to automate and accelerate the spectrum-to-abundance mapping process, potentially transforming the field of white dwarf science from manual individual characterizations to automated large-scale studies.

The Performance and Potential of cecilia

The cecilia pipeline stands out for its ability to predict the abundances of 10 elements with an accuracy comparable to traditional methods but without requiring manual oversight. Trained on over 22,000 synthetic models covering a wide range of stellar parameters and pollution levels, cecilia efficiently generates high-resolution synthetic spectra from a set of 13 underlying stellar properties, including the star's effective temperature, surface gravity, the abundance of hydrogen, and that of 10 other metals.

Testing against both synthetic and real-world observations, such as the well-characterized SDSS spectrum of white dwarf WD 1232+563, has demonstrated cecilia's capability to rapidly and accurately retrieve stellar parameters. It identifies abundances with an error margin less than 0.1 dex for up to 10 heavy elements, showcasing its comparable performance to that of traditional, more labor-intensive techniques.

Designed to be both fast and scalable, cecilia promises to unlock the full potential of upcoming massive spectroscopic datasets from wide-field astronomical surveys. With the expected exponential growth in white dwarf observations, cecilia offers a practical solution to studying metal pollution on a scale previously unattainable, paving the way for statistically robust insights into the geology and chemistry of extrasolar material.

Future Directions and Improvements

While cecilia marks a significant advancement in the automated analysis of polluted white dwarfs, there is still room for refinement and expansion. Future iterations could benefit from a more extensive training dataset covering a broader range of stellar properties and metal abundances. Additionally, integrating ultraviolet spectral data and cooler white dwarf models could further enhance the pipeline's utility and predictive power.

Moreover, the implementation of a complementary ML model for estimating effective temperatures and surface gravities directly from the abundance predictions could yield a fully self-consistent fitting routine, eliminating reliance on external photometric observations for these critical parameters.

Conclusion

Cecilia represents a pivotal shift in the paper of metal pollution in white dwarfs, offering a scalable, accurate, and automated solution that is primed to make significant contributions to our understanding of ancient extrasolar geochemistry. As we move into an era of big data in astronomy, pipelines like cecilia will be instrumental in unlocking new scientific discoveries from the vast datasets collected by future astronomical surveys.