Assessing the star formation history of all-sky and part-sky 100pc white dwarf samples

TL;DR

The paper evaluates how representative 100 pc white dwarf samples from Gaia-based all-sky catalogues and part-sky MOS surveys constrain the local star formation history and Galactic disc age. It uses a population-synthesis framework to compare observed luminosity and absolute Gaia G magnitude distributions with four SFH forms, examining the impact of sample selection, atmospheric assumptions, and footprints. Across all samples, a late-peaking SFH (as in Roberts_2025) provides the best overall match, with disc ages around 10–11 Gyr, and the 40 pc sample serving as a robust benchmark for the larger 100 pc volumes. The study demonstrates that ongoing and upcoming MOS surveys (DESI, 4MOST) can meaningfully refine local SFH estimates despite incomplete skies and magnitude limits, while highlighting the persistent challenge of the faint end of the white dwarf population for age determinations.

Abstract

Thanks to Gaia and large-scale spectroscopic follow-up surveys (4MOST, DESI, WEAVE, SDSS-V), it is now possible to build representative and minimally biased samples of the local white dwarf population. Here we analyse several volume-limited 100pc samples of white dwarfs, constructed from different surveys and studies, to evaluate their completeness and residual biases. We model the underlying star formation history and Galactic disc age via comparison with simulated populations of white dwarfs to quantitatively characterise completeness. We assess whether the benefit of Gaia XP spectra in datasets outweighs the reduction in sample size, and to what extent targeted, part-sky, and magnitude limited surveys can be used in comparison to all-sky volume limited surveys. Additionally, we simulate the 4MOST 100PC sub-survey and discuss its use to better understand the local star formation history.

Figures (13)

• Figure 1: The probability of each white dwarf candidate in the GentileFusillo_2021 sample lying within the 100 pc volume based on its determined parallax and uncertainty on its parallax. Fewer than 2 000 objects outside the 100 pc limit have any probability of being within 100 pc within 1$\sigma$, and all of them have $P_{100 \rm pc}<0.5$.
• Figure 2: Top: Luminosity function for each of the three all-sky 100 pc samples of GentileFusillo_2021GCNS_2021Vincent_2024 and the 40 pc sample OBrien_2024Roberts_2025. We also include the Poisson uncertainties for each bin. Bottom: Absolute Gaia G magnitude distribution for each of the all-sky 100 pc samples and the 40 pc case, with Poisson uncertainties for each bin. The bump at $\approx$12.5 mag in the 40 pc sample does not appear in any of the 100 pc samples at the $3\sigma$ level and so appears to be an artifact of the smaller sample size rather than a real feature.
• Figure 3: Top: The footprint in RA and Dec (degrees) of the SDSS sample taken from Kilic_2025_100pc, and the simplified footprint used to generate our simulation of the SDSS sample. Our intention was not to match the footprints completely, but mostly simulate the same parts of the sky. Middle: The footprint in RA and Dec of the DESI sample taken from DESI_2025. Bottom: The simulated 4MOST footprint based on the planned sky coverage of the instrument.
• Figure 4: Top: Results from the luminosity function method with population synthesis using all four assumed star formation forms (constant; Roberts_2025 direct age method; Fantin_2019; Mor_2019) compared to observed 100 pc white dwarf sample of GentileFusillo_2021. Error bars are the standard deviation of 100 runs of the simulations, and Poisson errors for the observed sample. Bottom: Results from the absolute Gaia G magnitude method with all four assumed star formation forms compared to the observed 100 pc sample of GentileFusillo_2021. Error bars follow the same convention as above.
• Figure 5: Similar to Fig. \ref{['fig:GF_results']}, but for the 100 pc GCNS white dwarf catalogue GCNS_2021.