Rb–Sr · York 1969

Isochron

An isochron is a regression. Break the assumptions. It will still print an age.

Fig. 1 · Suite BC

2σ bars sit inside the symbol

0.700.750.800.850.900510⁸⁷Rb/⁸⁶Sr⁸⁷Sr/⁸⁶Sr
  • York fit
  • Closed-system isochron
  • Far from the line
  • Omitted
Age comes only from the slope: t = ln(1 + slope) / λ, with λ(⁸⁷Rb) = 1.3972×10⁻¹¹ yr⁻¹. The fit is not told whether the rock stayed closed.

Apparent age

1.05 Ga

1050.7 ± 4.2 Ma

Closed system, common initial ratio, nothing omitted. This slope is an age.

Closed-system age of this suite: 1.05 Ga.

Initial ⁸⁷Sr/⁸⁶Sr
0.70423 ± 0.00007
MSWD
0.65
Probability of fit
0.69
Points in the fit
8 / 8
  • One initial ⁸⁷Sr/⁸⁶Sr
  • Closed after crystallization
  • Cogenetic, not a mixture
  • Every analysis kept
  • Scatter within quoted error

What you do to the rocks

Two-component mixing0%

Blends a young low-Rb melt with a young high-Rb rock that already has high ⁸⁷Sr. The line stays straight. The slope is not an age.

Open-system leak0%

Fraction of ⁸⁷Sr above 0.7042 that leaves. Uniform loss rotates a straight line. High-Rb loss bends it.

Inherited daughter0%

Initial ⁸⁷Sr/⁸⁶Sr rises, and rises more in high-Rb samples. Intercept and slope both shift. The line can stay perfect.

Partial reset0% lost

Later heat strips radiogenic ⁸⁷Sr from the aureole samples only. Reset every sample and it looks like a younger isochron.

Samples in the aureole0
Analytical noise0%

Extra scatter beyond the quoted errors. The error bars stay the same size, so MSWD climbs. Then drop the points you dislike.

The equation, live

(⁸⁷Sr/⁸⁶Sr) = (⁸⁷Sr/⁸⁶Sr)₀ + (⁸⁷Rb/⁸⁶Sr)(eλt − 1), only if the system stayed closed and every sample began at the same initial ratio.

slope = 0.01479 ± 0.00006
λ = 1.3972e-11 yr⁻¹

The game edits the left-hand side, then fits a line anyway. York (1969), uncorrelated errors. Decay constant from Nebel, Scherer and Mezger (2011).

Analyses

Uncheck a point to drop it.