Mauri J. McSaveney

Effects of debris on ice-surface melting rates: an experimental study

Here we report a laboratory study of the effects of debris thickness, diurnally cyclic radiation and rainfall on melt rates beneath rock-avalanche debris and sand (representing typical highly permeable supraglacial debris). Under continuous, steady-state radiation, sand cover >50mm thick delays the onset of ice-surface melting by >12 hours, but subsequent melting matches melt rates of a bare ice surface. Only when diurnal cycles of radiation are imposed does the debris reduce the longterm rate of ice melt beneath it. This is because debris >50 mm thick never reaches a steady-state heat flux, and heat acquired during the light part of the cycle is partially dissipated to the atmosphere during the nocturnal part of the cycle, thereby continuously reducing total heat flux to the ice surface underneath. The thicker the debris, the greater this effect. Rain advects heat from high-permeability supraglacial debris to the ice surface, thereby increasing ablation where thin, highly porous material covers the ice. In contrast, low-permeability rock-avalanche material slows water percolation, and heat transfer through the debris can cease when interstitial water freezes during the cold/night part of the cycle. This frozen interstitial water blocks heat advection to the ice-debris contact during the warm/day part of the cycle, thereby reducing overall ablation. The presence of metre-deep rock-avalanche debris over much of the ablation zone of a glacier can significantly affect the mass balance, and thus the motion, of a glacier. The length and thermal intensity of the diurnal cycle are important controls on ablation, and thus both geographical location and altitude significantly affect the impact of debris on glacial melting rates; the effect of debris cover is magnified at high altitude and in lower latitudes.

The October 1999 Mt Adams rock avalanche and subsequent landslide dam‐break flood and effects in Poerua river, Westland, New Zealand

Abstract On 6 October 1999 a very large (c. 10–15 million m3) rock avalanche from Mt Adams blocked the Poerua River 11 km upstream from the SH6 road bridge on the West Coast of the South Island. The 120 m high rock debris dam impounded a lake with a volume of 5–7 million m3before it overtopped on 7 October. The short survival time of landslide dams in rivers in Westland and around the world suggested a high probability of rapid dam failure and flooding downstream. This was confirmed when the dam breached 6 days later on 12 October 1999, during the first significant rain after the landslide occurred. The resulting dam‐break flood deposited considerable coarse gravel from the landslide in the valley downstream of the dam, and (mostly fines) on the alluvial fan below the Poerua gorge exit. The flow inundated farmland in the upper Poerua valley, but otherwise was largely confined to the river channel and did little damage at the time, mainly because of significant flow attenuation (c. 50% or greater) and sediment deposition on the alluvial fan below the gorge exit. Subsequently the remnant lake has been infilled, and c. 75% of the dam material has been transported downstream during floods. About 1.7 million m3 of alluvium has been deposited in the river channel and across farmland between the gorge exit and the SH6 bridge, changing the rivers course and causing lateral erosion of older terraces. These effects are continuing and will cause ongoing problems in the future. The rock avalanche and landslide dam failure in the Poerua valley were significant events which have had a profound local geomorphic impact. The landslide and subsequent downstream effects are typical landscape‐forming events. These events and the resulting community response to them have provided valuable information on the hazards, effects, and management of future landslide‐dam failures in Westland. Increased resources within local authorities for hazard assessments and response planning during such events would reduce the risk from dam‐break floods in the future, especially following the next Alpine Fault earthquake, when further landslide dam failures are expected.

Rockslides and Their Motion

The motion of landslides sourced from mostly bedrock (called rockslides) is controlled by the phenomenon of grain flow, and the frictional resistance of the constituent rock grains and their interstitial fluids. Modern understanding of grain-flow dynamics recognises that the important interactions between grains are irregularly distributed within the grain mass, with fortuitous alignments of grains carrying most of the stress in “force chains”, while other grains are only weakly stressed. In rapidly shearing grain flows, under substantial confining stress, force-chain stresses rise high enough to crush grains. Such comminuting grain flows develop a distinctive grain-size distribution that is fractal over many orders of magnitude of grain size down to sub-micron sizes. In the moment of crushing, grains are not solids, and behave as high-pressure fluids. As the grain fragments are injected into lower pressure surroundings, they behave as would any other fluid, lowering the effective stress on other grains, and thereby lowering frictional resistance to flow. We show how this affected the blockslide component of New Zealand’s prehistoric giant Waikaremoana rockslide; New Zealand’s Falling Mountain rock avalanche triggered by an earthquake in March 1929; and a small prehistoric New Zealand rockslide that was too small to be a comminuting grain flow, but which fell on and mobilized a fine, saturated substrate. We use grain-flow dynamics to explain the motion of these rockslides determined through field studies and physical and numerical modeling.

Orbital forcing of mid-latitude Southern Hemisphere glaciation since 100 ka inferred from cosmogenic nuclide ages of moraine boulders from the Cascade Plateau, southwest New Zealand

Cosmogenic nuclide (Be-10) exposure dating of moraine boulders in the Cascade Valley, southwest New Zealand, reveals three phases of glaciation with similar maximum magnitude since 100 ka. In this area, 8–10 lateral moraines were deposited during the Last Glacial Maximum (LGM) at 22–19 ka, and >15 lateral moraines and three end moraines were deposited during recession after the LGM. Also, three exposure ages of 29–33 ka from pre-LGM deposits may indicate increased weathering and erosion at the onset of the LGM in New Zealand, as has been suggested by other studies. An exposure age of 57.8 ± 2.7 ka from one of the highest moraines, combined with previous studies of cave speleothems, glacial features offset by the Alpine fault, the Vostok dust record, and sediment cores, supports the inference that a significant glacial phase culminated at 66–58 ka. A cluster of five exposure ages from older moraines reveals a glacial phase with at least three advance-retreat cycles at 79.0 ± 3.9 ka. Correlation between the ages of glacial periods and the timing of Southern Hemisphere summer insolation minima suggests that orbital forcing has played a first-order role in regulating glacial extent in New Zealand.

A possible coseismic landslide origin of late Holocene moraines of the Southern Alps, New Zealand

Abstract Moraine deposits of the Southern Alps, New Zealand, have been used to infer past climatic regimes, such as the “Little Ice Age”. Recent work has identified four major movements of the Alpine Fault since AD 1200. These earthquakes are inferred to have produced numerous large rock avalanches, some of which must have fallen onto the valley glaciers, with this debris subsequently deposited as terminal moraines. Evidence is presented which suggests that periods of moraine formation follow each of these major earthquakes. Considerable caution must therefore be exercised in inferring past climatic conditions from the size and location of these terminal moraines. These findings have implications for Southern Hemisphere mid‐latitude paleoclimatic reconstruction.

Rapid block glides: slide-surface fragmentation in New Zealand's Waikaremoana landslide

Physical modelling of part of prehistoric Waikaremoana landslide shows that the blockslide must have hit the valley wall at c. 40 m/s, after sliding 2 km on a 5.5–8° slope, in order to form the 150-m high mound of debris known as Raekahu. Both the blockslide and a distal rock avalanche were in simultaneous motion when the impact occurred. Finely ground rock on the slide plane suggests that a mechanism of dynamic rock fragmentation may explain the low friction necessary for acceleration to 40 m/s. When a rock particle fractures in a confined space, an isotropic dispersive pressure equal to the rocks Hugoniot elastic limit (in the GPa range) at the ambient pressure and strain rate may be exerted on its surroundings. Beneath the 275-m thick block, about one particle in 15–30 or so fragmenting at any instant (with lower density for higher rock strength at higher strain rate), could completely support the weight of the block by fragmentation pressure; but then there would be no frictional resistance (and hence no further fragmentation). Self-regulation of the process may explain the apparent coefficient of friction of c. 0.1 in the blockslide. Low friction through dynamic fragmentation may apply widely to blockslides with a basal layer of comminuted rock.

Determining Rockfall Risk in Christchurch Using Rockfalls Triggered by the 2010–2011 Canterbury Earthquake Sequence

The Canterbury earthquake sequence triggered thousands of rockfalls in the Port Hills of Christchurch, New Zealand, with over 6,000 falling on 22 February 2011. Several hundred families were evacuated after about 200 homes were hit. We characterized the rockfalls by boulder-size distribution, runout distance, source-area dimensions, and boulder-production rates over a range of triggering peak ground accelerations. Using these characteristics, a time-varying seismic hazard model for Canterbury, and estimates of residential occupancy rates and resident vulnerability, we estimated annual individual fatality risk from rockfall in the Port Hills. The results demonstrate the Port Hills rockfall risk is time-variable, decreasing as the seismic hazard decreases following the main earthquakes in February and June 2011. This presents a real challenge for formulating robust land-use and reconstruction policy in the Port Hills.

An early Holocene glacial advance in the Macaulay River valley, central Southern Alps, New Zealand

Abstract A 10 km long valley glacier terminating between Lower Tindill and Tom Streams, in the Macaulay River valley, central Southern Alps, New Zealand, at 8690 ± 120 years B.P. (NZ 6473A), is inferred from radiocarbon-dated deposits of till that were formerly thought to be of nonglacial origin. The glacial advance is one of three dated early Aranuian (post-14 000 years B.P.) advances in South Island. A group of undated moraines to the east of the Main Divide of the Southern Alps, collectively known as the Birch Hill moraines, may include moraines of similar age to the Macaulay River deposits,butrepresenta substantial interval of time. Available radiocarbon dates suggest that previous correlation of Birch Hill moraines with the c. 12 000 year old Waiho Loop moraine at Franz Josef Glacier is unlikely. Early Holocene glacial deposits of similar age to the Macaulay deposits are found in arctic Canada, western United States, and in the European Alps, so the triggering minor climatic change may have been glob...

Ice flow measurements on Franz Josef Glacier, New Zealand, in 1966

Abstract Surface ice-flow rates measured between April and September 1966 near the actively advancing snout of the Franz Josef Glacier show considerable variation over area and with time. Over a distance of 500 m, horizontal components range from 189 ± 4 cm/day to 68.1 ± 1.3 cm/day, and vertical components vary between 70.2 ± 2 cm/day downwards and 11.5 ± 0.6 cm/day upwards. The maximum flow rate of 790 ± 150 cm/day was measured during heavy rain in April 1966, but by May 1966 the flow in the same area had dropped to 189 ± 4 cm/day. The estimated quantity of ice discharged through an inferred cross section confirms that the glacier was much more active in 1966 than in 1956 (45,200 m3/day in August 1966, compared with 17,900 m3/day in March 1956). If the present discharge were maintained, the glacier should recover about half the length lost since the present recession began in 1951. A large kinematic wave (velocity 15.4 ± 1 m/day) suggests an abnormally high ice velocity (7.0 ± 0.4 m/day) , in which condi...

Rock avalanche deposits store quantitative evidence on internal shear during runout

We investigated the quantitative effect of internal shear on grain breakage during rock-avalanche runout, by means of 38 ring-shear experiments on identical sand samples at different normal stresses, shear strains and shear-strain rates. We compared sample grain-size characteristics before and after shearing. We found that grain size decreased with increase in normal stress and shear strain. Reduction in grain size was inferred to occur through grain breakage associated with grain interactions in strong force chains during strain. The results were consistent with observations of both inverse-grading structure in deep rock-avalanche exposures, and fining and grading of particles with increasing rock-avalanche travel distance. Our study suggested that with appropriate calibration, variations in grain-size distributions within a rock-avalanche deposit would provide quantitative information on the distribution of internal shear during its runout.