· · 2018-05-17
| fit | Teff [K] | log g | log n(He)/n(H) | Z | reduced χ² |
|---|---|---|---|---|---|
| three parameters | 23154 ± 327 | 4.955 ± 0.044 | -0.55 ± 0.03 | held 0 | 1.41 |
| four parameters | 23605 ± 352 | 5.014 ± 0.053 | -0.59 ± 0.04 | 0.75 ± 0.22 | 1.38 |
Observed minus model scatters by 2.64% of the continuum over the fitted lines, worst on He I 5876 (5.2%), Hζ (3.3%). That number, not the signal-to-noise, is what says whether the spectrum can carry a fourth parameter.
| pair | correlation |
|---|---|
| Teff ↔ Z | +0.407 |
| log g ↔ Z | +0.381 |
| helium ↔ Z | -0.341 |
| Teff ↔ log g | +0.743 |
Hβ, Hγ and Hδ give -78.8 km/s between them, so the star's radial velocity is +78.8 km/s; they scatter by 13.4 km/s about it.
| line | λ0 | shift [Å] | [km/s] | |
|---|---|---|---|---|
| Hβ | 4861.32 | -1.39 | -85.7 | |
| Hγ | 4340.46 | -0.87 | -60.1 | |
| Hδ | 4101.73 | -1.24 | -90.6 | |
| Hε | 3970.07 | -0.50 | -37.8 | |
| Hζ | 3889.05 | -0.94 | -72.5 | |
| Hη | 3835.38 | -0.92 | -71.9 | |
| Hθ | 3797.90 | -1.06 | -83.7 | |
| Hι | 3770.63 | -0.52 | -41.3 | |
| Hκ | 3750.15 | -1.66 | -132.7 | beyond any plausible radial velocity |
| He II 4686 | 4685.70 | +8.12 | +519.5 | model slid +8.1 Å onto a noise minimum |
| He II 5412 | 5411.52 | -4.51 | -249.8 | beyond any plausible radial velocity |
| He II 4542 | 4541.59 | +11.36 | +749.9 | model slid +11.4 Å onto a noise minimum |
| He II 4200 | 4199.84 | -7.16 | -511.1 | model slid -7.2 Å onto a noise minimum |
| He I 5876 | 5875.83 | -1.69 | -86.2 | |
| He I 4472 | 4471.60 | -0.48 | -32.2 | |
| He I 4026 | 4026.29 | -0.69 | -51.4 | |
| He I 4713 | 4713.28 | -1.22 | -77.6 | |
| He I 4922 | 4921.93 | -1.22 | -74.3 | |
| He I 5016 | 5015.68 | -1.06 | -63.4 |
Each line is centred independently: the fitter finds its core with a parabola and slides the model there. The sign is the fitter's, opposite to the star's motion. Lines far from the Balmer mean above are not measuring a velocity — near the Balmer jump the series crowds and pulls the apparent core blueward, and a weak line has no core to find at all.
The whole reduced spectrum. Shaded bands are the intervals the fit used; drag on either panel to zoom.
Offset from the line centre in Ångström across, normalised flux up. The strip under each profile is observed minus model for whichever fit the buttons select; they all switch together and share a vertical scale.
The resolution is measured from the interstellar Ca II K line, which is not in the star, so its width is the instrumental profile. It is an upper limit: an unresolved interstellar line can still be several velocity components. The same fit gives the wavelength shift, which is the zero point only in so far as the cloud sits near rest.