Lost landslides: Rock-avalanche occurrence and fluvial censoring processes on South Island, New Zealand

Bainbridge, Rupert (2017) Lost landslides: Rock-avalanche occurrence and fluvial censoring processes on South Island, New Zealand. Doctoral thesis, Northumbria University.

[img] Archive (Zip archive - GMZ files, geolocation data, etc.)
Rupert Bainbridge - thesis - Digital Appendices.zip - Supplemental Material

Download (1MB)
[img]
Preview
Text (Doctoral thesis)
Rupert Bainbridge - thesis - Lost landslides, Rock-avalanche occurrence and fluvial censoring processes on South Island, New Zealand.pdf

Download (28MB) | Preview

Abstract

Rock-avalanches (RAs) are a large (typically >10⁶ m3) and extremely rapid (30 - >100 m/s) type of landslide. RAs pose a significant hazard as they can runout over long distances and generate secondary hazards such as tsunami and unstable, cross-valley dams. Previous research on the distribution of rock-avalanche deposits (RADs) on the South Island, New Zealand has suggested that there are fewer deposits than would be expected for a seismically active, high-mountain region. This is due to their removal from the sedimentary record (censoring) by fluvial erosion, glacial entrainment, vegetation cover, sub-aqueous occlusion and deposit misidentification. Censoring of deposits skews magnitude-frequency relationships of RA occurrence and hinders hazard planning. This research examines processes acting to fluvially censor RADs on the South Island. 268 known, and 47 possible RADs were identified to provide the first RAD inventory for the entire South Island. The temporal distribution of RADs indicates censoring of the record over the Holocene. >500 year intervals exist between RA events from 12,000 to 2,000 years ago; a more complete record is shown for the last 1,000 to 100 years with intervals of >50 - <150 years. The last 100 years shows phases of co-seismic RAD generation, a period of RAD quiescence and a recent increase in aseismic RAD occurrence. The spatial distribution of RADs suggests that the West Coast, Fiordland and Nelson could have experienced fluvial censoring of deposits. The sediment routing characteristics of catchments in these regions, where the majority of rivers have direct pathways from RADs to the ocean, suggest that fluvially reworked RAD material could be stored within alluvial flats and braidplains.

Agglomerate grains (microscopic grains which are diagnostic of RAs) were used to identify fluvially reworked RAD material. Grains were detected in dam-breach flood terraces up to 1km downstream of known RADs. Contemporary river sediment samples showed no agglomerate presence, this suggests that 1) agglomerates break down under extended fluvial transport, 2) they are not supplied to river systems outside of flood events, 3) agglomerates become diluted by other river sediment or 4) they become buried in discrete sedimentary layers. In order to investigate the redistribution of coarse RAD material within South Island rivers, a micro-scale flume model was developed. Using ultra-violet sand as a novel analogue for a RAD, the redistribution of material through an idealised South Island catchment could be examined. The model showed that RAD material is deposited in discrete aggradational layers in dam proximal locations. Downstream, the sedimentary signal is rapidly diluted by ordinary river sediment flux.
The research shows that the RAD record for the South Island is incomplete and that fluvial censoring is prevalent within the West Coast, Nelson and Fiordland. The agglomerate tracing method can be used to identify the presence of RADs in fluvial systems proximal to RADs but the signal is undetectable after ~1km from the deposit. Both field sampling and flume modelling show that localised flood derived aggradational layers, close to deposit locations, will archive reworked RAD material. These results have important implications for understanding the magnitude and frequency of RADs within New Zealand and other similar high-mountain, tectonically active regions of the globe.

Item Type: Thesis (Doctoral)
Uncontrolled Keywords: geomorphology, flume modelling, agglomerates, database, hazard and risk
Subjects: F600 Geology
F800 Physical and Terrestrial Geographical and Environmental Sciences
Department: Faculties > Engineering and Environment > Geography and Environmental Sciences
Depositing User: Paul Burns
Date Deposited: 23 Nov 2017 11:06
Last Modified: 16 Sep 2022 17:45
URI: https://nrl.northumbria.ac.uk/id/eprint/32621

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics