Properties of an inverse isochron:

• Argon data can be plotted on an isochron to help assess the isotopic composition of Ar trapped at the time of argon closure, thereby testing the assumption that trapped argon isotopes have the composition of modern atmosphere which is implicit in age spectra.
• The x-axis plots the ³⁹Ar/⁴⁰Ar ratio of the individual heating steps. Ratios are corrected for extraction line and mass spectrometer blank contributions and nucleogenic isotopes. Correction for atmospheric argon is not applied.
• The y-axis plots the ³⁶Ar/⁴⁰Ar ratio of the individual heating steps. Ratios are corrected for extraction line and mass spectrometer blank contributions and nucleogenic isotopes. Correction for atmospheric argon is not applied.
• The "best fit" line through the data array yields the trapped ⁴⁰Ar/³⁶Ar value from the y-intercept and the ⁴⁰Ar/³⁹Ar value (age) from the x-intercept.
• The MSWD or Mean Sum Weighted Deviates is a statistical "goodness-of-fit" indicator where the higher the MSWD value, the poorer the line fits the data. Commonly accepted MSWD values are less than 2.5. The error on the x- and y-intercepts is calculated from the regression results of York (1969)
• Isochrons are most useful for step-heated or total fusion data which have a significant spread in radiogenic yield: highly radiogenic data points plot near the x-axis while low radiogenic data points plot near the y-axis. The spread of data points along the 'chron line' help better constrain the intercepts, thus reducing uncertainties.
• For the above isochron analysis, a step-heated groundmass concentrate yielded a very poor age spectrum with a spread in ages from 2.5 to 5.0 Ma (no plateau). The isochron indicates that the sample contained an elevated trapped ⁴⁰Ar/³⁶Ar ratio (303.7 ± 1.2 instead of 295.5). The red triangle on the y-axis indicates the present day ³⁶Ar/⁴⁰Ar value (0.00338). The excess of ⁴⁰Ar is corrected for in the isochron yielding an age of 2.59 ± 0.08 Ma (compared to approximately 3.5 Ma for the age spectrum).

---

York, D., 1969. Least squares fitting of a straight line with correlated errors, Earth and Planet. Sci. Lett., 5, 320-324.

 

Back to Data Tutorial