Assessing the statistical attributes of sensitivity and equivalent dose estimation of multiple grains of K-feldspar single aliquots, central Garlock fault, California
- Author(s): Okubo, Steven Garrett;
- Advisor(s): Rhodes, Edward J;
- et al.
The central Garlock fault in California offers a unique opportunity to collect a large volume of useful data to determine its incremental slip rate during the Holocene. The slip rate of the central Garlock fault through the Holocene is based on slip-rate studies and the timing of events interpreted from paleoseismic studies. Understanding the ages of displaced sediments in offset geomorphic features will develop a better understanding of the slip rate. Optically stimulated luminescence (OSL) can be used to date the burial age of these displaced sediments. However, OSL dating has its challenges. Many models may either consider the possibility of incomplete bleaching or the mixing of different aged grains. At the luminescence laboratory at UCLA, the informal "discrete minimum method" (DMM) attempts to consider both of these challenges. In order to examine the reliability of the DMM, the luminescence data of samples collected along the central Garlock fault at the Christmas Canyon west site were analyzed. This analysis included the statistical assessment of each sample's sensitivity and probability of estimated equivalent dose values. Also, an evaluation of the same data was conducted using Galbraith's (2005) finite-mixture model as well as the minimum-age models by Galbraith et al. (1999). Probability distributions indicated at least two sensitivity groups within each sample. Because differences in sensitivity are probably derived from different types of source rock, multiple types of feldspar are likely contributing to each sample's luminescence signals. Polymodal, wide and skewed unimodal PDF's and KDE's of the equivalent doses indicate mixed dose populations within the majority of CCW samples. Finite-mixture modelling is an effective indicator of a sample having internal De variability. However, it does not indicate which component corresponds to the burial age of a sediment layer. Minimum-age modeling may be effective at determining the minimum age of a sample, but does not consider the possibility of a younger component being introduced into a sediment layer. This may result in an underestimation of De for some samples. The "Discrete Minimum Method" appears to address the existence of mixed dose populations at the CCW site while separating the groups of unbleached and any other mixed grains unreasonable to use for De estimation.