Blue horizontal branch stars: weighing the Milky Way, testing the assumptions

Why these stars connect the papers

In this selected coverage, blue horizontal branch (BHB) stars are tools for measuring the Milky Way’s gravitational mass, rather than primarily objects of stellar-evolution research. Their luminosity makes them observable far into the halo, and their colour provides an estimate of absolute magnitude and hence distance. Combined with DESI line-of-sight velocities and Gaia proper motions, they supply positions and velocities for dynamical modelling. The DESI overview is the natural starting point for that observational logic.

These four papers form a connected project, not four independent confirmations: an overview combines mass estimates, two companion analyses develop contrasting estimators, and a mock-catalogue paper supplies their shared validation framework. Their common question is how much the inferred mass depends on the tracer population, modelling assumptions and calibration—not simply how many BHB stars are available (overview; mock catalogues).

Two methods, different tracer comparisons

Read the Galactic Mass Estimator (GME) and empirical distribution function (emPDF) analyses together. GME uses hierarchical Bayesian inference, assumes power-law forms for the gravitational potential and tracer density, and accounts for distance and velocity uncertainties. Its selected outer-halo sample contains 279 BHB stars and 213 RR Lyrae stars at roughly 50–80 kiloparsecs from the Galactic centre. After mock-based correction, the enclosed masses at 80 kiloparsecs are respectively 0.64 ± 0.06 and 0.53 ± 0.05 trillion solar masses. BHBs give systematically higher masses across the analysed radial range; including Sagittarius stream stars lowers the estimates by about 10% (GME analysis).

By contrast, emPDF constructs the tracer distribution from the data without imposing its functional form. That flexibility is not assumption-free: the analysis still depends on steady state, spherical symmetry and an adopted NFW density profile. Unlike GME’s explicit uncertainty treatment, the current emPDF implementation does not explicitly model observational errors; their effects are calibrated with mocks. Its abstract reports overall consistency between the two tracer populations, with the RR Lyrae central virial-mass estimate above the BHB estimate—the opposite ordering from GME (emPDF analysis).

That contrast does not establish a contradiction or identify its cause. The methods differ in assumptions and error treatment, and a mass enclosed within 80 kiloparsecs is not the same quantity as a virial mass. GME’s estimates beyond approximately 80 kiloparsecs extrapolate its power-law potential; its virial masses require anchoring an NFW halo. Comparing headline numbers without tracking radius and model extension would obscure those distinctions (GME analysis).

Read the calibration paper before the combined result

The AuriDESI catalogues turn simulated galaxies into mock BHB and RR Lyrae surveys, adding selection effects, observational incompleteness and measurement uncertainties. They let the companion analyses compare recovered masses with known simulated values. But the mock-catalogue paper also documents mismatches: Auriga stellar haloes produce too many tracers, and the normalised radial distributions differ from DESI. Proposed explanations include the simulated stellar distributions and the stellar-evolution prescriptions; the excerpts do not isolate a single cause (mock-catalogue study).

This makes calibration transfer an unresolved question: how faithfully do biases measured in those mocks represent biases in the actual Galaxy? The overview uses 24 halo–viewpoint combinations from six simulated haloes, while noting that views of the same halo may remain correlated. It explicitly cautions that its calibrated joint inference relies on mock realism (overview calibration).

The overview combines four methods while accounting for biases and residual correlations. Its abstract reports 0.636 trillion solar masses inside 80 kiloparsecs, with uncertainties of +0.053/−0.048 trillion, and similar results when combining only BHB-based or only RR Lyrae-based measurements. This combined agreement does not erase GME’s tracer offset or make the component estimates independent (joint analysis).

Evidence boundaries

All four supplied records contain selected paper-text excerpts plus original abstracts, rather than abstract-only evidence. Nevertheless, extraction visibly disrupts equations and author lists; numerical statements here use the clean supplied abstracts where available. Abstract-level results support reported outcomes, not unshown implementation details. Missing material in these excerpts is not evidence that a paper omits it.

The supplied versions are arXiv v1 preprints with draft/preprint headers; these records do not establish subsequent peer-review status. Together they offer a focused reading route through one DESI effort—not evidence of field-wide consensus on BHB tracers or the Milky Way’s mass.