Debris flow mechanics for New Zealand mountain catchments

Abstract

Debris flows consist of mixtures of soil, rock and water that travel at high speeds down slopes – often within stream channels, but sometimes also on open slopes. They tend to occur frequently in mountainous areas when there is heavy rainfall and a good quantity of debris available, either from natural erosion supplied by weathering and small landslips, or from human activity such as logging. Debris flows may add to their volume by eroding material in their paths or may lose volume by depositing material as they travel downslope. Usually, they end on relatively shallow open slopes as a fan of debris. Unfortunately, often this is where infrastructure is built, including roads, houses and bridges.

Debris flows are common in New Zealand because of the high level of precipitation and the relatively weak rocks that make up the mountain landscape. Despite this, the public recognition of debris flows is relatively low in New Zealand because of its low population density, although this is gradually changing.

In New Zealand, there has been very little detailed assessment of debris flows in terms of how large they are, how far they run, what sort of materials are involved, how frequently they occur and how they are triggered. This knowledge is needed in order for risk assessment of these hazards to be carried out. To begin to address this, a study was conducted that included the detailed mapping of 20 debris flow events in relatively undisturbed areas (i.e. areas that had not been affected recently by wildfire or logging, for instance). The resulting data-set covers four regions of New Zealand, representing a wide-range of climatic conditions and geology. The information gleaned from this study is aimed at putting New Zealand’s debris flows in the context of the world-wide state of knowledge, so that simple empirical and statistical models can be used for risk assessment and mitigation of these dangerous events. While these field investigations are useful to place New Zealand flows in context and suggest reasonable methods of hazard assessment, they are not very useful in understanding the underlying physics of debris flow movement, as many important variables influencing the debris flow are unknown. To overcome this, a parallel study was conducted using a small-scale experimental debris flow channel housed within a geotechnical centrifuge. The small-scale of the experiment enabled debris flow behaviour to be examined in a carefully controlled manner, a situation that is not possible to achieve in the field, while the centrifuge allowed important processes to be scaled up to more closely model a larger debris flow in the field. In particular, these tests explored the influence of moisture and soil mass on debris flow velocity and travel distance. Sensors placed in the channel enabled the model flow to be tracked before it exited on the debris flow fan, while a camera was used to provide high speed footage of the flow. The results showed clear relationships between flow velocity, mass, momentum, slope, moisture and travel distance which can be used to give insight to debris flow mechanics and generate better models for risk mitigation.

---

Technical Abstract

Debris flows, high speed, gravity-driven mixtures of soil, rock and water, are ubiquitous mass-wasting processes in areas of high relief and rainfall. New Zealand’s position in the mid-latitudes of the Pacific Ocean results in periods of high-intensity rainfall – leading to high rates of physical weathering. Combined with extremely high rates of uplift and highly indurated, fractured bedrock, these factors result in a particularly high temporal occurrence of debris flows. Despite the danger they pose, the public recognition of the losses wrought by debris flows is relatively low in New Zealand. This is largely a result of the country’s low population density, especially in regions most prone to them. This situation is gradually changing as hilly and mountainous terrain, once thought marginal, is developed.

Despite some notable examples of large debris flows, such as that which occurred at Matata in the Bay of Plenty in May 2005, the state of knowledge on New Zealand specific debris flow occurrence is sparse. While there has been some qualitative work done on specific cases, there has been very little detailed quantitative assessment of these hazards in New Zealand. To address this, a detailed field investigation of twenty debris flows covering four regions within New Zealand was undertaken. The flows examined were located in the North Island in the Southern Rimutaka Mountains and in the South Island, near Cass, Mt Cook and Franz Joseph Glacier. These flows cover a wide range of debris flow types, from the bouldery, channelized events to littlestudied, smaller, hill-slope debris flows. They also cover a wide range of geoclimatic conditions, from extremely high-rates of rainfall west of the Southern Alps to the comparatively drier conditions east of the divide. The methodology used to map the debris flows was adapted from that used extensively in British Columbia, Canada. The method examines the flows on a reach-by-reach basis, with details of erosion anddeposition being mapped as well as slope angles, channel widths, and flow thicknessbeing measured or estimated from field evidence. Observations of constrictions in theflow path, the entry or exit of stream flow and other details help to provide a detailed picture of the history and path of each flow.

Values of total deposition in the data-set range from 10,000 m3 to 300 m3, which is typically smaller than most of the event magnitudes discussed in the debris flow literature. The detailed investigation of small, non-anthropogenic induced events makes this New Zealand data-set unique. The results of the data can be further utilised in developing empirical models of runout for New Zealand specific conditions and can be compared with statistical-empirical models developed in similar conditions elsewhere. Details of each individual flow within a dataset can be examined to assess the influence of moisture and flow geometry on the overall behaviour. The aim is to determine the specific mechanisms that lead to departures of behaviour from the average within a locality and therefore to better understand the risks and uncertainties within each dataset.

While forensic field studies are invaluable in describing the flow behaviour typical of an area, they are less useful in elucidating the underlying mechanics of the behaviour, as many important variables such as moisture content, the geotechnical properties of the material, and exact volumes are extremely difficult to determine after an event. Therefore, a series of physical modelling experiments using a small-scale flume housed within a geotechnical centrifuge was undertaken in parallel with the field component of the study. The focus of this study was to better understand the mechanisms underlying debris flow behaviour in general, by measuring and varying parameters of interest and carefully observing the result.

The experiments were designed to examine the effect of varying the moisture content and flow volume on debris flow velocity, discharge, and runout. The effect of porefluid rheology was also examined by using both Newtonian and non-Newtonian pore fluids. Detailed measurements of pore pressure were taken at positions along the flume, while a high speed camera was used to capture the flow behaviour close to the exit of the confined flow to the unconfined fan area. Measurements were also taken of the thickness and extent of the deposition area and the flume bed after each test.