DIFFERENTIATION OF ROCKFALL EVENTS: SPATIAL ANALYSIS OF MODERN AND PALEO ROCKFALL, BANKS PENINSULA, NEW ZEALAND

Rockfall events create long-lasting imprints on topography and in society, and their deposits provide a record of those impacts. In the Canterbury Earthquake Sequence (CES) thousands of rockfalls occurred, highlighting the susceptibility of the Miocene volcanics of Banks Peninsula to rockfall. This study analyzes a region of Banks Peninsula, Purau, in which rockfall occurred during the CES and previously. We present a field-based methodology for the recognition of paleo rockfall beneath reactivated source regions, and attempt spatial analysis of boulder populations. Field measurements of boulders (dimensions, lichen coverage, surface roughness, and sediment wedge) formed a collated database for analysis in ArcGIS10.2. Within ArcMap, shadow angles were calculated as a method of analyzing runout distances. Using the shadow angle, spatial analysis and correlation of lichen, surface roughness, and the size of the sediment wedge of a boulder was used to separate the data set into populations (increasing values for all three indicates boulder residence time), which in this study identified one modern and three paleo rockfall populations.

Paleo rockfalls typically have a distribution with the largest boulders closest to the source, while modern rockfall follows an opposite trend, an aspect we link with historic (late 1800s) deforestation increasing runout distances. Topography also controls spatial distribution, with gullies acting as a mechanism to arrest and direct falling boulders. The three paleo rockfall populations can be linked to source regions. Rockfall distribution extents analyzed for each paleo event indicate different source areas were more active in certain populations than in others, and that some sources were active in all three populations. This study highlights the use of relatively simplistic and low-cost methods to identify populations within paleo rockfall events, which can be used as a preliminary methodology prior to precise dating of rockfall events.