Testing quartz and feldspar luminescence dating to determine earthquake and tsunami recurrence in the area of the giant 1960 Chile earthquake

Coasts on active subduction zones are recurrently affected by great earthquakes and associated tsunamis. Deposits resulting from both tidal inundation after coseismic subsidence and from tsunamis can be radiocarbon dated to infer patterns of earthquake and tsunami recurrence. Yet, finding suitable plant material for radiocarbon dating is not always easy. Luminescence dating provides a more ubiquitously applicable approach, but in such settings remains challenging due to three main reasons: (i) The quartz and feldspar minerals formed on active subduction zones tend to show inadequate luminescence properties; (ii) tsunami deposits may be affected by incomplete resetting of luminescence signals; and (iii) the dosimetry in sequences of tsunami and tidal sand sheets intercalated in marsh soils is often complicated. To evaluate the impact of these factors on luminescence dating accuracy, we re-date previously historically-constrained and radiocarbon dated tsunami and post-earthquake tidal deposits marking some predecessors of the giant 1960 Chile earthquake. While the impact of complex dosimetry on luminescence ages was negligible for both quartz and feldspar, the selection of an appropriate luminescence signal and aliquot size was crucial for generating robust chronologies. Due to unstable luminescence signals, quartz optically stimulated luminescence (OSL) ages and fading corrected infrared-stimulated (IR₅₀) feldspar ages significantly underestimate the control ages. Only post-infrared-infrared (pIRIR₁₅₀) signals of feldspar provide ages without any systematic age underestimation. Incomplete signal resetting was successfully addressed by using small aliquots in combination with the bootstrapped minimum age model. Remaining remnant ages of a few decades to a few centuries for the historical 1960 and 1575 tsunami sand sheets reflect a combination of laboratory residuals and incomplete signal resetting. Our results show that low-temperature pIRIR dating of feldspar is a valuable tool for reconstructing late Holocene earthquake chronologies in regions where quartz is geologically young, volcanic in origin, and holds few transportation cycles, as it occurs in active subduction zones. Although relatively large residual doses introduce dating uncertainties for sediments younger than a few centuries, feldspar pIRIR dating appears to be a good alternative to radiocarbon dating. On the contrary, quartz OSL and feldspar infrared ages were associated with systematic age underestimation, making them problematic for obtaining adequate chronologies in this type of geologic setting.