Controls on 10Be dilution in catchments affected by coseismic landsliding: a 2016 Kaikōura earthquake case-study
Dr Clare Wilkinson1, Dr Tim Stahl1, Katie Jones2,3, Dr Toshiyuki Fujioka4,5, Dr David Fink5, Dr Kevin Norton2
1University Of Canterbury, Christchurch, New Zealand, 2Victoria University of Wellington, Wellington, New Zealand, 3GNS Science, Lower Hutt, New Zealand, 4Centro Nacional de Investigación sobre la Evolución Humana (CENIEH), Burgos, Spain, 5Australian Nuclear Science and Technology Organization (ANSTO), Menai, Australia
The 2016 Mw 7.8 Kaikōura earthquake triggered tens of thousands of landslides across the northern Canterbury and southern Marlborough regions in the South Island of Aotearoa New Zealand. The influence of landslides generated by this earthquake on sediment generation, transport and deposition in stream networks has varied across the region and through time — some catchments show significant and near-immediate responses while others show little to no change despite the extensive landsliding. We measured 10Be concentrations in detrital quartz sands over a two-year period in the Conway River catchment, which has a total area of ~475 km2 and had ~13 M m3 of new landslide material liberated from hillslopes during the 2016 Kaikōura earthquake. Samples for 10Be analysis were collected at the rangefront of the Seaward Kaikōura Mountains and near the catchment outlet on three sampling campaigns between 2017-2018. We also carried out a similar sampling regime in the nearby Hurunui catchment, which was unaffected by the 2016 Kaikōura earthquake. Measured 10Be concentrations were converted to basin wide mean denudation rates using accepted GIS-based elevation and shielding programs. Our results indicate that apparent catchment-wide erosion rates in the Conway River (i) did not change through time, (ii) have remained similar to basin mean erosion rates for the Hurunui, (iii) overlap with the range of values for exhumation rates of the region (from previously published low-temperature thermochronology data), and (iv) are influenced in part by selection of grain size. We also compared our 10Be concentrations for the Conway catchment to values derived by modelling mean landslide 10Be concentrations constrained by local production rates, detailed mapping of all landslides across the catchment, area-volume scaling, and landslide-channel connectivity estimates. Our modelling estimates show that the mean 10Be concentrations derived from landslide sediment would have been sufficient to dilute pre-earthquake catchment-wide 10Be values by up to a factor of 3. We explore various landscape and landslide parameters that may explain the mismatch between measured and modelled 10Be and ascertain that the combination of storage, site specific channel connectivity, and landslide geometries/failure mechanisms likely exert first-order controls on in-situ 10Be concentrations following such a large catchment-wide disturbance event. We conclude that using fluvial quartz grains to characterise catchment response to landsliding and the mass balance of earthquakes, is subject to a number of factors that are highly site-specific.
Clare Wilkinson recently completed her PhD at the University of Canterbury in Christchurch, NZ. She studied landscape and fluvial change following major disturbances, with a focus on the 2016 Kaikoura earthquake. She applied mixed-methods approaches and bicultural research to her doctoral work and is now a fluvial geomorphologist at Tonkin and Taylor.