Rates and Controls of Erosion on the Western Southern Alps of New Zealand over 102 yr timescales

Abstract:

The Southern Alps of New Zealand are one of the fastest-eroding mountain ranges in the world, but many open questions remain about the magnitude and controls of erosion rates on their western side: Erosion rates have been estimated based on decadal landslide inventories, but these represent a relatively short observational time-span, and they are limited to erosion from one process with observations below the tree line. Frost-cracking has been shown to be the dominant erosional mechanism on the eastern, drier side of the Southern Alps, but its role in the erosion of the western side has not been examined. Whether the locus of maximum erosion is located closer to the Alpine Fault, at the main drainage divide, or in between, and if this position is controlled by tectonics or glaciation, is also debated.

To examine the rates and controls of erosion along the western Southern Alps, we present new ¹⁰Be catchment-averaged erosion rates from 17 catchments between the Paringa and Maruia rivers. These estimates integrate over 10² yr timescales and are mostly representative of inter-seismic erosion. We find that catchment-averaged erosion rates span an order of magnitude, from 0.8 to 9 mm/yr. Erosion rates experience a sharp drop north of the intersection with the southernmost fault of the Marlborough Fault System, the Hope Fault, in a similar manner as the vertical uplift rates at the Alpine Fault. Previously published ¹⁰Be erosion-rate estimates from three of these catchments show good reproducibility with our new rates, despite the widespread landslide activity. To a first order, catchment-averaged erosion rates are consistent with exhumation rate estimates from thermochronology, which integrate over 10⁶ yr timescales, and with erosion-rate estimates from river sediment yields and landslide inventories, which integrate over 10¹ yr timescales. Catchment-averaged erosion rates correlate well with several geomorphic metrics, but the factor that best explains their variability is the proportion of the catchment area at elevations of 1500 to 2000 m. We examine if this elevation control arises from uplift, slope or precipitation gradients across the range, frost-cracking, or a transient hillslope response following deglaciation.