In-Situ Sampling and Spatially Resolved Measuring Approaches for Optical Surface Exposure Dating for Late Quaternary Applications

Abstract

Optically stimulated luminescence (OSL) exposure dating utilizes OSL-at-depth signals to extrapolate an exposure age from rock surfaces. Exposure ages are commonly obtained by fitting the forms of luminescence depth profiles, which depend on parameter estimates of light attenuation and defined rates of luminescence bleaching. Current procedures for obtaining these parameters for a rock surface require matching luminescence depth profiles from compositionally and morphologically matched rock surfaces with known exposure ages, which limits the accuracy and applicability of the technique. Further, traditional measuring procedures for depth profiles involve measuring the OSL of millimeter slices from surface core samples, offering poor resolution datasets and limiting parameter and age fitting accuracies. With the aim of improving the accuracy and applicability of OSL surface exposure dating, modified sampling and measuring procedures incorporating controlled exposure experiments and spatially resolved OSL laser scanning measures are performed in a trial study on 11-year exposed quartzite rocks. The first experiment involves trials of controlled exposure sampling approaches, to attempt to reliably determine exposure dating model parameters directly from the rock surface of interest using OSL saturated core samples subjected to controlled light exposures. Measured from wafer derived datasets, natural sunlight controlled exposed parameters were able to produce decadal ages, equivalent in magnitude to that of ages produced from proximal rock sourced parameters. Parameters acquired from simulated light, however, produced centennial to decamillennial ages. Data scatter in the luminescence depth profiles substantially limit parameter and age precision of all techniques, however, warranting more resolute OSL measuring protocols to make more valid conclusions about the use of controlled exposures for rock surface parameterization. Resultingly, OSL scanning measures were trialed on the Lane Mountain samples to record sub-millimeter resolution OSL depth profiles of the core surface samples, to compare if the parameterizations from these higher resolution datasets would be more precise than parameters acquired from lower resolution wafer derived datasets. Scanning electron microscopy with energy dispersive spectroscopy analyses (SEM-EDS) were used additionally on OSL scanned samples to identify and filter out non-quartz OSL anomalies in scan data, with the aim of generating depth profile data which more closely follows model expectations for quartz-OSL exposure dating, and to see how parameterizations are influenced with this modification to data. The use of spatially resolved OSL depth profiles from scan datasets improved parameter extrapolative precision over wafer derived datasets for individual and combined core datasets, although individual datum error was generally higher than in wafer datasets. The use of non-quartz anomaly filtering on spatially resolved scan data improved data error bounds of anomaly affected regions, and improved the data scatter of spatially resolved depth profiles for both individual and combined core datasets, but minorly improved or matched parameter precision to nonfiltered OSL scan data. Once core parameterizations were acquired, age calculations using spatially resolved data were performed using proximal rock sampling techniques. For individual age fits from each core sample, filtered and nonfiltered spatially resolved OSL depth profiles produced inconsistent ages between known age surfaces, but the calculated precision was improved when using scan and scan filtered data. Combined core age fits of each rock surface using scan and scan filtered data produced more accurate age results to 11 years than age fits from individual cores, but each fit type produced comparable fit precision. Ages using controlled exposure experiment techniques were unable to be extrapolated from spatially resolved data, given the scan data from the controlled exposed samples were too low in OSL intensity to acquire viable depth profiles for adequate parameter extrapolation. The promising results in using controlled exposure experiments for parameterization, and the observed precision improvements when using spatially resolved OSL to parameterize rock surfaces, invite the opportunity to apply both sampling and measuring approaches to exposure date quartzite erratic members of the Foothills Erratics Train, which are interpreted to have been deposited during the final retreat stage of the Laurentide and Cordilleran ice sheets and the opening of the ice-free corridor. Inconsistent deposition timelines of the erratics and the opening of the ice-free corridor warrant the application of an alternative surface exposure chronometer more sensitive to millennial timescales, such as luminescence exposure dating, to attempt refining the timeline of Erratics Train deposition. The use of controlled exposure experiments allows for exposure dating trials at the sites, where no known age proximal rocks are available for the erratic surfaces. Further, the use of higher resolution spatially resolved OSL may offer precise age characterizations for decamillennial time. Lab simulated, unidirectional sunlight exposures were used to extrapolate parameters from the erratic surfaces, given difficulties in facilitating natural sunlight exposures, and depth profiles were measured using spatially resolved OSL laser scanning techniques. Age results for the erratics offered sub-annual to centennial exposure ages, providing no realistic insight on the deposition timelines of the Foothills Erratics Train members. However, the results indicate that the parameterizations of OSL depth profile development in the exposure dating model may be too simple, and that other influences in the experimental setup may affect parameter accuracy and precision. For instance, mineralogical variations in OSL may impact fit precision, which are unaccounted for in the model form, and were not able to be considered for this assessment, as X-ray element maps produced for the cores were not usable for filtering assessments. Second, the higher intensity OSL acquired from the controlled exposure samples indicates that the use of unidirectional, single flux rate solar simulated light, which follows model conditions for light exposure, may oversimplify exposure conditions over decamillennial timescales, producing inaccurate exposure ages. Third, the fit approach for parameterization may also not be effective in obtaining accurate parameters for the rock surface, given inconsistencies observed between physically derived parameters and curve fitted parameters both from Foothills Erratics Train and Lane Mountain samples. Fourth, the presence of weathering rinds may impact the feasibility of controlled exposure rock surface parameterization, and surface altered cores should be incorporated in controlled exposed samples to attempt best emulating rock surface exposure conditions. Finally, erosion rates are likely to impact depth profile evolution in the Foothills Erratics Train study, but are not quantified in the model form, potentially causing the younger than expected exposure ages. With continued physical experiments on depth profile evolution, and application trials to test modified parameterizations, the ability to date millennial-decamillennial exposed surfaces can be more effectively evaluated. Specific physical experimental studies on depth profile effects from the light source, such as photon flux variations and angle of illumination variations, as well as sample characteristics light surface coverage, weathering rind impacts, and erosion, can offer improved insight on the effective parameterization of depth profile evolution for exposure dating applications. Still, the use of higher spatially resolved OSL measurements, and the use of controlled exposure parameter sampling, have shown potential in expanding the applicability and precision of the exposure dating technique, and provide more detailed measures of OSL, which can benefit trial experiments aiming to better parameterize depth profiles for exposure dating methods. With these new procedures, continued research on the parameterization of depth profiles for exposure dating can be more effective in execution, and can improve OSL exposure dating to become an established geochronometer.