Testing the assumptions of turbidite paleoseismology using the 2016 Kaikōura earthquake

Understanding the magnitude and frequency of earthquakes at subduction zones beyond historical and instrumental records is paramount to understanding the potential hazard these earthquakes pose in the future. However, terrestrial evidence of past earthquakes is generally patchy and poorly preserved. One method that circumvents some of these limiting factors is the study of seafloor turbidites. Turbidite paleoseismology uses the observed relationship between widespread triggering of turbidity currents and the co-occurrence of large earthquakes to build up a paleoseismic record. However, the application of turbidite paleoseismology is underpinned by largely untested hypotheses, due to a lack of modern examples of earthquake-triggered turbidites. The 2016 Mw 7.8 Kaikōura earthquake that struck New Zealand triggered turbidity currents in ten consecutive canyons along hundreds of kilometres of the southern Hikurangi margin. Their occurrence provided an unprecedented opportunity to examine earthquake-triggered turbidites within the framework afforded by paleoseismology. The Kaikōura event bed (KEB) is examined in detail herein to test some key assumptions made in turbidite paleoseismology. The first is the assumption that a single, well-placed core can contain the full history of earthquake-triggered turbidity currents in a given distributary system. Studies have shown that deposition of turbidites can be highly variable, both along the flow path of a turbidity current and with height above a canyon/channel thalweg. The KEB is deposited ubiquitously where the canyons broaden in their lower reaches, and in the channel, ~30 m above the canyon/channel thalweg. This result offers confidence that representative coring locations can be found for turbidite paleoseismology, and the methods currently used to identify optimal core sites are likely robust. Some turbidite paleoseismology studies also use the confluence test, a method used to identify synchronous deposition of turbidites. The confluence test counts the number of turbidites above and below a confluence, and if the number below the confluence is the same as in the tributaries above, the turbidites are inferred to be triggered simultaneously. In this study, turbidites linked to the 2016 Kaikōura (KEB), 1855 MW 8.2 Wairarapa and 1848 MW 7.4 Marlborough earthquakes are examined through the confluence of Cook Strait, Campbell and Opouawe canyons with the Hikurangi Channel to provide the first evaluation of the confluence test following observed earthquakes. The number of turbidites present below the confluence crucially does not exceed the number observed above, showing that these historic earthquakes produced turbidites that adhere to the assumptions of the confluence test, thus showing the viability of the method. Examination of turbidites through the last millennium in the Cook Strait region highlights the potential utility of turbidite records to unravel complexities in fault rupture styles. By linking turbidite records to terrestrial paleoseismic evidence, instances where upper plate faults rupture synchronously with each other and/or the subduction interface can be identified. From this, the most recent Hikurangi margin rupture (1473-1434 CE) does not show any evidence of synchronous rupture. However, the penultimate rupture (1153-1072 CE) potentially involved synchronous rupture with the Wairarapa and/or Wellington faults. Paleoseismic records derived from marine turbidites, while they have often been debated in the literature, are shown here to be a valuable tool that can shed light on the potential hazards posed around subduction margins.