The day in conversation for 2026-11-10

discussion

Transcript

The discussion: Vera: That requirement for over one million photons per spectral bin just to hit that 0.1% polarization accuracy on GMT targets seems astronomical; how do we justify that photon budget when we’re looking at things like the faint stellar populations we analyzed last month? Jocelyn: It depends entirely on the source geometry and magnetic field complexity; if the spectral resolution is high enough, perhaps those photons are concentrated in a specific line profile, mitigating some of the noise floor you’re concerned about. Did they specify what kind of objects—stars or galaxies—this threshold applies to? Subrahmanyan: It seems tied specifically to resolving the structure around magnetic fields rather than just photometry; that pushes the requirement toward extremely high signal-to-noise ratio demands, which is something I remember struggling with when we were calibrating those early interferometric baseline measurements. Vera: Exactly, and that makes me wonder how this relates to the denoising work in The PAU Survey; if we can reduce scatter by 33% or 37% on faint galaxies just using flux-aware methods, does that less stringent photon threshold for polarimetry become achievable in practice? Jocelyn: Not directly, because polarimetry is fundamentally different—it’s about measuring the cross-correlation of light polarization states, not just intensity fluctuations. The PAU survey improves flux precision by handling repeated exposures; the GMT

More episodes

← Home