= Supernova discovery of accelerating expansion
The 1998 <High-Z Supernova Search Team> paper by <Adam Riess>, <Brian Schmidt> and collaborators presented <supernova> evidence for acceleration.<Observational Evidence from Supernovae for an Accelerating Universe and a Cosmological Constant by Adam Riess and others>[Byline and abstract, PDF pp.1–2]{ref}
The 1999 <Supernova Cosmology Project> paper by <Saul Perlmutter> and collaborators independently analysed 42 distant events. <Alexei V. Filippenko> coauthored both papers.<Measurements of Omega and Lambda from 42 High-Redshift Supernovae by Saul Perlmutter and others>[Byline and abstract, PDF p.1]{ref}
The <High-Z Supernova Search Team> search used the CTIO Blanco 4-m telescope and a $2048\times2048$ <charge-coupled device>, comparing fields across dark runs. Difference images located rising candidates; Keck, MMT and ESO <spectroscopy> classified them and measured host <redshifts>. Follow-up <photometry> used filters approximating emitted B and V bands. Calibrated field stars transferred the <magnitude zero point>; late host images removed <galaxy> light, <point spread function> fits measured brightness, and inserted artificial stars tested recovery.<Observational Evidence from Supernovae for an Accelerating Universe and a Cosmological Constant by Adam Riess and others>[Sections 2.1–2.2, PDF pp.5–8]{ref}
The <Supernova Cosmology Project> likewise discovered rising events with CTIO 4-m <charge-coupled device> imaging, then followed their <light curves> with several telescopes. Keck LRIS and ESO <spectroscopy> supplied classification and <redshift>, preferably from narrow host lines. Cross-filter <K corrections> connected received R/I observations to emitted B/V light.<Measurements of Omega and Lambda from 42 High-Redshift Supernovae by Saul Perlmutter and others>[Section 2, PDF pp.2–3: apparatus and corrections]{ref}
Why does fainter mean accelerating? First it means a larger inferred <luminosity distance> only at fixed standardized <luminosity> and corrected attenuation. Next it favours acceleration only after a comparison family is specified: an <FLRW metric>, representative propagation of light, common nearby and distant standardization, and controlled <measurement calibration> and selection. For the usual <general relativity> models containing nonnegative <pressureless matter>, <spatial curvature of an FLRW universe> and a <cosmological constant>, the shape of the <Hubble diagram> constrains $\Omega_{m,0},\Omega_{\Lambda,0}$ and thus $q_0=\Omega_{m,0}/2-\Omega_{\Lambda,0}$. A positive residual relative to one decelerating model rejects that prediction; it does not alone prove $q_0<0$ across arbitrary expansion histories or arbitrary opacity laws.
The essential advance is that a nearby sample fixes an intercept while a range of distant <redshifts> measures the shape. The two teams supplied independent observations and analysis, although common nearby calibrators and <Type Ia supernova> physics can produce shared uncertainties. The <dark-energy acceleration criterion> explains the interpretation within <general relativity>; the <worked supernova distance comparison> shows the size and sign of the effect without fitting any survey data.
Back to article page
