Dating intramontane alluvial deposits from NW Argentina using luminescence techniques: Problems and potential

Intramontane basin sediments are an archive of the interaction between basin bounding faults, and alluvial fan and fluvial systems. The chronologies of intramontane basin sedimentation enable an understanding of the cycling of sediments within a basin through time, can be interrogated to identify periods of alluvial storage and erosion, provide rates of sediment accumulation and storage and date fault movement. If suitable dating methods (in terms of resolution and timescale) are applied to develop the chronologies of alluvial archives, it is then possible to discriminate between climate and tectonic forcing mechanisms on long-term basin behaviour. Optically stimulated luminescence (OSL) dating of quartz grains from alluvial sediments is an ideal technique for establishing a chronological framework of basin sedimentation as the method directly dates sedimentation events. However, our experience of OSL dating of quartz minerals extracted from Late Quaternary alluvial sequences in the quebradas of the Eastern Cordillera of NW Argentina has presented a number of challenges concerning selection of appropriate facies to analyse, mineral contamination, failure of fundamental protocol tests, proximity to saturation, and broad and multi-modal age distributions. Through careful analysis of the alluvial sedimentology and choice of sampling environments we have been able to locate suitable samples in most vertical sequences studied. A post-infrared-OSL approach demonstrated that contaminant signals were resulting in protocol test failure and, conversely, circumvention of this problem has increased confidence and reliability in the dating results. Assessment of dose–response characteristics suggests that the luminescence for the oldest samples is not likely to be saturated and in turn ages are not considered to be underestimated. Finally, different statistical tests have enabled objective identification of single low-dose populations in complex distributions and confirmed that broad distributions in some samples may be treated as single populations. These statistical approaches have considerably lowered uncertainty on OSL ages and confirmed concordance with other dating evidence.