Statistics for Fission-Track Thermochronology
This chapter introduces statistical tools to extract geologically meaningful information from fission-track (FT) data using both the external detector and LA-ICP-MS methods. The spontaneous fission of 238U is a Poisson process resulting in large single-grain age uncertainties. To overcome this imprecision, it is nearly always necessary to analyse multiple grains per sample. The degree to which the analytical uncertainties can explain the observed scatter of the single-grain data can be visually assessed on a radial plot and objectively quantified by a chi-square test. For sufficiently low values of the chi-square statistic (or sufficiently high p values), the pooled age of all the grains gives a suitable description of the underlying ‘true’ age population. Samples may fail the chi-square test for several reasons. A first possibility is that the true age population does not consist of a single discrete age component, but is characterised by a continuous range of ages. In this case, a ‘random effects’ model can constrain the true age distribution using two parameters: the ‘central age’ and the ‘(over)dispersion’. A second reason why FT data sets might fail the chi-square test is if they are underlain by multimodal age distributions. Such distributions may consist of discrete age components, continuous age distributions, or a combination of the two. Formalised statistical tests such as chi-square can be useful in preventing overfitting of relatively small data sets. However, they should be used with caution when applied to large data sets (including length measurements) which generate sufficient statistical ‘power’ to reject any simple yet geologically plausible hypothesis.
The author would like to thank reviewers Rex Galbraith, Mauricio Bermúdez and editors Marco Malusà and Paul Fitzgerald for detailed feedback. The mica counts of Fig. 6.1 were performed by Yuntao Tian.
- Bernet M (2018) Exhumation studies based on detrital fission track analysis on sand and sandstones. In: Malusá M, Fitzgerald P (eds) Fission track thermochronology and its application to geology, chapter 14. Springer, BerlinGoogle Scholar
- Carter A (2018) Thermochronology on sand and sandstones for stratigraphic and provenance studies. In: Malusá M, Fitzgerald P (eds) Fission track thermochronology and its application to geology, chapter 13. Springer, BerlinGoogle Scholar
- Chew DM, Donelick RA (2012) Combined apatite fission track and U-Pb dating by LA-ICP-MS and its application in apatite provenance analysis. Quant Min Microanal Sed Sediment Rocks: Mineral Assoc Canada, Short Course 42:219–247Google Scholar
- Galbraith R (2010a) Statistics for LA-ICPMS fission track dating. In: Thermo2010—12th international conference on thermochronology, Glasgow, p 175Google Scholar
- Galbraith RF (2010b) On plotting OSL equivalent doses. Ancient TL 28:1–10Google Scholar
- Gallagher K (2012) Transdimensional inverse thermal history modeling for quantitative thermochronology. J Geophys Res: Solid Earth (1978–2012) 117(B2)Google Scholar
- Gleadow A (2018) Future developments in fission track thermochronology. In: Malusà M, Fitzgerald P (eds) Fission track thermochronology and its application to geology, chapter 4. Springer, BerlinGoogle Scholar
- Hurford AJ (2018) A historical perspective on fission track thermochronology. In: Malusà M, Fitzgerald P (eds) Fission track thermochronology and its application to geology, chapter 1. Springer, BerlinGoogle Scholar
- Ketcham RA (2018) Fission track annealing: from geological observations to thermal modeling. In: Malusà M, Fitzgerald P (eds) Fission track thermochronology and its application to geology, chapter 3. Springer, BerlinGoogle Scholar
- Ketcham RA, Donelick RA, Donelick MB et al (2000) AFTSolve: a program for multi-kinetic modeling of apatite fission-track data. Geol Mater Res 2(1):1–32Google Scholar
- Tufte ER (1983) The visual display of quantitative data. Graphics, CheshireGoogle Scholar