A Detailed Model Atmosphere Analysis of Hot White Dwarfs in DESI DR1

We present a detailed model atmosphere analysis of hot white dwarfs in the Dark Energy Spectroscopic Instrument (DESI) Data Release 1. Our sample includes 19,321 unique targets with $G_{\rm BP}-G_{\rm RP}\leq0$. We use the DESI spectra along with Gaia parallaxes and SDSS, Pan-STARRS, and SkyMapper photometry to perform spectroscopic and photometric fits. We find a significant discrepancy between the photometric and spectroscopic masses for DA white dwarfs (a systematic offset of 0.05-$0.06M_\odot$), indicating problems with the broad hydrogen line profiles in DESI spectroscopy data. Our photometric fits are consistent with a peak at the canonical mass of $0.6M_\odot$. A remarkable feature of the mass distribution is the prevalence of magnetic white dwarfs among the ultramassive DA population and that of warm DQs in the non-DA distribution. We identify 70 DQs in the DESI hot white dwarf sample, including 9 DAQs with carbon and hydrogen atmospheres. We constrain the ratio of non-DA to DA white dwarfs as a function of temperature, and discuss the implications for the spectral evolution of white dwarfs in the temperature range $10^5-10^4$ K. We also discuss unusual objects in the sample, including metal-rich white dwarfs and extremely low mass white dwarfs. This analysis provides the first look at the large sample of Gaia-selected white dwarf candidates that will be observed with multiplexed spectroscopic surveys like DESI, SDSS-V, 4MOST, and WEAVE over the next several years.

💡 Research Summary

This paper presents the first large‑scale, detailed model‑atmosphere analysis of hot white dwarfs observed in the Dark Energy Spectroscopic Instrument (DESI) Data Release 1 (DR1). By cross‑matching DESI DR1 with the Gaia DR3 white‑dwarf catalog, the authors assembled a sample of 19 321 unique hot white dwarfs with Gaia colour G_BP − G_RP ≤ 0, corresponding to effective temperatures ≳10 000 K where hydrogen and helium lines are clearly visible. The study combines DESI spectra with Gaia parallaxes and multi‑band photometry from SDSS, Pan‑STARRS, and SkyMapper (GALEX data are shown but not used in fits) to derive stellar parameters via two independent approaches: (1) a photometric method that fits broadband fluxes and solid angle using the known distance to infer radius and, through evolutionary models, mass and surface gravity; and (2) a spectroscopic method that fits normalized Balmer (for DA) or He I/He II (for DB, DO) line profiles to obtain effective temperature (T_eff) and log g, later converted to mass with the same evolutionary tracks.

The atmospheric models are LTE grids with extensions for NLTE effects and 3‑D corrections where appropriate. For DA white dwarfs, pure‑hydrogen models from Tremblay & Bergeron (2009) are employed, covering T_eff = 8 000–120 000 K and log g = 5.0–9.5, with 3‑D corrections from Tremblay et al. (2013, 2015). Hot DA, DA O, and DO stars are also modeled with mixed H/He compositions to account for metal contamination, although the grids remain metal‑free, which the authors note as a source of systematic error. DB/DBA stars are analyzed with Bergeron et al. (2011) models that include updated van der Waals broadening; warm DQ stars are fitted with He‑free C+H grids (Kilic et al. 2025b) because helium abundance cannot be constrained spectroscopically; metal‑rich DZ/DBZ objects use extended Blouin et al. (2018) grids with chondritic scaling of heavy elements. Evolutionary sequences from Bédard et al. (2020) (CO‑core) and Althaus et al. (2013) (He‑core) provide the mass–radius relations needed for both methods.

Photometric fits use all available ugriz, grizy, and uvgriz measurements, corrected for interstellar reddening using mean A_V values from Gentile Fusillo et al. (2021) and the Fitzpatrick (1999) law. Gaia G, BP, RP photometry is excluded from fits because the three bands are not independent. The χ² minimization is performed with a Levenberg–Marquardt algorithm; problematic photometry (e.g., blended sources in SkyMapper) is flagged and omitted.

Spectroscopic fits rely on DESI’s three spectrograph arms (3600–9800 Å) with a median signal‑to‑noise ratio of 11 (≈20 % of spectra have S/N ≥ 20). The authors visually inspect each fit, classify objects into DA, DB, DC, DO, DQ, DZ, magnetic, He‑DA, and other sub‑types (e.g., DA O, DBA, DBAZ, DZA).

Key results:

• The photometric mass distribution of DA white dwarfs peaks at the canonical 0.6 M_⊙, consistent with previous Gaia‑based studies.
• Spectroscopic masses for the same DAs are systematically lower by 0.05–0.06 M_⊙, indicating a systematic issue with DESI’s hydrogen‑line profiles (likely under‑estimation of line depth and omission of metal opacity).
• Among the ultramassive (M > 1.0 M_⊙) DA population, ~70 % are magnetic, confirming the known over‑representation of magnetism in high‑mass white dwarfs.
• The non‑DA/DA ratio rises toward lower temperatures, supporting spectral evolution models where helium‑rich atmospheres become more common as white dwarfs cool.
• 70 warm DQ stars are identified, including 9 DAQ objects that show both carbon and hydrogen features.
• Metal‑rich DZ/DBZ objects display Ca/He abundances from −12 to −6 and span T_eff = 4 000–19 500 K.
• A handful of extremely low‑mass (ELM) candidates with log g ≈ 5–6 are found, distinct from sub‑dwarfs and requiring follow‑up high‑resolution spectroscopy.

The authors discuss the implications of the DA mass offset for future large‑scale spectroscopic surveys (DESI, SDSS‑V, 4MOST, WEAVE). Correcting the hydrogen‑line systematics and incorporating metal‑contaminated model atmospheres will be essential for accurate white‑dwarf mass functions and for probing the spectral evolution across the 10⁵–10⁴ K temperature range. The paper also highlights the geographic bias of DESI (northern sky, avoidance of the Galactic plane) and the need for complementary observations to achieve a complete census of nearby white dwarfs.

In summary, this work demonstrates the power of combining Gaia astrometry with multiplexed spectroscopy to characterize tens of thousands of hot white dwarfs, uncovers a systematic spectroscopic bias in DESI data, and provides a benchmark dataset for refining white‑dwarf atmospheric models and evolutionary theories in the era of massive spectroscopic surveys.