With 130 hours on one telescope, a team measures primordial helium to half a percent

The precision came from giving up extrapolation. The older approach estimated the primordial value by extending a trend line through many galaxies and inferring the starting level. The uncertainty in the trend became the uncertainty in the answer ^[3]. The Large Binocular Telescope campaign instead spent its time on 15 of the most chemically pristine small, remote galaxies ever found, objects unevolved enough that the team calls them time capsules of the period shortly after the Big Bang ^[3][4].

Sub-percent spectroscopy is mostly a fight with systematics. Using spectrographs built at Ohio State University, the team analysed more than 10 helium lines and 15 hydrogen lines simultaneously. That let them account for small effects earlier analyses had treated as negligible ^[5][6]. "The MODS spectrographs took 12 years to build from conception to first light on sky," said Richard Pogge, a Distinguished Professor of Astronomy at Ohio State ^[10].

If 0.5% is three times better than the previous standard, that standard sat near 1.5% ^[1]. And 130 hours divided across 15 galaxies is roughly 8.7 hours each ^[2]. Faint, chemically bare dwarfs take that long to measure this way.

The precision yields a count. From the helium abundance, the team calculated the number of neutrino families present in the early universe ^[7]. "It has diagnostic power that speaks directly to the Standard Model of Physics," said Evan Skillman of the University of Minnesota Twin Cities ^[15], who also said, "We promised a half-percent uncertainty in our proposal, and we got there" ^[9]. The phys.org account of the work did not report the resulting neutrino number ^[16]. The collaboration lists ten institutions, among them the Universidad Nacional Autonoma de Mexico and TRIUMF ^[13][3].

Of the three pillars the Big Bang theory rests on, light-element abundance has been measured least thoroughly; cosmic expansion and the cosmic microwave background have had far more attention ^[11]. A 0.5% helium value is useful because of that neglect, and the same gap is why I would not expect the next factor of three to come from more hours on sky: once effects previously deemed negligible are inside the error budget ^[6], the limit moves to atomic physics and calibration. The first of the five papers, covering project description, sample selection, observations and methodology, appears in The Astrophysical Journal ^[14].