| Résumé | (disponible en anglais seulement)
It is very common for sedimentary rock samples to have compositionally-variable detrital apatite with overdispersed apatite fission track (AFT) ages. Proper
characterization of AFT annealing kinetics is essential for determining whether differential thermal annealing and/or variable provenance are the cause for the age dispersion. Although the widely used single kinetic parameters, Cl content and D(par),
may work for simple apatite compositions, they are generally inadequate for resolving multi-kinetic annealing behavior in more typical detrital samples with heterogeneous apatite compositions. The r(mr0) parameter, determined using elemental data, is
able to resolve multiple statistical AFT kinetic populations for a large suite (> 150 samples) of Phanerozoic age (Cambrian through Eocene) samples from northern Canada that contain apatite with highly variable cation (Fe, Mn, Mg, Na, Sr, La, Y, Ce)
and anion (F, Cl, OH) concentrations. In general, these same populations cannot be distinguished using Cl or D(par) due to substantial or complete overlap of population age and length data in kinetic parameter space. Analysis of 340 duplicate sets of
apatite elemental data from 50 of these Phanerozoic samples indicate that r(mr0) values, when expressed as "effective Cl" values, are reproducible to within ±0.03 apfu for the majority of the data (>80%). This uncertainty represents ~5-10% or less
of the typical range of r(mr0) values for the apatite grains comprising these multi-kinetic samples, meaning much less overlap of age and length data across population boundaries. In contrast, >400 duplicate D(par) values for the same samples are
reproducible within ±0.25 microns and ±0.50 microns, for 56% and 90% of the data, respectively, representing ±25% and ±50% of the typical range in D(par) values (~1 micron) for these samples. The large uncertainty in D(par), and the use of Cl only,
means that application of these parameters can result in mixed, poorly-defined AFT kinetic populations that can bias model thermal history results.
Multi-kinetic modelling, constrained using elemental data, extends the time-temperature range and
resolution of thermal histories in comparison to more conventional methods that use the single kinetic parameters, D(par) or Cl content. Although r(mr0) values are more precise than D(par) or Cl for defining kinetic populations, calculated values
are less accurate beyond the dominant range of the annealing experimental calibration data (0.75 < r(mr0) < 0.84; Ketcham et al. 1999) and must be estimated through thermal modelling. AFTINV, an inverse multi-kinetic AFT thermal model, is used to
obtain statistically-acceptable, geologically-constrained thermal solutions that fit observed AFT and vitrinite reflectance data. Multi-kinetic AFT samples from northern Canada can have two or three kinetic populations with geological annealing
temperatures that can range from ~75°C to >200°C, consistent with temperatures inferred from annealing experiments. Such a broad temperature range can constrain thermal histories further back in time and allow for the resolution of multiple heating
events.
Ketcham R.A., Donelick, R.A., and Carlson, W.D. 1999. Variability of apatite fission-track annealing kinetics: III. Extrapolation to geological time scales. American Mineralogist, 84: 1235-1255. |